Startups, company culture, technology, writing and life

Page 2 of 2

iPad: an Apple for Mom

To much of the geek-sphere, the iPad is somewhat of a disappointment. It’s underwhelming – there’s nothing all that surprising in there, no “oh, and one more thing” to throw some real techno-magick spices into the pot. We saw it coming from a hundred miles away, so now that it’s finally here, there isn’t much surprise left.

I think part of the reason for this is, as usual, that the first round of people interpreting the meaning and purpose of the iPad are all geeks – developers, designers, and assorted Macheads who already own one or two macs at least, and an iphone. I believe the collective “Meh” is absolutely right from this crowd. Yes, the iPad is a nice-to-have, yes, I’ll probably be getting one, but it’s not really all that revolutionary. It doesn’t do anything that we can’t already do with our iPhones and Macs, after all.

But geeks and assorted Macheads constitute a relatively small percentage of the computer-using people around the world. The vast majority of the world is still using Windows PCs. And for them, an iPad may be exactly what they’ve been waiting for. Let me elaborate on this…

Most people don’t need a computer

Most people have a computer at home. For some (like my mom), it’s an ancient Dell laptop they bought years ago. Others have bought into the netbook trend and invested a small amount of money into a machine that, for most people, can only be irritating to use (slow, small, ugly, and burns your lap – not a recipe for success). A few have actually paid good money for what was supposed to be a modern machine, and actually turned out to be yet another annoying slow, painful-to-use, Windows-based machine infested with trial software, spyware, and sometimes viruses.

In today’s (western) world, not having any computer at home makes life difficult. My mom needs some way to check airline ticket prices, to find out the weather, to go on Facebook, to buy movie tickets, to check her email, to call me on Skype, and a thousand other little uses that aren’t very taxing or challenging for either her or whatever device she’s using.

She doesn’t really need a computer in the same sense that I do, though. As a programmer, I need a machine that is powerful, that I can mess with under the hood, that I can do everything with. My mom needs a reasonably priced machine that Simply Works and does all those simple things that she wants to do when she’s at home.

In short, most people don’t really need a proper computer at all.

And they mostly don’t want one. Many people spend all day working in front of a computer, and they simply don’t like the idea of coming home to yet another computer that looks just like the one they use at work.

iPad: the uncomputer for the masses

Apple has grandiosely claimed that the iPad is creating an entirely new product category, and I think they’re right.

Many people are comparing the iPad launch to the iPhone or the iPod – dismissed by tech critics, who were ultimately proven wrong by mass adoption of the devices. There is something to this comparison, but it’s not quite fair. Both the iPod and the iPhone attacked existing product categories and made them so much better that they wiped the floor with their competition. The iPad is attempting to create a new market. A great many of the people who bought an iPod had other MP3 players beforehand. I’d wager most of the people who end up buying an iPad have never owned a dedicated tablet computer, and never considered that it would be worth owning one.

A better comparison is with the Nintendo Wii. While Sony and Microsoft competed in the cut-throat market of consoles for gamers, the Wii also created a new product category: consoles for everyone else. It worked pretty well for them – it turns out that there’s a lot more non-gamers than gamers, and making a device that appeals to 95% of the population sells better than making one that appeals to only 5%.

And that’s exactly what Apple is doing: making a slick “uncomputer” that’s tailored to those people who don’t actually need a computer. Many gamers ended up buying Wiis too, and I’m sure many geeks will buy iPads, but the real money-maker will be those who don’t even have a Mac, and probably won’t ever have one because it’s too expensive and they don’t need it.

The price point is also just right for this market – $499 is around the range of a cheap laptop or an expensive netbook. The applications balance is right – focus on email and web, with a good helping of medias, games and various practical apps. The gorgeous look is of course going to appeal to everyone. Once it has a video camera (inevitable for the next iteration), it will do everything that normal people do with their laptops at home today.

The only question, in my my mind, is, what will these people do when their cheap old Dell finally clonks out? Right now, to use an iPad and iPhone effectively, it seems you still need some kind of base station. So when the old Dell gives up the ghost, will people buy another one? Pony up for an expensive Mac? Or simply decide that the iPad is good enough and they don’t want another laptop?

What problems does Google Wave solve?

There are countless pundits and other tech gurus describing Google Wave as a disappointment, lately. Most of that seems to come from the fact that nobody seems to get what Wave is for. So they compare it to social media.

Is Wave the next Twitter? Nope. Is it the next Facebook? Nope. Is it going to replace Instant Messengers? Possibly, in some circumstances, but not any time soon.

I believe this is partly Google’s fault: they released Wave to geeks and hackers and social media folks first. But Wave is not a geek/hacker tool, or a social media tool, it’s a corporate tool that solves work problems (more on that later). On the other hand, they never claimed it would be a Facebook replacement or a Twitter killer. Google calls wave an “online tool for real-time communication and collaboration”. The way Google should have advertised Wave is: “it solves the problems with email”.

At Woobius, we’ve been working at resolving the problems with email for some time. Woobius is a solution to some of the problems of email within the construction industry. We’ve blogged and given talks about it. Perhaps that’s why it was immediately obvious to me and my team why Google Wave is awesome.

What’s the problem with email, anyway?

To most geeks, the main problem with email is spam. They don’t have a problem with online collaboration – they use Google Docs, Etherpad, Skitch, screen sharing tools, or any number of collaborative whiteboard applications. So the main problems for geeks are that they’re signed up to so many services that they get inundated with notifications, monthly newsletters, automated messages, and shreds of spam that manage to get through GMail’s spam filters. But when they want to collaborate on a document or picture, they can find the tools they need, most of the time.

But then again, most geeks don’t do all that much document-based collaboration, by email or otherwise. Programming doesn’t require a whole lot of collaboration, beyond that provided by source control tools and bug tracking system. Being Robert Scoble probably doesn’t require you to spend days working on a specification document for some finicky aspect of project X, or at least not very often, and he’s probably not the one collating everyone’s suggested changes and resubmitting the document for further review.

In your average corporate environment, though, this happens all the time. People work on documents, presentations, etc. They have lengthy discussions over email. Pieces of work bounce back and forth across one or multiple organisations for weeks before they’re finalised. People are brought on to the conversation late in the day. Attachments get lost. Inboxes fill up and emails bounce. It’s a major pain.

So what are the problems with email in a corporate environment, and what does Wave do to address them?

Problem 1: Collaborating on a piece of text

It’s hard to use email like you would use, say, Etherpad – to collaborate on a document that later needs to be sent out. Most such collaborations end up being done either via a Word document with change tracking, or, when they’re more ad-hoc, via a long thread of email with corrections coming in from every direction. It’s a nightmare to keep track of and collate all that feedback. Even giving the feedback is difficult sometimes: you have to quote the context and make sure your change is clearly outlined.

Google Wave resolves that by effectively integrating Etherpad’s features into the email client. Putting an email to an important client together, with feedback from the team, becomes a breeze.

Problem 2: Adding new people to the conversation

With a typical email thread, you can forward the whole thread to a new participant, or add them into the next reply, but they’ll only get a garbled, over-indented mess, in reverse chronological order. If you’ve ever been added late into an email thread that had already been going on for a week and involved two dozen replies, you know what I mean.

Google Wave solves that by giving exactly the same view to everyone, regardless of when they’ve been added.

Problem 3: Keeping added people added

Many times, when you add new people into a conversation, they get dropped again later, when someone replies to all from an earlier email that didn’t include the new participants. Sometimes it takes a while before you realise that key people have been dropped out of the conversation. That costs time and hassle both for the people who were dropped and those who weren’t.

Wave solves that by making “dropping people” an explicit action, rather than something you can do by mistake.

Problem 4: Attaching files

Most large companies have an email storage problem, so they limit the size of people’s mailboxes. Because of that, it’s not uncommon to see “Inbox full” bounces when sending large documents around. Not only that, but sending documents is iffy at best. The SMTP protocol doesn’t seem to be all that good at sending large files.

Now, to be fair, Wave will probably suffer from the same limitations as any HTTP upload applications – but that’s still a whole lot better than your average email. Sending emails over 10MB usually fails. Attaching a 10MB file to a Wave is no problem at all.

Problem 5: Lost attachments

When you reply to an email with an attachment, the attachment is dropped. This is a good thing with email, because it stops a single email thread from unnecessarily clogging up both the mail server storage and its bandwidth. Since the whole email is transmitted down the wire when you click “send”, this kind of limitation is unavoidable.

What this means, however, is that if you bring new people into a conversation, by adding them as recipients or by forwarding them the latest mail in the conversation, they won’t get any of the attachments. Not only that, but if you’re looking for that first attachment, and the conversation has been going on for weeks (and, like everyone else, you receive upwards of 50 relevant emails a day), finding that attachment can be quite difficult. If there were multiple attachments throughout the life of the discussion, gathering them all to send them to a new participant is exponentially more tedious.

With Wave’s model, however, the attachments stay there, where you put them. They’re only sent down the wire, from you to the email server, once. You never need to re-forward an attachment to someone. When you add new people to the conversation, they get access to all the attachments right away.

Problem 6: Multiple conversation branches

Email conversations are, basically, flat. If you try to have multiple branches of conversation in email you end up with a sordid mess. You might do that a few times in your life, but you quickly learn not to. But flattening everything has its own share of problems – every email ends up containing replies to several other emails. It becomes very difficult to track what was replied to and what wasn’t. And it’s hard to collate all the suggestions effectively.

Google Wave resolves this by allowing clear, obvious threading. Yes, if you use a lot of threading in an instant messaging context, it will be hard to manage. But within the typical email collaboration context, it will keep things a lot more clean and tidy than not having threading.

Problem 7: Small corrections

With email, if your only comment on someone’s email is to fix a dozen typos, you still have to do almost as much work as if you were making substantial changes to their proposed text. You need to quote the context, highlight which bit you corrected, and then rely on the other person applying your changes back to the original document (which they often forget to do — after all, it’s just a few typos).

With Wave, no such problem – you can just edit the original text and make those changes. If the person who submitted that document wants to review your changes, they can play them back.

Problem 8: Email to IM to Email

Instant Messaging is a powerful, useful technology that has proven its worth. But it’s not very well integrated with email. If you rely on your inbox to keep track of conversations, there’s still this gaping black hole of IM which is tracked somewhere else (if at all). GTalk tried to resolve that by storing IM conversations in your inbox – and that was a good step.

What Google Wave does, however, is much bolder: it recognises the fact that a lot of IM conversations, in corporate environments, begin with an email exchange that’s just getting too rapid. When you send more than 3 emails to the same person in one minute, it usually makes sense to either pick up the phone or IM them. With Google Wave, this doesn’t need to be a conscious decision: if you’re replying quickly, Wave smoothly turns into an IM-like platform. When your replies get slower again, it, once again smoothly, turns into an email-like platform.

This means that the whole conversation, whether email-like or IM-like, is tracked and searchable in the same place, and visible to all those who are invited to the conversation.


I believe that people who don’t see what Google Wave is for are simply looking at it from the wrong angle. Wave is not a social tool. It’s not Twitter, it’s not GTalk, it’s not Facebook. It was never designed to appeal to the crowds of geeks who are currently trying it out.

Wave is built for the corporate environment. It’s a tool for getting work done. And as far as those go, it’s an excellent tool, even at this very early stage.

It will probably take years before Wave fully penetrates large corporations and replaces the email systems everyone is used to. But it solves so many thorny problems with email that it might well manage to do so, where so many other tentative “email fixes” have failed.

In the meantime, we should stop judging it as a social tool and start looking at how we can use it for real work. Invite your colleagues to it, and get working.

A tale of two users

Eric von Hippel’s concept of early adopters versus lead users is well established. How does it work in practice, though? And why is it so hard to cross that famous chasm?

Let’s have a look at how this process can unfold, looking at a real life example.

It is Easter weekend, a few months ago, and my friends and I have just had a delicious Chinese hot pot. As we sit around and talk of everything and nothing, the conversation suddenly focuses on Twitter.

“Twitter seems like unwanted noise,” Bob declares, “I don’t need more noise in my life. I can’t see any use for it and I don’t think I’ll ever use it.”

Others echo his feelings.

“Believe me,” I reply, feeling the weariness of a conversation I’ve had a hundred times with a whole variety of people, with little result, “Twitter is worth your while. You can use it to network with people, to find out what’s going on, to get answers to obscure questions, express interesting thoughts… it’s very versatile, and the way you end up using it won’t be quite the way that I use it, but you will use it.”

“Maybe Twitter is only useful for people like you, who feel a need to express themselves and communicate with lots of people online. I don’t need that extra distraction in my life. I’m happy with the way things are.”

Here we go again…

Struggling for the right words

Why is it so hard to convince people that an application (like Twitter) is good for them?

I’ve been using Twitter for over a year. Back in those days, I was skeptical too. But some people that I trusted told me, “look, just try it, it rocks.” I did try it, and they were right. Now nary a day passes when I don’t tweet about things ranging from articles that I’ve read and liked to interesting ideas that have passed through my head. I have 500 followers, and I follow as many. Twitter feels as much a part of my social world as my mobile phone does — perhaps more.

I am an active, avid Twitter user. Not a “Social Media expert”, desperate to gain huge amounts of followers. Not a celebrity with a million followers. Not a Twitter obsessive who tweets a hundred times a day. I’m a normal human being, using Twitter actively, but with moderation, for what it is, with no specific goal (marketing, business, networking, etc) in mind.

And to me, the value of Twitter is so blindingly obvious that I can’t quite explain it convincingly to others. Shortly after MiniBar, a month or so ago, I was observing to a group of entrepreneur friends:

Isn’t it funny how at the same time, in one circle of “real life” friends, we struggle to convince people that Twitter is something worth their while, yet in another circle of more connected friends, the fact that Twitter is worthwhile is so obvious that to even bring up the subject is basically boring and tedious, lame old news not worth discussing?

That, in fact, I believe, is the problem. Many of us hackers are early adopters. We’re willing to jump into a new service on a whim, and figure out what to do with it as we go along. We can see the potential of something as simple and basic as Twitter even without being told all the different ways we should use it. We can make our own path. To us, Twitter is a huge blank canvas that we can paint in whichever way we want, and we love it that way.

But because of this vision, we find it hard to distill the essence of what Twitter is to others. Because to us, there is no singular essence, but a cornucopia of possible uses. But this is not how other people function, they want concrete benefits before they will dip their toe. By the time those more obvious benefits finally emerge, however, we’ve lost interest in trying to convince others, because we’ve already been trying so damn long.

And that’s fine, because, here’s the good news, we’re not alone in this dynamic.

Lead users

The typical product adoption curve is broken up into multiple segments. If we’re the early adopters, who comes next? The leaders of the early majority, the people who won’t put up with a product that doesn’t work, but are willing to dip their toe when the time is right, are called “lead users”. They’ll only use a new product when they can see a use for it, but once that use becomes obvious, they’ll jump in and start to adapt the product to the uses they can see.

It turns out, Bob is a lead user. Shorlty after declaring his eternal distaste for the additional distraction of Twitter, he gave in and created an account. Like everyone else, he ignored it for the first month. And then, one day, he started using it in earnest. He spotted a use for it, when he noticed that a lot of influential architects (people who are early adopters in the architectural space, like him) are on Twitter.

Once he discovered that, he took to Twitter like a fish to water. Soon he was following a hundred people, and a lot of them followed him back. He made the effort of tweeting once or twice every day. All the while, he declared that he was only using it for business, and couldn’t see any personal purpose to it, that maybe this was just my cup of tea and not his (does that discourse sound familiar yet?). Yet he used it, more and more.

And something else happened. Bob started convincing others to join Twitter. In the few weeks since he started using Twitter in earnest, he’s convinced 3 times as many people to join Twitter as I did in a year. Sure, they’re not all that active yet, and they mostly have a very small circle of followers/followees, still, but I’m not worried. They’ll get there (such is the power of Twitter). Everyone eventually gets there, once enough lights blink online in their social network.


So why is this worth thinking about? What can we learn from this?

Well, early adopters like myself can learn from people like Bob how to be better at convincing the next wave of people to adopt a new product or technology. If you’re an early adopter, you can probably think of a few people who tend to eventually adopt new technologies and gadgets that you’ve unearthed in your jaunts through cyberspace (or meatspace). Talk to them. Find out how they convince others, and you could become more effective at spreading the message when you want to.

That’s an important caveat, because often we don’t really want to. For the most part, although we, early adopters, do care about convincing our friends to follow our lead towards progress and the greater good, we don’t care that much. We’re far more interested in uncovering the next new fun thing and painting on that blank canvas than in spending all that tedious effort figuring out what will convince the masses to adopt something that we discovered a whole six months ago. But we do care enough to try, at least with our close friends.

More importantly, I think, lead users can learn from this. If you’re a lead user, identify who the early adopters are in your social circle, and pay attention to them. They’ll lead you on to the next big thing. What you’ll get out of it is that you’ll be able to catch those trends earlier and be even more of an influencer of the people who revolve around you.

And please, please, never declare that you see no use in something new that they bring you.

Otherwise, someone might bring up your comment about “not using Twitter for personal purposes” when you’re caught tweeting pictures of defrosting watermelons.

Many thanks to Bob Leung and Dale Harvey for reviewing drafts of this article.

How to make your application viral

Image courtesy of flushirts

Everyone wants their application to “spread virally”. And why shouldn’t they? Viral growth resolves at least part of the expensive and complicated headache of actually marketing your application, by getting the application to grow all by itself. So, then, the question that forms on the lips of any entrepreneur is:

“How can I make my application viral?”

There’s a misconception built into that question: the idea that virality is something that you can just slap onto any product, like a magical pixie dust that will suddenly grant your application the gift of users.

“Making an application viral” is as silly a proposition as “making a car fly”. Sure, it’s technically possible, but that requires a lot of fundamental changes, and what you end up with is not a car anymore (or perhaps it was never a car to begin with).

However, what you can do is set out to build a plane in the first place. There are some building blocks that you can use or omit in your application design that will enable or hinder your application’s virality. This article explores some of those building blocks, with the aim of at least helping you make your next application spread virally1.

Viral basics: the viral loop

There are plenty of articles talking about the viral loop, so I’ll cover it quickly in this article. The main focus is not the theory of viral growth, but some practical things you should keep in mind while designing your product, so that this viral loop does work for you2.

The cornerstone of all discussion of the viral loop considers the cycle from acquiring a new user to having them invite others. In equation form:

viral coefficient = (average number of users invited by each active user who invites someone) x (proportion of invited users that actually join or become active) x (proportion of active users that invite others)

or, using variable names that we’ll refer to later:

VC = N x P1 x P2

If your viral coefficient is greater than 1, then over time your growth will increase exponentially, and you will saturate your market (or whatever parts of it you have access to, if your market is highly fragmented). If it’s smaller than 1, your growth spurts will always end and you will have to keep pumping marketing energy into your application to grow it (which, as this article points out, may not necessarily be a bad thing for some, since it puts you in control of your growth). If it’s 0, you won’t get any viral growth at all.

This idea of a viral coefficient is very powerful. If you can measure it, and break it down into the correct factors for your application, it can tell you what specifically you need to improve in order to increase your viral spread. So it’s definitely worth measuring this coefficient, even if your application is already spreading virally.

Practical virality

So, with that in mind, how do you actually get your users to invite more users? The brutal, dirty answer would be to slap on a form that asks for their email password and hoovers up their contacts. Many early Facebook applications used similar tricks (in fact, Facebook themselves did so), and the result of that was incredible growth, but also a lot of spam. Facebook soon shut down these holes as far as applications were concerned, and they were right (if a tad hypocritical) in doing so. In my opinion, if you rely on abusing people’s contact lists for your viral growth, you’re only a few rungs above ordinary spammers. Many reputable companies still do it today, but I don’t think this is something they should be proud of.

Before we go into the principles of how to build a viral application, it’s worth emphasising this again: viral growth does not and should not require dirty tricks. I do not condone any form of spam or other abuse of your users’ precious trust, if only for the reason that once users stop trusting you, it is very difficult to regain that trust – as far as business applications are concerned, this sort of behaviour can kill your reputation dead. None of the principles below require any form of underhanded behaviour to be effective.

The golden rule of the ethics of viral spread is: try to only do things on a user’s behalf when they’ve explicitly done something to request that thing, and they know that what they’ve done will result in a communication being sent on their behalf. If you can’t link an invitation directly to the inviter’s action, then you probably shouldn’t send that invitation.

Virality principle 1: Inviting as a core process

Once upon a time in 2007, I was tearing my hair out as to why my Facebook application (now defunct) just wasn’t spreading, and I cornered R. Tyler Ballance (who created the highly successful “Top Friends” facebook application) into a private chat on IRC, to exhort him to give me the secret of building a viral app. Although he didn’t quite give it away, he did provide a handful of hints that pointed to one of the key answers.

The first step in increasing your Viral Coefficient is to increase N, the average number of users that each of your active users invite. The most effective way to do this is to ensure that inviting other users is a core process in your application, rather than something that people are asked to do as an afterthought.

In other words, the most viral applications are those that involve inviting others to the application as part of the daily usage.

Update: Some people have pointed out that this section is not entirely clear. What I mean is this: certain kinds of applications (e.g. collaboration tools) lend themselves to the concept of inviting others, as part of the core usage of the application. If you can architect your product so that it is one of these applications, that will greatly enhance your viral spread. This cannot be latched onto any application. It needs to be as integral to your application as “getting people wet” is integral to the use of a water pistol.


Top Friends did that extremely well. The core usage of Top Friends is to select your top friends and rank them. As part of this process, you are naturally asked to tell the friends about this ranking business, and thus they are invited to use the app themselves.

Woobius (my business) has also integrated the invitation process successfully: part of the core usage of the application is to invite collaborators to a project.

Hotmail, the grandaddy of them all, was one of the first mega-viral applications because each email included a small advert at the bottom, that invited the recipient to join hotmail.

Paypal also hit the nail on the head by allowing users to send money to people who were not registered yet (they later doubled up on that by giving a referral bonus if you got the recipient to join). Paypal is about sending money to people, and inviting them as part of the money-sending process was a stroke of genius.

Virality principle 2: Keep pulling people back in

It’s not enough to invite people, of course, you also have to get them to actually join up (coefficient P1 in the earlier equation). There are several ways to do that, but one of the essential, practical things that your application should do to get them to join is be persistent.

What this means is that you should not just send them one invitation and then wait for them to join. And once they’ve joined, you should not just wait for them to start using your system and inviting more users. You should keep sending useful communications and/or showing signs that will encourage people to use the system. Of course, this also needs to be a core process of using the application. You can’t send people reminders all the time without a good reason – that would be spam.

When you pull people back in, you should not only remind invited people that they can join, but also remind your active users that they can invite more people, and that there are benefits to doing so.


Hotmail, again, got that exactly right. Since every email included an invitation to join Hotmail, and promoted hotmail by coming from an @hotmail.com address, every email implemented the “keep pulling” effect.

Mob Wars, a Facebook game, applies this concept by constantly reminding its users that new features will be unlocked if they invite more of their friends.

Facebook itself does this extremely well. I can’t count how many times I’ve drifted away from Facebook, only to be dragged back when someone posted to my wall or sent me a message. Because of this, I (and many others) still pay attention to Facebook every week or two, even though I don’t actively use it.

Virality principle 3: Be useful even with no other users

This is also known as the “chicken and egg” problem. It’s what killed my Facebook application, back in the days, because that application was useless unless some of your friends had also signed up. It’s possible to succeed despite this problem, of course, but it’s much, much easier to spread without it.

This doesn’t mean that your application shouldn’t benefit from network effects (i.e., the effect of becoming more and more useful the more users join). It does mean that there should be some modicum of useful functionality that works and is immediately useful even when none of the user’s contacts have signed up. Even in applications like Twitter or LinkedIn, which are really all about the network effects, the fact that there is something meaningful for the user to do right away, even when they know nobody, is a huge plus.

Most users will not invite other people when until they’re familiar with an application. If you don’t provide them with something to do before they’re ready to invite others, you will probably lose them long before they reach that stage.


Twitter, funnily enough, got this somewhat right. By presenting itself as a micro-blogging platform and providing a “What are you doing?” prompt, it provides an immediate piece of functionality that users can start using even when they don’t know anyone else on the network (even though it is a little demoralising for some people to write mini-blog posts that no one will read).

Woobius achieves this because a new user can use it to send files to other people even if those other people have not joined. Similarly, the revision tracking functionality works well even if only one person uses it.

LinkedIn, despite being mostly used for its connections, provides an immediately compelling functionality for new users, by allowing them to recreate their CV online (which is a big deal for many people who are not so tech-savvy).

Virality principle 4: Remove artificial invitation limits

If you’ve ever built an application with viral spread, you’ll have noticed something that’s not very obvious when you look at the Viral Coefficient equation. At first glance, it would seem that growth should be smoothly distributed… each user invites an average of 10 other users, of which, in turn, 20% end up becoming active users and inviting another 10 users. It all seems quite easy and… smooth.

Except it isn’t. In practice, most people invite no one. Just like with other human relationship networks, invitation patterns have a few strong focus points, highly connected nodes that result in dozens or hundreds of invitations, and a lot more weak end-nodes who rarely invite anyone.

We’ve noticed that on Woobius too: most people invite no one, but a handful will invite 40 people at a time.

What this means in practice is that the last thing you want to do is to limit how many new users each existing user can invite. This is an important lesson, because when designing Software-as-a-Service applications, it seems very reasonable to charge based on how many users use the system (after all, many leaders in the field do so). Even easier, one might be inclined to make the “free” version support only one user.

One way to resolve this problem is to ask yourself what would happen if someone who just joined my application thought it’s great and wanted to invite 20 of their friends/colleagues? If you can’t provide an easy way for that person to market your application for you, you’ll lose out on your best inviters.


MinuteBase, a recently released minute-taking application, started off with a limit on the number of users, but decided to remove that limit. After only a couple of months live, this has already resulted in faster viral spread than the initial concept of limiting free accounts to a single user.

As a counter-example, most products built by 37-Signals do the exact opposite. They limit the number of users (so you’re unlikely to invite someone unless you really must), and even force their own users to fragment into multiple subdomain. This obviously works for them, so I’m not criticising that as a way to do business. However, from the point of view of viral growth, it is very ineffective, and it is likely that 37-signals gets most of its users not through viral spread but through their more traditional marketing activities (their blog, their conferences, and, of course, their link to Ruby on Rails).

Summary and conclusion

There are many other things you can do to tweak the viral coefficient and improve the spread of your application, but carefully considering these points will get you further than most:

  • Invitation should be a core process, that is essential to using the application – this will maximise the chances that your users do invite new users.
  • Keep pulling people back in, rather than letting them forget you after the initial invitation, and make this “reminder” process also be central to the use of the application.
  • Be useful even to the lone user, because that lone user is the source of all your other users.
  • Remove artificial invitation limits, to recognise the reality that most invitations come from a few very active users, and help those users spread the word.

These concepts, as I mentioned earlier, cannot be applied to all applications. Some applications are simply not going to spread virally. But if it is possible to tweak the application design, early on, to accommodate these principles, the result should be increased, viral, self-sustaining growth.

Many thanks to Bob Leung, Tom Hastjarjanto and Cliff Rowley for reviewing drafts of this article.

1 Some existing applications may even lend themselves to retrofitting wings and turbines. Your mileage may vary.

2 It’s worth noting that not all apps have to be viral. There are many ways to achieve high user growth, and virality is only one of them (though it is certainly one of the cheaper and more scientific ones).

Dealing with impossible crises

In large corporations, almost everything new is impossible. Try to do anything new, and typically you are met with dozens of reasons why it can’t be done. As a consultant (which I was throughout my time in the corporate world), however, you’ve been hired to get something specific done, so you don’t get to echo the “it can’t be done” line back to your client. Your job is, effectively, to do the impossible.

What happens after you do this for a few years? Well, you gain a healthy disrespect for anyone who tells you that something is impossible (because it very rarely is), and a productive attitude to dealing with such situations. Some things really are impossible (so far), like teleporting to the other side of the world, or bringing someone back to life, or traveling back in time. But most situations deemed “not possible” are nothing of the sort. The lessons I’ve learned about the nature of impossibilities from my years as a corporate consultant are also useful in start-up life.

Now when I run against so-called impossible tasks, the mantra I usually repeat to myself is: “There are very few things that are truly impossible. This is not one of them.”

A practical example of getting the impossible done

Last Christmas, my girlfriend and I were going to spend Christmas in Geneva, with my parents. We bought our tickets in September, booked a hotel to stay in, and, on December 24th, turned up at London City airport to climb on the plane. What we didn’t know (and what Air France failed to inform us of) was that Switzerland had just entered the Schengen area three weeks earlier (why they timed that for three weeks before Christmas, I don’t know). This made no difference to me, but my girlfriend, being Taiwanese, now needed a visa to go to Geneva.

Of course, the airline wouldn’t let us on the plane. The Swiss embassy in London was closed for the holidays, so there was no way to get an emergency visa. This was a typical “it can’t be done” situation. French companies are notoriously procedural and uncooperative. Air France proved to be just that. They wouldn’t help us, beyond giving us the phone number of the Geneva airport border police. The Swiss border police, predictably, had no intention of breaking the Rules to let in a couple of people who didn’t arrange their affairs in advance.

The situation looked lost, Christmas was ruined, my girlfriend was close to tears, suggesting that I go by myself and leave her alone in London for Christmas.

In the end, we made it through. My father, from Geneva, inexplicably convinced the border police, on the phone, and without any prior contacts, to give us a temporary visa upon arrival (impossible!), and I convinced the uncooperative Air France to bend ever so slightly (unthinkable!) and actually bother to call up the Swiss border police to confirm that we could board (at first, they insisted on requiring the border police to send them a fax confirmation). We still missed our plane (largely due to the inflexibility of Air France, who I shall not be flying with again), and had to pay a surcharge to get the next plane1.

But we made it to Geneva, and that was the most important thing. Christmas was saved — in fact, we had a great Christmas holiday.

Principles for doing the impossible

Overcoming these brick walls in the airport required the same approach as it did in large corporations. I later discussed it with my father, and have broken it down below into a handful of essential practices. Some may be obvious, some may be less so. All can be useful when trying to get your way and facing the faceless brick wall of “it can’t be done2”, though they’re not applicable to all situations.

1. Calm down, smile and stay polite

You cannot win this kind of battle through anger. The very first thing to do is to observe that you are angry, and calm yourself down. Acknowledge your anger, and put it aside for later. There’ll be plenty of time for cursing the system that put these problems in your way later, after you’ve removed these obstacles. The surest way to guarantee that you won’t get your way is to get angry. Angry people are always wrong, and they’re rarely worth helping or cooperating with.

When faced with sudden, unexpected obstacles, it’s easy to let your emotions flare up, but in most cases, the other side doesn’t have to help you, and a heated exchange may even ensure that they actively hinder you.

Next, you need to make a positive effort to stay polite and pleasant. This is more than just “acting normal”. You need to be extra nice with all the people involved in the problem you’re facing, because any of them might be the key to helping you move past that problem.

As a start-up, you have very little power to bully people who throw brick walls in your way, so when you hit such a problem, remain apparently cheerful and polite even if your business is going up in flames in the back of your mind. Rule 1 is: smile and stay polite and positive. You’ll win more with that attitude than any other, particularly when there’s plenty to be emotional about.

2. Become a human being

It’s easy to say no to a statistic, or a number, a traveller, a buyer. It’s easy to ignore a ticket in a queue. It’s much harder to ignore a real human being with a real, personal story, someone you’ve met, who’s looked you in the eye, and spoke to you one-to-one.

Chances are the system that is blocking you is not treating you like a person, but like an item in a queue of things to be dealt with. To get people within that system to help you, you need to snap them out of that world-view, into considering your case as a real, personal, human situation.

The easiest way to achieve that is to provide them with personal details. No, I don’t mean your name and address! Provide some elements to build a story of what is happening to you. You’re not just two faceless entities who want to get on a plane without a visa, you’re a couple traveling home to spend Christmas with your parents. You’re not just yet another customer trying to solve some production issue, you’re an inspiring start-up preparing to go-live with your first and best customers.

Add as many personal details as reasonable to become a real, live human being in the mind of the people you converse with.

3. Be persistent

There’s a saying that the “squeaky wheel gets the fix”. Persisting, and conveying that you’re probably not going to give up until you get your way, makes it more likely that even uncooperative people will help you, if only to get you out of their way.

It’s of paramount importance that you continue to be very pleasant, friendly, and polite, while firmly persisting. It is socially difficult to forcibly remove or ignore someone who is polite, positive and friendly, even if they’re continuing to insist on something past the point where they should have reasonably given up. The minute you frown and raise the tone of your voice, though, you’re a goner.

As a start-up, simple polite persistence is often enough when dealing with banks (when you want to get a merchant account, for example), clients, or suppliers who are ignoring you.

4. Be prepared to lose

If you are not prepared to lose, you probably won’t win.

To get what you want, it is, paradoxically, more productive to start by assuming that you have possibly lost already. This results in a psychological shift where you go from trying to defend what you thought you had (which can make you overly emotional and defensive) to trying to gain as much as you can from a bad situation.

This gives you a freedom of both thought and action. You can think laterally and come up with unexpected compromises and pragmatic solutions, and when you see an opportunity, you can seize it without hesitation.

5. Define your objective clearly to yourself

In the airport example, our objective was to get to Geneva within the next day or so. In a corporate work environment, your objective might be to get a production bug fixed today despite all the procedures that say that you need to wait 3 months and get twenty approvals. In a small business environment, you might want to save the relationship with a client who suddenly blew up, called you names and cancelled their account.

You should phrase this objective in terms of results, not in terms of how you’re going to get there. Doing so will also help free you to come up with workable ways to get there, rather than try to make things happen the way you originally thought they should have happened.

6. Find who can

Often, the person you’re dealing with right now cannot help you. What you need is impossible to them… but not to someone else. Perhaps someone in another department, another institution, or another country can help you. You can tease out who you need to talk to with questions like “Could anyone change this?”, “What do you need to do this? Who can provide it?”

Suddenly, the authorisation that was previously impossible to grant may turn out to merely require the supervisor to sign a piece of paper.

7. Take an active part

In these kinds of situations, things often fail to happen because of the inertia of interactions between people. You can enormously enhance your chances of success by acting as a go-between, the social grease between people. If you get a response like “Oh, I’m going to have to send an email to so-and-so, and I can’t do anything until I get the reply,” offer to send the email yourself (and then probably follow it up with a phone call). You’ll shave hours off the time from zero to done.

Use this opportunity to escalate and widen your contacts within the power structure that you’re trying to bend to your will. If you actively collect those contacts, you may soon find that you have more access to decision-makers than many of the people who you are dealing with (“You need a signature from a director? I just spoke to one, give me a few minutes.”).

8. Make the other person feel good

This is particularly important when the situation is such that you are in a position of overwhelming inferiority (for example, in the airport, or when facing an executive secretary). The natural tendency is to make the other person feel bad for not helping you, and that may work sometimes, but it can often backfire because that person then associates feeling bad with your presence, and may try to get rid of you without helping you.

It is much more effective to present the situation in such a way that the person who can help you will feel that they are doing something Good by helping you. So, taking the airport example, don’t make the person at the ticket desk feel that they will destroy your Christmas if they don’t help you. Instead, make them feel like they will have been the most essential actor in saving someone’s Christmas by helping you.

Another example: let’s say you’re dealing with a supplier who has treated you badly. You have every reason to be annoyed at them, but if you really want to get what you want, you should present things in a way so that the person helping you feels like they are a good person — not a cog in an evil machine. So rather than accuse the company that they work for of various evils (justified or not), present your case in a way so that your interlocutor will feel good for helping you.

It’s not over yet

Finally, these kind of nightmare situations have a habit of lasting longer than you think and surprising you a couple of times. Taking the airport example again, it wasn’t over when we got Air France to agree to let us board. It wasn’t over when we climbed on the plane. It was only over when we left Geneva airport, after going through the Swiss border.

When you’ve gone through the first step in an impossible task, it can be tempting to relax and assume that everything will go fine after this, and maybe let the stress of the situation show through. Don’t. As they say, it’s not over until the fat lady sings, and things can still go terribly wrong at any moment between now and your objective being achieved.

Until you’ve got what you want in your hand, stay focused and expect things to continue going wrong.


People react to being told “it’s not possible” in a variety of ways, and not all those ways are productive. To get around the brick walls which large corporations, bureaucracies and other social organisms put in our way, it is important to:

  • Calm down, smile and remain polite to maintain any chance of success
  • Become a human being rather than a faceless number
  • Be persistent to grind away the brick wall
  • Be prepared to lose to expand your freedom of thought and action
  • Be clear about your objective so you can be flexible about how to achieve it
  • Find who can, since often the first person you speak to cannot help
  • Take an active part in making things happen more efficiently
  • Make the other person feel good about helping you so that they are more likely to help you
  • Don’t relax this stance until it’s over, it’s easy to snatch defeat from the jaws of victory.

I hope those tips are helpful, both in a personal and a professional setting. If you have any further suggestions, please do post them as comments, and I’ll try to add them into the article as an update.

Thanks to Josh Catone for reviewing a draft of this article, and to Bob Leung for the excellent drawing.

1 It worth noting that we did send a letter of complaint to Air France about this and requesting a refund, but they have not responded. Fail.

2 Some of these practices may even help you in other situations, but they are particularly helpful when facing the impossible, and especially so when that impossible situation is wrapped in a crisis.

How to get a merchant account

Update: This article was written in 2009. Back then, the advice below made sense, since getting a merchant account was necessary to be able to take payments online.

Today, this is no longer the case. Skip the article below: just use GoCardless or Stripe.

Back to the article…

Getting a merchant account is a necessary step before you can take credit and debit card payments. That’s essential for the vast majority of Software-as-a-Service (SaaS) start-ups, and for many other kinds of business too. Unless you’ll be making all your money from advertising, you’ll need to be able to take payments, and if you want to be able to do so efficiently, you need a merchant account.

Considering how many businesses there are out there, one would think this process might be smooth and painless by now. Just apply for a merchant account, sign on the dotted line, and receive your merchant id in the mail a couple of weeks later. Unfortunately, this is not the case. Obtaining a merchant account can be a long and complex process, particularly if you’re a new, small business that’s going through this process for the first time.

I don’t claim to be an expert on this process. However, I’ve gone through it a few times, and I know people who are very familiar with it, so I put together this guide to obtaining a merchant account, from the cash-strapped start-up’s point of view.

Why do you need a merchant account?

If you’re building a typical SaaS business, you’ll need a merchant account in order to process credit and debit card payments. You could use Paypal, but some customers may feel you’re less credible because you don’t charge like everyone else. You could use one of the gateways that don’t require a merchant account, but typically they take fees of up to 10% of the transaction amount.

As of the publication of this post, the standard, most cost-effective way to take payments for SaaS products is still merchant accounts, managed by a regular bank, and going through a payment gateway.

1. When should you start this process?

Getting a merchant account can take anything from a few days to more than six months (if you make many mistakes). If you’re going through it for the first time, you may find yourself stuck at various points in the process. These sticking points can result in massive delays. You don’t want to be sitting there with a business that has users who want to pay you, but can’t because you don’t have a merchant account. Unforeseen delays in this process can cost you dearly, and if you’re going through this for the first time, there will probably be unforeseen delays.

On the other hand, there is little to no cost in obtaining a merchant account too early. Until your first transaction, the account just waits, unused, costing you nothing.

Therefore, you should start this process as soon as possible — today, if you can. Don’t underestimate how long this might take. Banks are often ridiculously slow in processing merchant account applications, when compared to the pace of the internet.

A small addendum to this advice: you need to have some sort of website (even if it’s just a placeholder) up before you start the application process. The bank will look at it. But this needs not be the same website that you actually launch with.

2. Apply to more than one bank

When you apply for a merchant account as a fresh new start-up, you’re in a position of overwhelming inferiority. The bank can dictate whatever terms they want to you, ask you for money up-front, impose ridiculous charges, etc. There’s a simple way around this: apply to several banks.

It takes more time and effort, but it gives you bargaining power. Banks don’t talk to each other much, so you can play them against each other. If all the banks are demanding an up-front payment, you can tell each one that another bank that you’ve applied to didn’t demand it. Banks do want your business, especially if you’ve conveyed the right impression, and so they’ll be willing to negotiate terms if it looks like you’re going to go somewhere else.

If you end up with more than one merchant account, so much the better. Sometimes banks can act irrationally and freeze your merchant account (particularly early on when you’re still relatively tiny), at the first hint that maybe there could be fraud going on (whether there really is or not). Having another merchant account set up gives you the ability to minimise the damage by switching over to your backup account. And, as I said earlier, unused merchant accounts don’t cost much.

3. Exaggerate your volumes

When you’re preparing your internal business plan, you should be conservative with your sales forecasts. However, when you’re talking to the bank, you should convince them that you will be a successful business that will bring them significant revenues. This shouldn’t be too hard — after all, you are optimistic about your start-up’s prospects, aren’t you?

Don’t lie. You need to be able to justify those numbers. However, if your internal, conservative goal is to get 500 sales a month by the end of the year, say you’re looking at internal projections of 1500 to 2000 sales a month. Be realistic, but optimistic. Ideally, you want to convey a turnover that will soon rise above a million dollars, so that your account manager feels that it’s worthwhile dealing with you and giving you a good deal. Don’t tell them that you’re hoping for those numbers. Tell them that those are the numbers you expect, based on current usage patterns. Now is the time for unwarranted confidence and optimism, not for cautious realism.

Your projected volume will also influence how big a cut the bank asks for. If your volumes are very small, they will want a comparatively large percentage of your transactions. If you’re planning to process a lot of transactions, on the other hand, they may be happy with a smaller cut.

Once your rates are negotiated, the bank is extremely unlikely to increase (or decrease) them if you fail to reach the volumes you predicted. So make sure you get good rates from the start by being optimistic.

4. Know your fraud-stuff

Banks still exist in a universe where “online businesses” are a scary new idea. They often don’t quite understand it, and they are scared to death of fraud. Even though you are honest, some of your customers may not be.

Your account manager will need to convince her fraud department that you’re not a great fraud risk. In order to do this, you need to convince her that you know all about online fraud, and that your business has covered all the bases to prevent fraudulent activity.

You should make sure you know all about chargebacks, 3D-secure, and AVS/CV2. You should also think ahead and figure out how someone might attempt to defraud you. This is harder because there is not all that much advice about how fraudulent individuals screw over businesses. Most of the readily available advice focuses on helping consumers avoid fraudulent businesses, but many businesses go broke dealing with fraudulent customers. It’s important to think ahead to what you’re going to do when (not if) someone uses a stolen card number on your service.

Do you pay commissions to affiliates? You’re automatically a higher fraud risk. Do you sell intangibles (like a web service)? People might call in chargebacks to get refunds. Sell to businesses? You’re a lower fraud risk, so make sure you highlight that. Adult industry? Multi-level marketing? Other “shady” businesses? Chances are the bank will not even touch you.

It is important for your own business that you think about fraud ahead of time. And it is extremely important for your merchant account process that you convey that you’ve thought about fraud and have it covered. Make the bank feel safe, or they’ll delay your application forever.

5. Be serious

The last thing any bank wants, understandably, is someone who’s going to blunder their way through figuring out how to charge credit cards, making costly mistakes along the way, getting defrauded left and right, and finally going broke and winding down the business because they couldn’t handle the complexity of it all. As part of your discussions and emails with your account manager, you must convince him that you’re serious, you know what you’re doing and, ideally, that you have help from someone who’s done it before, successfully.

If you don’t have access to such a helper, find one. Again, this shouldn’t be too hard — presumably you know other people who have started businesses before and they had to apply for a merchant account too. Once you’ve found that person, make sure your account manager is aware that you’re not helpless.

Be serious about security too. There are a lot of security regulations tied to internet transaction processing, and you want to be aware of them and convey that you are aware of them. For example, it’s probably a good thing to name-drop PCI-DSS in the conversation, and explain how you’re compliant with it.

Finally, do convey that you’re a serious business and you’re going to succeed. This is not the time to declare that you’re a fresh new business and you don’t have formal titles within the company, and don’t care much for rigid formality. This is the time to introduce yourself as the managing director and sound exactly like your stodgier competitors might.

6. Read the fine print and negotiate the charges

Banks will often present you with complex charges that are hard to compare with the offers from other banks. Unscrupulous ones may even use tricks like high termination fees that hit you when you least expect it. Others may ask for high up-front fees, suggest they hold your money for 3 months, etc.

All those points are negotiable.

The extent to which you can negotiate is largely driven by how much they value your business (see point 3) and by how many other offers you can choose from (see point 2). None of the merchant accounts I’ve been involved in have ended up having to pay an up-front fee, although the bank always asked for one.

Rates are an obvious one to compare, but don’t forget to also negotiate how long the bank will hold your money before passing it to you (many banks hold your payments in a separate account and pay your corporate account at the end of the month). Anything more than 1 month should be considered unacceptable. You should consider how important that is for your cash flow.

In conclusion…

Getting your first merchant account can be a real nightmare, but it doesn’t need to be so. Following these tips will maximise your chance of looking back at this process and thinking “Heh, it wasn’t so hard after all”. Doing the opposite of these tips can practically guarantee you’ll never get a merchant account. Here they are again:

  • Start the process early because it might take much longer than you think it should
  • Apply to several banks to give yourself leverage in negotiations
  • Exaggerate your volumes (realistically) to obtain good terms and good attention
  • Know all about fraud to make the bank feel safe about doing business with you
  • Be serious to ensure the bank feels you’re a trustworthy business
  • Read the fine print and negotiate the terms, as you may be able to get much better terms than initially offered

I hope these tips help others avoid some unnecessary headaches. Do you have any other advice? Please leave it below.

Thanks to appletizer and lifo for reviewing a draft of this article.

Starting up with a friend

It seems like a fool-proof plan: start up with a close friend. You’ll get along (obviously), and you’ll get to share the exciting, fantastic, scary experience of starting up with someone you care about. It’s not a bad idea, but there are a few caveats that you should be aware of before you proceed1.

When I started my first company with one of my closest friends, I expected things would go very well between us. We understood each other in ways that would take years to build up (and did take 10 years). We knew each other, and we knew we could rely on each other. We were prepared to have many surprises along the way — starting a business is always going to be a scary adventure.

What we weren’t prepared for was that the main problem would come from us and the dynamic between us.

What happened, in brief

I’m not going to go into all the details of what exactly went wrong, for a number of reasons (among them, it would be a one-sided account and inherently unfair on my friend and first cofounder). The long and short of it is, we had different expectations about the business. I left my safe, comfortable corporate job to work on it, so I needed it to succeed, or else I would find myself back in the corporate world. By contrast, my friend had already started several companies and was comfortably well off, so he didn’t have the same expectations and requirements.

It turned out we have a different definition of “the business isn’t working out”. For me, it was working out if it was making enough money to cover my expenses. For my friend, it wasn’t working out unless it was making enough money to also add to his existing wealth and thus justify the time and effort which he poured into it. Both those views were correct, but because we knew that we understood each other, we didn’t realise that our views were different until that difference had grown into a huge misunderstanding.

This core divergence of views could have been resolved easily if we’d known about it and discussed it ahead of time, but we didn’t know about it, so it festered and turned into dozens of other misunderstandings, so that by the time it finally became clear what our main divergence was, much of the damage was already done and it was entangled in a huge mass of emotional misunderstandings2.

This almost cost us our friendship. We got through this thanks to the help and mediation of another very good friend, who helped us to communicate to each other how we felt, so that we could move forward together rather than against each other.

I’m glad to say the mediation worked, and we’re still friends (perhaps even stronger than before). Nevertheless, I learned some important lessons from this.

1. Make your agreements explicit

The first lesson is to keep agreements explicit. It’s not enough to think that your friend understands what you think: make sure he does by discussing it openly with him. As my mediating friend phrased it, “unspoken promises” have a tendency to turn into broken promises (which are always hard to swallow). Avoid unspoken promises.

Here’s an example of a really bad thing to keep implicit: “We’ll only call it quits if the business is bankrupt and can’t raise any more money.” The promise here is that we’ll keep going until the very end. This may seem obvious to one party in the business, but it may not be so to the other. One partner could, for instance, feel that the time to call it quits is when the business has 3 months of cash flow left. Another may feel that it’s worth going deep into credit card debt territory before giving up.

Don’t make this mistake: keep those agreements explicit.

2. Detail your agreements

Once you make some agreements explicit, it should become clear that you need further discussion to figure out exactly what your explicit agreement is. Don’t be afraid to do this. It’s not “too early to discuss this”.

Here’s an explicit agreement that’s not detailed enough: “We want the business to make a lot of money”. Really? How much are you happy with? 10’000 pounds a month? A million? What is the definition of success? It’s almost certain that you and your business partner have different views as to what “a lot of money” is. Being on the same page about what you expect out of your business will ensure that you don’t pull in different directions when things are going well. Think of how mortifying it would be to find out that your partner wants to pull the plug when you think that the business is successful.

3. Don’t be afraid of discussing the bad stuff

This really happened!

There are a number of subjects which seem almost embarrassing to discuss when things are going well. For example, “What if one of us decides to pull out?” Your first reaction to this topic might be “What? We’re barely getting started, and already we’re talking about what happens if one of us pulls out?”

The reality is that people’s life circumstances change through time. They get married, or decide to leave the country, or get engrossed in a different pursuit, etc. Many things can get in between a founder and his start-up. Similarly, many things can go very wrong with a start-up. When those things do go wrong, or when one of the founders decides to pull out, is not the time to discuss these things. You need to discuss them with a clear head when no one is thinking of pulling out and the business looks healthy and hopeful.

When you discuss your start-up’s future, do not be afraid to talk about the disaster scenarios. Also, when you negotiate what will happen if a partner quits, don’t be so sure that it won’t be you.

4. Write things down

There are two reasons to write things down: first, people’s memories of conversations are faulty. Writing things down also ensures that there is no disagreement, later, about what was decided. You don’t need a long document for this — even just one or two pages describing your agreement is enough to avoid later misunderstandings.

The second reason is that people may think they have reached an agreement when in reality they never agreed about the details. Once you put something in writing, you give it a certain air of finality that teases out those last remaining disagreement. Basically, putting an agreement in writing is like putting a new piece of functionality in code. Until it exists in that form, it’s just vapour.

Halfway through my misunderstanding with my friend, we thought we’d figured out a way forward. I wasn’t sure that we were both thinking the same thing, so I made the effort to put it in writing, in the form of a business plan. When my friend read it, and understood more clearly what I meant, he recanted, and the agreement fell through. It’s a good thing that it fell through, because it would likely have resulted in even more problems later on if we’d gone through with it based on our flawed understanding of each other.

5. Don’t make it work at all costs

Yes, I know this is your friend that you’re starting up with, and this is your great opportunity to start your own business. However, if, in those discussions, you find that there’s an intractable disagreement, don’t fall into the trap of thinking that the most important thing is to smooth things over and start the business.

Starting up with someone is almost like marrying them (temporarily), in a way. You’ll be talking to them almost everyday, and possibly even more than with your significant other. You’ll be working on a “baby” (your business) for many months. It’s a big commitment, basically, and much like any other kind of significant commitment, you shouldn’t go into it if you think there are major problems, because those problems will only get worse.

6. Don’t assume things will get better with time

It’s easy to rationalise away big problems by assuming that things will get better with time. In some cases, they will, but in a majority of cases, they won’t. What this means, for example, is that you shouldn’t assume that your inexplicably small share of the business will magically grow to 50% later on. This is even less likely to happen if the business is working well (if the business isn’t working out, chances are it doesn’t matter anyway).

Sample questions

This article wouldn’t be complete without a list of questions that you might go through and discuss with your cofounder. Use them as a guideline or as a checklist, as you please.

  • What do we both mean by “the business is successful”?
  • What do we both mean by “the business is not successful”?
  • What happens if one of us needs to voluntarily pull out, for any reason?
  • What happens if one of us cannot work on the business anymore, for involuntary reasons?
  • What are the conditions under which we’d call the business a failure and pull the plug?
  • What is plan B for each of us if we do pull the plug? Are we both prepared for that plan B?
  • What do we expect of each other, both in terms of responsibilities and in terms of attitude and effort?
  • What is and is not an expense? What is the maximum amount someone can spend on an expense without checking with the other? (from Sebastian Marshall)
  • When and how will profits be distributed? How much will be reinvested? What will the reserves be? (from Sebastian Marshall)
  • What happens if one partner needs cash and the other wants to reinvest it into growth/expansion? (from Sebastian Marshall)
  • How will you handle it when (not if) the hours each partner is working are unbalanced? (from Sebastian Marshall)

This is not a final list by any means, but it should at least provide some starting points to make the implicit explicit. If you have other suggestions, please do add them in the comments below.


I don’t regret starting that business with my friend, but I do regret not clarifying those kinds of questions upfront. It would have saved me a lot of worry. If your business is struggling, you don’t need the additional pain of seeing your friendship unraveling under the stress of accumulated misunderstandings.

So, do yourself a favour, and set out to:

  • Make your agreements explicit so that you don’t break implicit promises
  • Detail your agreements so that your promises are clear
  • Don’t be afraid of discussing negative scenarios, so that you don’t add the stress of misunderstanding to already bad situations
  • Write things down so you’ll remember
  • Don’t make things work at all costs, so that you don’t spend the next years living with a deal that’s not acceptable to you
  • Don’t assume things will get better with time, so you’re not surprised when they don’t

I hope this helps others. Your comments below are much welcome.

1 It’s worth adding that this advice can be useful for any kind of adventure, not just business. However, given the propensity of businesses to crank up the pressure to diamond-producing levels, and what can often be at stake, it’s particularly important in this context.

2 Although this sounds like a barely mitigated disaster, I must add that, on the whole, the business was a success (it made money, I learned a lot from it both about myself and about start-ups, and it provided a jumping board from which to start my next business). It wasn’t as much of a success as it might have been, and there were some times when it looked like it might turn into a small disaster, but on the whole it turned out reasonably well.

College vs Startup

Let’s say you’re 17, a high school student, passionate about web start-ups, and you just know that you’ll be starting your own one day. Should you go to university first? Should you get a degree? What use will it be? Is it worth it? Should you study computer science? IT management? An MBA? Physics? Maths? None of the above? Which will best prepare you for start-up life? More importantly, is that why you should go to university, or are there better reasons?

It’s a tough question, particularly since it’s going to have a big influence on the rest of your life. If you skip university, your life will be radically different than if you don’t. If you’re keeping track of the pulse of conversations online, you might have spotted a few discussions on the topic recently, such as The Great College Hoax (HN link), published on Forbes, which makes the bold claim that colleges are outright scams, or a milder one by celebrity VC Fred Wilson, which claims that you don’t need a degree to be an entrepreneur (HN link).

My opinion on the subject is simple: if you have a thirst for learning, and you don’t have to enter the workforce immediately (i.e. you can afford, somehow, a degree, without being financially irresponsible), then you absolutely should go to university, even if you have a start-up that you could work on right away. This is not because you need the degree for your future career, but because it’s a great thing to spend your next 4 years on.

Here’s a breakdown of reasons why you should go to university rather than work on your start-up 1:

Business ideas are a dime a dozen

Perhaps you feel that you have the idea of the decade, that will make you millions. If so, I encourage you to read this essay by Paul Graham, the friendly uncle of the start-up world. Don’t worry about ideas. You’ll have just as many, if not more, ideas when you come out of college as when you went in. Having a great idea is not a good reason to skip university.

You can drop out

If you go to university and you decide that it’s really not for you 2, or if you work on your venture on the side and it takes off, you can always drop out like many other famous entrepreneurs did before. No one’s forcing you to stay. But at least, you’ll drop out with the knowledge of what you’re giving up, and, hopefully, you’ll do it for something concrete that needs all your attention right now.

A good time

University is an enjoyable way to spend four years. Start-ups aren’t the only thing in life. If you focus yourself on work alone, without enjoying other pleasures along the way, you may well find out later that you missed out on the more fun parts of life.

You don’t need a degree to start a start-up, that’s true. There are many other things that you don’t need for your start-up, but which are still enjoyable and worthwhile parts of life. Having a girlfriend/boyfriend, and eventually a wife/husband and children, is not essential to your business. Neither is a university degree. But both can be great things to have in your life.

Learn things you would never have learned by yourself

Many people think they need to do a degree in the profession that they will follow later (and, for some, such as architects, doctors and lawyers, that is true). For web entrepreneurs, an IT-related degree can seem like the obvious option. I believe that it’s far more useful and interesting to study something else, something that you’re interested in and that you know that you’d never get around to studying in depth otherwise.

In my case, I’ve always had a thirst for figuring out how things work, so I quenched it by doing a degree in Physics. Many things I learned there are not directly useful to my life. Indirectly, however, my four years of Physics have shaped the way I think, the way I approach problems, and my approach to learning complicated topics. I would not have absorbed all these things if I hadn’t done a Physics degree. There are some things you really have to be taught by someone, to learn them properly.

I was genuinely interested in Physics, and that’s why I chose that subject. This leads me to an important point: you should choose a subject based on your interests, rather than career usefulness. There is an unhealthy fashion these days to link degrees to jobs, as if a degree was the best way to train for a specific job (e.g. degrees in “Communications and Media” as training for Marketing or PR people). This isn’t helping anyone, least of all the universities. You should do a degree because you care about the subject, not for the career prospects. Ultimately, university is supposed to train you for life, not for a job.

Doing a degree in a subject that does not end up being your career gives you an extra perspective to look at things, an extra string on your bow.

The last place where people care to teach you

Once you enter the workforce (for your own business or otherwise), no one cares what you learn. Corporations will make a token effort to give you some opportunity to better yourself via a training budget and various work-related courses, but university is truly the last place where you’ll have teachers who really care about teaching you things (if they’re any good at all).

Moreover, they’ll care about it for a long time. For years on end, a relatively small group of people will care about what you learn, will care that you become a better person through their teaching, and will care about your intellectual development. Believe me when I say that that will never happen again in your life (unless you go back to university, of course). The most attention you’ll ever get from a corporate course is a couple of weeks of somewhat bored oversight.

A shelter where you can develop yourself

The “real world” is a harsh place, where, as I have said, no one will care whether you’re growing as a person. Once you’re working full time, on your start-up or otherwise, it becomes incredibly difficult to find the time to invest in extra-curricular things which you really want to do. At university, it’s easy — you might decide to learn a new language, or even start a start-up on the side, and still have time for your social life and to take your studies seriously. Once you leave that shelter, however, you’ll have to surmount tall obstacles to find the time to do what you want.

Meet interesting people

Last but not least, to me, the most fascinating aspect of university was the people I met there. I’ve yet to find myself again surrounded by so many smart and interesting people, from college professors to students. Moreover, you get to meet all those people in a low-stress, low-competition environment 3 where you can really become friends with them. Making friends in the work-place is much more delicate and takes a lot longer 4.

Even if you’re looking at it from a start-up point of view, you will meet many more potential start-up cofounders at university than by going straight to work — whether for your own business or for someone else’s.

But the important thing there is really that you’ll make friends that you otherwise wouldn’t have made, at a stage of their life where they’re growing and learning many new things, and build personal connections that are priceless not from a business point of view, but from a personal point of view. Life is about more than business.


There are many good reasons to go to university. You’ll notice that I didn’t list job training among them, because, unless your vocation requires a degree, that is not the best reason to go to university. If you can go to university without being financially irresponsible, then it is personally irresponsible not to — especially if your reason for avoiding college is that you have a start-up idea to work on.

I hope this helps. There are no doubt many other reasons to go to university. Let me know your thoughts below.

1 These are largely based on my personal experience of university, in Oxford University in England. Different universities can vary greatly, though in my opinion you can get a fairly good idea of what life will be at a specific college before you apply, through open days and other ways.

2 Please give it a couple of years before deciding that… the first year is usually significantly less interesting than the rest of a degree course, because the university first has to ensure that everyone has the baseline level of competence in the subject.

3 It’s worth noting that not all universities, and not all parts of the world, are equal. Asian universities, as my friends who studied there tell me, have a much stronger emphasis on study and don’t care so much about developing their students socially. That’s a mistake, in my opinion.

4 It can happen, and does happen, but in many cases, a work relationship will get in the way of a personal relationship and, at the very least, slow things down so that what might have taken a couple of months of hanging around each other will take a year of careful exploration. And then there’s always the potential for becoming your friend’s boss, which throws spokes in the wheels.

Thanks to Kelvin Koh for reviewing a draft of this article.

Counting hours doesn’t make sense

There’s recently been some interesting discussion on the internet about “How many hours” people are putting in each day, inspired by a Slashdot poll.

This can be a sensitive topic. After all, everyone wants to do a good job, and the number of physical hours that you spend “working” is considered an objective measure of that. In some cases, that measure is considered positive (Fire people who are not workaholics, claimed Jason Calacanis before he changed his mind and toned it down to “people who don’t love their work”) or negative (Fire the workaholics, responded the ever controversial 37-signals blog).

When working in a self-employed, services job (e.g. freelancing), the idea that hours matter is deeply ingrained, because hours are the measurable thing that we charge for (even though what the client really cares about is getting the job done). In regular jobs”, hours are also important, because they are the basis of the long-term contract between employer and employee (“Your working hours are from 9am to 6pm on weekdays, excluding public holidays”), and they are the first mechanism an employer will use to make sure you’re working hard enough.

This is all very interesting from the point of view of a technology start-up 1, because when you run your own start-up, and you have no one to impress with your long working hours, you end up realising that hours are immaterial. What matters is the quality of what you do, not how or when or for how long you do it. And that quality is not correlated with how long you are sitting down and focusing on your immediate “work tasks”.

When we measure results instead of hours, something interesting happens: the distinction between work and not-work blurs away and vanishes, for two reasons. First, clever ideas can make a huge difference to results, and ideas occur anywhere, at any time. In fact, they’re least likely to occur while sitting at a desk working. Secondly, it soon becomes obvious that our actual output of things done is correlated far more to how we feel on the day than to how many hours we spend “working”. The real measure of work is not hours – it’s energy.

We all have a certain amount of energy each day, that can fluctuate depending the day, on our general level of fitness, nutrition, health, state of mind, etc. Some activities (such as going to the gym) increase our daily pool of energy. Others (such as staying up all night or getting drunk every evening) decrease our daily pool of energy.

Asking the wrong question can have very detrimental effects. It forces our mind to focus on the wrong approach. I don’t get more done by sitting in front of my computer looking at work for an extra hour. I get far more done by investing time to exercise regularly so that I feel energetic.

So the useful question is not “how many hours do you work?” but “how much energy do you put into your work?” Other useful questions that come with it are:

  • How much of your daily energy do you spend increasing your total energy? Do you feel you spend enough? Do you feel you spend it on the right things?
  • How much of your daily energy do you waste each day? How do you define waste? Is all that waste really unproductive or does it have some beneficial side-effects? Are those side-effects sufficient to justify spending that energy?
  • Do you spend energy on things which actively hurt you?
  • Has your daily energy increased or decreased in the last 6 months? year? 5 years?

Any of these questions is more worthwhile than “How many hours do you work each day?”

1 This varies depending on the kind of business you run. If your business is to produce birthday cards, chances are there’s a direct relationship between your hours and your success. The relationship breaks down for start-ups where the output is largely intellectual and leveraged (like a web-based application).


Newer posts »

© 2014 danieltenner.com

Theme by Anders NorenUp ↑