Monday, May 12, 2014

Will your startup idea be successful? I don’t know. But here’s a checklist before you approach people who could help you.


(this post contains two startup ideas that you might consider doing, if so, please contact me!) 

Judging startup ideas. This is notoriously difficult but simultaneously enormously relevant. It is hard to predict if an idea (or your execution of it) will be successful. However, it may be possible at an early stage to determine that it likely won't be.

This post is intended to be useful to rid your startup idea of easily avoidable errors before you approach (angel) investors, incubators or other people that could help you.

PowerDNS
My personal history in judging startup ideas is shoddy - when PowerDNS was launched late 20th century, I did not realise our idea was doomed by design.

We wanted to sell expensive DNS software to a world used to getting it for free with the operating system. The people we tried to sell it to (system administrators) understood our sales pitch all too well: Database driven DNS would save a lot of manpower. Back in 2000 that meant the people we aimed our marketing at would personally lose their jobs if they took us up on our offer. And we didn’t know anyone higher in the food chain. We didn’t get anywhere.

However, over the years, my batting average probably improved. PowerDNS is doing very well, both as open source technology and as a company. My joint-venture with Fox-IT had sound beginnings (because I learned from PowerDNS), and achieved the three-year goals we set for it in only 5.

Recently, I was asked by the good people of Yes!Delft if I could participate in a startup event, but I decided I had better things to do, since I judged the attending startups to be hopeless. And I was rightfully chided for that - it is incredibly hard to judge a startup idea

New ideas sound bad
Almost by definition, great new ideas sound really bad when you first hear of them. Because if the idea were so obviously good, it would’ve been done already! So prima facie plausibility is not a good way to judge an idea. For example, when launched, Twitter looked like a truly stupid blogging service (2 sentence blog posts?!). When the first digital cameras arrived, the only available storage devices were fragile power hungry hard disks. It made no sense.

But still we need ways to distinguish the right kind of bad idea from the truly bad idea. Are there things that just have to be right, even if we believe in the concept itself?

Here’s a checklist of things to consider:

  1. Is the idea (near-future) implementable?
    • But not so doable that anyone can do it
    • If it is possible, can you explain why it hasn’t been done yet?

You'd be surprised how often startups ideas flunk these simple tests. Wishing something is possible doesn't make it so. The laws of physics (generally) do not change for you. It is important to separate the two kinds of “impossibilities”:
  • breaking the speed of sound (hard work, requires lots of R&D),
  • breaking the speed of light (requires Nobel prizes, whole new laws of physics).
Usually you should prefer 'speed of sound'-hard problems.

  1. Do you have a good mental model of who would buy and/or use your technology
    • How do they buy things?
    • Would they buy it from you?
    • Is it conceivable that they’d pay your prices?
    • How much better do you have to be than what they are currently buying?
    • If they won’t be paying, do you have a plan to make money anyhow?

Note that this item does not demand a 5-year business plan. No such thing makes sense, least of all for a startup. The item does ask if you have a working mental model of your potential customer base. What makes them tick? Does your team have personal experience? Do you know people who could help you there? Or are you operating on a “of course they'll buy our stuff, it is cheaper”-grade model? 

  1. Do you have (people with) special knowledge, capabilities, access that competitors don’t?
    • Technologies
    • Skills
    • Execution
    • Unique team
    • Relations
    • Intellectual property rights

Companies, startups, are in perpetual competition. If you don't have some kind of head start, don't enter the game. Note that you don't have to have all of these, but never play fair. Make sure you bring something special to the game.

  1. Are you serving a market that isn’t too crowded already? 
    • Or one that nobody previously wanted to serve, or thought of serving?

This one is key too. Unless you have vast resources, don't try to make a better iPhone from scratch. It is tempting to target a well known problem, but the problem with well known problems is that the are, well, well known. Again, don't play fair in this respect. Tackle something in a field that is not too crowded. Don't try to out-Google Google! Do however try to do things other people found beneath their scale, or too boring. At least don't do what everybody is doing already, unless you can do it 100 times better.

  1. Do you have access to your market, or can someone easily block you from entering it?

When pondering revolution, don't expect help from the establishment. They've been at it for decades. Intel kept AMD processors away from everyone by promising sweet prices only in case of total vendor loyalty. Microsoft set up its licensing so they got paid per computer shipped, not per copy of Windows – thus removing any incentive for shipping anything other than Windows. Don't expect to compete with Amazon if you need AWS to do it.

Does it all have to be perfect?
These things can compensate each other, and it is a rare startup idea that will score well on all items of the checklist.

For example, great execution can make up for an idea that is not too novel in a sparingly addressed market. Conversely, a truly great technology can take off even in the face of adverse markets and lack of relevant business experience (the launch of Google comes to mind).

A cautionary note
No battle plan survives contact with the enemy. And it is a rare startup that delivers what they originally thought they’d be doing. WhatsApp did not start as a texting service. American Express was a courier service which later added credit facilities, and only then dropped out of the delivery business. Google did not start with advertising. “The Facebook” launched as a college hangout.

But if you don't have an initial narrative on how you could succeed, you'll have a hard time gathering funding, talent and advice!

Some worked examples.
A cell phone battery that never dies. If we score this idea with the list above:
  1. yes, it might be possible (pacemakers used to have nuclear batteries, and many decades old spacecraft are still powered by them. Mars Curiosity is a recent example)
  2. Everybody who is sick of short battery life would want to buy it. So everyone. Possibly cell phone companies too, if they weren’t tied up with their own technology yet
  3. But no, I don’t actually have the special technology, nor do I know anyone who does. The nuclear batteries are not compatible with today's world.
  4. Everybody is already trying to serve this market.
  5. Batteries are highly patented, so I could easily be blocked even if I had an idea.
So this idea likely won’t get anywhere. Sorry.

Let’s try another one. Cheap hearing aids. The most basic hearing aids sell for 500.00 euros a piece at least in large parts of Europe, since that is what insurance companies reimburse. However, they only reimburse once every three to five years. If in that time, you lose one, or break it, you are on your own. Basic hearing aids are very basic, there is no need for them to cost 500 euros. And the fabled ‘programming’ consists of three knobs. Changing the clock on your microwave is harder. So, does this idea fly?

  1. Yes, the idea is possible. A quick web search shows that hearing aids, somewhat different from the ones we know, sell for less than 10 dollars a piece in other parts of the world.
  2. I personally don’t know a lot about retail, but I know enough about hearing & speech, compliance, the medical world, that I’m sure I can get these hearing aids acceptable to European markets and regulators.
  3. I know quite some (younger) hearing aid users, and they are all unhappy and want change.
  4. No one is doing this since it is easy to make stupendous amounts of money with regular hearing aids.
  5. Siemens, Phonak and others have lots of patents on advanced hearing aids, but basic ones have been around for ages and should be public domain. Traditional hearing aid shops won’t touch it (they are part of the racket), so we’ll have to work around them. Don't expect this to be easy.

Summing it up, I guess it is worth it to ponder this idea. Doesn’t mean it would work, but it would get out of the starting gate. Please contact me if you want to do it, I have some more ideas on this field. There's a (little) bit more to it.

Early stage startup patterns
Here are some common early startup patterns, some of which work:

  • Having a powerful idea, but you need to find people and suppliers for everything, since your own team (you?) lacks the native ability to implement the core of the idea.

    This means you might fail item 1, because you don’t know. You fail on item 2. And all the suppliers you need to approach might be beholden to your competition, so you fail 4 too. This never works. If you need to pay for all the work (instead of doing it as a company and owning the result), you are doomed. Gather more talent first, and give them equity (stock).
  • Launching a ‘platform’ for something that would be great once it was there and everybody was on it.

    I recently received a pitch: “.. will be an online multimedia platform that brings together four customer groups: biologists, press officers of research organisations, journalists and the public. We will identify the supply and demand related to nature information and will organise and facilitate the efficient, easy and quick search and exchange of content and contacts via web, apps and social media”.

    Once there, it would be great. But the greatness of the idea does not make it happen - the challenge is how to get there. Half a platform is no use. The probable reason it hasn’t been done is that it is very hard to get it off the ground. So by the metrics above, this fails 4 and 6. It can be done, it is not too easy, but lots of people would want to do it already, and everybody who has achieved partial success will try very hard to block you.

  • Great execution versus sleepy industry. I keep trying to get people to do this. Traffic lights. They are stupendously boring, and nobody’s main industry. An afterthought for an intersection. At least in Europe, most traffic light installations come with a giant cabinet that is placed somewhere next to the road. Refrigerator sized. What the f*ck do traffic lights require that even in 2014 they need controllers that big?

    So even without innovating a lot, you could enter this market simply by being nimble and efficient. You could power 250 intersections from the computing abilities of a cell phone. And once you have connectivity, add lovely control centers for the city that manages the intersections.

    Gather statistics, hand out magic devices to ambulances that pre-announce their trajectory to the hospital miles in advance, guaranteeing green lights all round!

    There’s no fundamental innovation going on, but it requires great execution to get this sleepy industry moving! And by the time the competition wakes up, the only thing they can do is buy you. Again, if you want to do this, contact me, I have lots more ideas on traffic lights. I'm like that.

  • Relying on partners as a shortcut to scaling. Thing is, actually scaling an idea is hard work. With notable exceptions, you’ll have to get people to sell, support and market the stuff around the world. And even if you have one of the rare ideas that sell themselves, you still need to build the infrastructure to support your customer base. It is a common theme for early stage startups to focus on their idea, and claim to find partners for the rest to speed up their time to market. I’ve fallen for this one too.

    So here’s the problem. Partners that could truly help you, since they are in the right place and have the capabilities, are likely to be invested in your competition. Even if not literally, they will ‘think like the established players’, and not like you. Secondly, to them, you are really small fry. How much of their attention will they focus on what to them might end up as 1% of their revenues? Will you get their best work?

    So any idea that starts with “we’ll focus on the core and speed up time to market with partners” needs more work, like for example, which partners would be interested and why. Count on having to do a lot of the heavy lifting yourself. (Note that it is fine for many places to buy commodity services like hosting, but this is different from ‘partnering’).

  • Ideas that deliver ‘millions’ in savings to ‘billions’ sized companies. So the thinking goes like this. You have a technology that objectively saves money, for example by allowing for slightly more efficient procurement. Say, 5% savings. For large enterprises, that adds up to millions and millions. If you’d be able to appropriate 20% of those savings, you’d be making good money. Business case done! I encounter a lot of this thinking among technology students, and I admit I used to fall for it too. (This is an example of having a bad mental model of your customer base, by the way).

    There are countless opportunities for increasing efficiency and saving 5% here and there - if only people would sign up. And there’s the rub. Your 5% savings are worth their while in absolute terms. However, unless you make *everything* 5% more efficient over at your customer, their bottom line (‘profit’) impact will be a lot smaller. However, the disruption you cause by getting them to move to your new technology is real. This means that a company has to decide to move to a startup’s technology to eventually increase profits by 1% - whereas profits go up and down by 30% all the time for other reasons. Seen this way, it is not a compelling case.

    To overcome this, your startup will need solid credentials and a great network. As a great example, see “Looking bigger”.

Wrapping up
Even good ideas tend to look bad initially, but as a founder, you’ll love your idea anyhow. But the idea is not yet a company, and you should try to work out the obvious kinks before looking for help. A potential investor or employee that can shoot instant & serious holes in your idea will not join you. If you show up with a compelling narrative that at least in theory could work, you might get the help you need!

Further reading