BLOG: 11 Lessons I learned from my failed tech startup
By Daniel Williams. Posted 10 March, 2016
The day I began working on my previous startup, I was under the usual self-delusion all us entrepreneurs fall under — that we are working on the web’s next big thing and we’re going to change the world.

As it turns out, we weren't, and we didn't change anything.

Our elevator pitch was "Microsoft Access, for the web, for dummies," and in an early blog post I described a day in the life of an early-stage tech startup which detailed the daily chaos of my startup life.
A screenshot of Collecto at launch
1. Failure is likely
We developed our product over 9 months; then two weeks after launching our beta we pulled the plug. Most users were abandoning the app quickly and the few users we had were unresponsive when we tried to engage with them.

Starting a new business is tough and most don't make it past one year. For those that do, failure is still likely within 5 years. Pretty depressing, but the fact is, success is not guaranteed. You're likely to fail, but at least you'll learn a lot if you have the courage to try again. And you know what they say about falling off your horse...
2. Avoid unnecessary expenses
We took a two-year lease on an office, set up a complex server network, and hired a developer, who was very expensive. We justified all this because we were convinced we had a billion-dollar idea. Compared to Silicon Valley startups our budget was tiny, but at $10,000 a month in expenses (which came out of our own pockets) we quickly realized the numbers we needed to make this work were years away. In round numbers, finding 1,000 customers to pay $120 a year was a massive goal. And that was just to pay expenses. We also needed to pay the salaries of two founders. After all, we have to eat too.

Work from home instead of renting an office (or get a hotdesk at a startup lab); buy a shared hosting account instead of an AWS server; and outsource instead of hiring. Until you have revenue, question every expense. Even if you have funding, behave like you don't. Every dollar you spend had to be earned by someone (most likely you), and that someone wants a return on it (see point 1.)
3. Look for problems, not ideas
In our first week we sat down and tried to define the problem our product solved. Alarm bells should have been going off when we found this difficult. This is the big startup mistake—we were trying to create a problem to fit our idea.

There are an infinite number of ideas out there, and most of them are terrible. If you can identify a common problem that causes a real pain for users, and create a solution, or simply do it cheaper or better than others, you'll have a better chance at success. Ideas on their own are usually worthless.
4. Get to market quickly
Our vision for the product became so grand it was overwhelming. We started off with a single function the product should perform—the core essence of what the product was—allow people to save text to a database and order the display of it in different ways.

We then decided that images were just as important as text, and we spent a few weeks getting this to work. Then we decided text shouldn't be the only data type, so we added support for numbers. Of course, with numbers you should be able to produce sums and averages, so we spent a few weeks adding support for this. And on and on this went for months as we decided we needed just one more feature before launch.

Ironically, when we were using the product ourselves, we were only using the original function: saving text to the database and ordering it in different ways.

Make a list of everything your product will do, then order them by the most to the least important. Then make your product do the first one and launch it. Your users (if you're lucky enough to acquire any) will tell you what features to add. If you're not embarrassed by the first version of your product, you launched too late.
5. Nobody cares (sorry!)
We handed over a not-insignificant amount of money to our lawyer to draw up a non-disclosure agreement, in the fear that someone might steal our idea. The first investor we approached said he wouldn't sign an NDA and that we shouldn't expect anyone else too either. Investors get proposals all the time and signing NDAs is too risky.

Skip the NDAs (non-disclosure agreements) and come out of stealth mode—people just aren't that interested in what you're doing—which is a tough pill to swallow when you think you're working on the next big thing; but unless you're Apple, no one is impatiently awaiting your launch.
6. Don't do anything that doesn't help validate the product
Spending weeks setting up the infrastructure to support the product was a waste of time, because no one ended up using the product. We could have achieved the same result with a $5-a-month hosting account.

Everything you do should be working towards validating your product. Before you spend two weeks adding that next feature, ask yourself "will this help us determine if people will use our product?" It's easy to get caught up in "it just needs this feature before launch" talk, which can spiral on forever.

Sometimes, validation can come from finding a buyer for a product you haven't even made yet.
7. Make sure you'd use your own product
We built the product for ourselves, but we didn't really end up using it. Managing data is a pain, not a pleasure, and there were simpler ways to do it. Our whole pitch was that people misuse spreadsheets to manage data, and our product could fix that, but I've since returned to spreadsheets since we switched off our servers, and it's working just fine.

Ideally your product solves a problem you've had yourself, and is a problem others are likely to have. If you find yourself not using your own product, chances are, no one else will.
8. Find a paying customer
Our product was free to use and we figured we'd work out the commercialization later. It should be easy being a SaaS product (software as a service.) It would have made more sense to find some people for whom our product would solve a problem they had, and get them to commit real dollars to our service—before we built it.

Businesses exist to produce revenue and generate profit. The sooner you find a paying customer, the better. If you're lucky enough to gain traction with a free product, you can always fall back on advertising to at least pay the bills, but that's hardly an inspiring business model, and relying on something like Google Adsense as your sole  source of income is risky.

Build your business one customer at a time. Overnight successes are exceptionally rare.
9. 40-hour weeks won't cut it
If you want to work 40 hours a week, get a regular job. Startup life requires devoting most of your waking hours. Starting a new business requires commitment and sacrifice. Expect to put in at least 60 hours a week for the first few years.
10. Know when it's time to quit
It's easy to say "we've come this far" or "we've invested so much" when evaluating your startup's future, and then pour in more in the hope that things will improve. If the writing's on the wall, just be totally honest with yourself about it, and move on. You're better off investing your time and money into your next effort.
11. Make sure your heart's in it
I was convinced we had a billion-dollar idea, and I was spending the money in my head before I'd made it. If you're starting a tech business for fame and riches, STOP RIGHT NOW. Your users won't be inspired by a vision that includes you being a billionaire driving a Ferrari.

Thank you! Your submission has been received!

Oops! Something went wrong while submitting the form

Recent posts: