DjangoCon Europe 2017 - Autopsy of a slow train wreck: The life and death of a Django startup

Speaker: Russell Keith-Magee (Django core team member, Founder of BeeWare project)

Startup culture

Russ was CTO of TradesCloud until earlier this year, when he and his partner shut it down.

As an industry, we are very proud of super successful startups. This is not the normal startup experience. VC investments have ~5% success, 16% might reach break even, the remainder fails (often hard and in flames).

People whose startup fails feel often alone due to the silence and stigma attached to startup failure.

TradesCloud: The 13 mistakes.

TradesCloud survived for 6 years.

Find tradespeople who advertise in the region, rate them, recommend them, stuff like that. They found people who were interested in the problem from the other side.

Mistake 1: Validate, then build.

No battle plan survives first contact with the enemy - same with software. Consider (strongly. Repeatedly.) talking to your potential users and customers first, to find out if you're on to a hobby project or an actual business. Don't conflate the two.

Mistake 2: If you can't sell it, it's not a business.

You need customers. If nobody is interested, if you haven't got a sales channel … you might aswell shut up shop now. Consider to change directions to move to a profitable business.

Mistake 3: Humans gonna human.

  1. Humans don't behave rationally, this goes especially for economic decisions.
  2. Software is hard to sell to humans. Software is, by tradition, over-promised and under-delivered.
  3. Software is hard to sell to humans. (This is also why FLOSS funding is hard.)
  4. Humans are people. Firing people is not attractive.

Mistake 4: Beware favourable patterns.

Finding good customers at first does not mean that your streak is going to continue.

Mistake 5: Do the math.

Calculate best-case, worst-case, and expected incomes. Do this repeatedly.

Mistake 6: Pricing is what the customer will pay, not your burn rate.

Don't go off "how am I going to pay the landlord".

Mistake 7: Establish your sales channel

Word-of-mouth is a great sales channel - but relying on it fails to teach you how to establish trust and get into the door without personal introduction. The dangerous zone is: expensive enough to require sales people, too cheap to pay them.

In order to close sales, people will have to believe you, or the cost of trial needs to be near zero.

Mistake 8: Sales don't stop when you sell.

Sales stop only when the product is actually integrated at the buyer's business, to prevent churn. It is most important to get the people who will have to use the product on board. You need to sell to both the owner and the user. Show that jobs get easier but will not get them dismissed.

Mistake 9: Establish failure criteria.

Establish when to cut your losses and then do it.

Mistake 10: You are who you are. Don't deny it.

Don't forget what your core business and your core customers are. If you want to change your tracks, research if you can have success, and what you need to change to fit in.

Mistake 11: Take the hint.

Dead horse. You know it - don't stumble along hoping for the final blow. It just costs money and emotional resources.

Mistake 12: Quitting is always an option

Don't forget about this: you are allowed to quit. Don't conflate success of the company with your own success - it's still just a job.

Mistake 13: Partnerships require actual partners.

Partnerships need to be equal. In partnerships, you need to discuss all your issues as equals. If you don't see eye to eye, the partnership will not work out.

Aftermath and Future

This was a large learning experience. It showed Russ new ways for the future (in the form of BeeWare).