3 reasons software startups fail — backed by data
Finding the perfect data requires the perfect study of startup failure.
Building the perfect study is hard:
We’d need a LOT of exit interviews from founders, members, and stakeholders: enough for significant sample size and we’d need to account for every bias.
Considering how much money is poured into startups every year, I’d think someone out there would find it worth the time.
One study we reviewed from Penn State University, that tried to leverage other studies to help reach a data threshold, concluded:
“…there is a lack of relevant primary studies on software development in the startup context.”
We can, however, gain a lot of value from the studies that DID try to quantify what causes startups to fail — despite the flaws of each study.
I reviewed several studies but only leveraged a few that did a decent job at containing their biases.
From the founders’ perspective, this often meant they ran out of cash.
Through analysis, the studies found the money they did have was not resourced properly.
Traditional software development bled into their processes — many of the companies from the study built a full “MVP” before even trying to get their first customer.
In growth environments, building software can create massive waste. We need to find clever ways to gather feedback fast.
Often this means interacting with people and not writing any code until we’re sure it’s worth it.
Growth problems require identifying our riskiest assumptions and using a scientific process.
The failed startups evaluated did not validate the riskiest assumptions before trying to scale.
In one study, failed companies did so because they tried to build features to get users to pay — ignoring all data and feedback that suggested their users weren’t even engaging with their product at all.
They tried to force their own needs rather than considering the current problems the customer was facing.
If you’re building any new app, version, or feature and customers aren’t involved in the process frequently, we risk massive waste.
Risky assumptions happen at every level of the process:
- the idea itself
- will users even use the new app?
- even if they use it, will they find the experience friendly enough to continue using it?
- will it be good enough for the user to open it a second time?
- will they find it valuable enough to upgrade for?
- will the sales page spark enough interest to complete the payment process?
…Ok, I think we get it.
If we don’t measure and gauge customer feedback, both qualitative and quantitative, we will fail.
This contributes to the failure of everything above.
I’ve been guilty of this myself. The first startup I ever co-founded could’ve been part of an interview for these studies — we ran out of cash, we didn’t measure properly, but our biggest violation was probably that we were not a good personality fit.
If the founders can’t communicate well, everything else a successful company depends on will fail, too.
Trendcraft’s goal is to give people a competitive advantage in the app space.
Subscribe to our newsletter to stay on top of the latest trends.
Stanford also while not calling it a study, gathered why they’ve seen startups fail too.