When startups go under, outside observers rush to catalogue the fallout. Did anyone use the product? Did they run out of money? Was the CEO too inexperienced?
Now we have some answers from startup graveyard Autopsy.io. A new site launched this year, Autopsy publishes a list of tech startup deaths, which to date includes 112 examples dating back to 2006. More important, Autopsy includes links to articles where the entrepreneurs themselves attempt to explain, in their own words, what went wrong and why they failed. For example, if you want to know why KOLOS, the first iPad racing wheel startup, or Fastr, a WhatsApp for customer service, both failed, Autopsy gives it to you straight.
To analyze this growing dataset, semantic software company Expert System performed an advanced content analysis of the articles looking for common themes in the text on why startups fail. They shared their new report with Forbes to help startups learn from the lessons of those that came (and went) before.
While many entrepreneurs worry about marketing, revenue, and employees, none of those were top 10 reasons Expert System identified as critical reasons for failure. Their methodology: what words were mentioned in the same sentence as “fail” or “failure.” While not foolproof, this led to an interesting result: “product” was mentioned 15 times in those sentences, compared to employee, marketing, or revenue at only 2 mentions each. This suggests that product is the key cause for many startup problems.
When Expert System expanded their textual analysis to include all words mentioned (not just those in sentences with “failure”), it showed that money is also critical. After all, most startups close shop at the end because they simply run out of cash or can’t prove they deserve a new funding round. “Money”, “capital”, or “fund” were mentioned by 78% of the startups. “Product” was close behind, mentioned by 66 startups, “Idea” 65, “Team TISI +%” 55, and “Investor” or “VC” 50.
Unfortunately, the data analysis doesn’t go deep enough to identify specific underlying concerns within a category like “money.” Did they fail in fundraising? Manage cash flow poorly? Or just never generate sustainable sales? These would have to be analyzed on a case-by-case basis.
The report does show that entrepreneurs tend to have specific problems (mentioned 196 times in total) they blame for their failure. More than half of those are “mistakes” (92 times) or “wrong” (39) decisions. Others simply said their business became too “difficult” (59).
Further textual data from Autopsy suggests that startups may have aspired to create something new, but couldn’t deliver on their hopes and promises. The top five mentioned action verbs—make, build, work, want, start—point to product creation. But words like “launch” (102 times total) or “sell” (80) were mentioned much more frequently than “deliver” (27) or “execute” (18). Getting the product right in practice is a lot more difficult than it sounds.
Then again, many startups may be too focused on the product, and not enough on examining the needs of the end user or the necessary steps to take it to market. While product was mentioned 454 times in total, customers (202) and market (176) came in far lower.