Newsberry

Launched: ~2005

Shut Down: January 2012

Founder: Wildbit

Funding: none

Newsberry was an email marketing product that was founded by the then consulting company Wildbit. It was shut down after 7 years, which is a reasonable run in itself, but the unusual thing about this story is that at the time of shutting down, Newsberry was actually profitable (to the tune of $75,000 per year).

So why did they shut down? Why didn’t they sell? Wildbit have written two blog posts (here and here) explaining their decision to kill a profitable product without selling out. The crux of the problem was that Wildbit themselves didn’t particularly like the product, weren’t proud of it, didn’t even use it and basically ignored it. The second of the posts exlpains why they didn’t sell – they wanted to, but couldn’t find a suitable buyer.

This probably doesn’t qualify as a failed startup, but it does illustrate some of the complex reasons people have for creating, maintaining and letting a startup go.

Advertisements

Exambuff

Launched: ~ January 2009

Shut Down: ~ January 2011

Founder: Tim Ruffles

Funding: none

Exambuff was an exam preparation service where¬† one could “hand-write an answer to an example exam question from your course, upload it to a PhD student, and they’d comment on it and tell you where you were going wrong”. Unfortunately, it never gained any traction, nor did it ever make any sales. Around two years after starting, Exambuff shut down. However, Exambuff isn’t totally defunct – founder Tim Ruffles has open-sourced the project.

Given that ExamBuff never made a sale, it could probably be described as a failed project more than a business startup. But as a case study, it does provide some insight into a project idea that never quite found a market.

Why Exambuff Failed

Founder Tim Ruffles wrote a post-mortem and pins the failure down to three main factors:

  1. Not understanding customers – this was the big one he identified (and I entirely agree). Tim only tested his assumptions with a Facebook survey, but never bothered to talk to actual or potential users.
  2. Harmful ramifications of knowing how to code¬† – he was a technical guy who wanted to code and he essentially re-invented the wheel by creating a complex app from scratch. To be fair, I don’t think this was a proximate cause for the startup not succeeding, but rather a waste of his time.
  3. Market or industry forces – the university exam market is seasonal (exam time) and he was targetting the universities themselves, and institutional market. Again, this probably wouldn’t have broken the startup if he was more flexible in his target market (e.g. one suggestion on the Hacker News discussion thread was to target the legal niche, or alternatively build an open platform for scanned and annotated documents).

Ultimately, there wasn’t a paying market for ExamBuff, probably due to a number of separate reasons, for example: the target demographic (students) generally aren’t interested in paying such services, it’s hard to know how many students study for essay-based exams by actually writing practice essays, students can solicit third-party reviews for free from friends or at local student centres, etc.

Moreover, the site design was mediocre at best, with the dubious choice of a canary yellow theme. It was also noted that the demo was lacking, not for the handling or technical aspects, but for the actual content of the example essay. The comments were in mildly unclear handwriting and were stock, banal comments and writing advice that would have been at home in a primary school setting. Were I the customer, I would have felt it was a total waste of money.

Lessons Learned

The main lesson from ExamBuff is to test your assumptions and constantly seek feedback on your product. A common theme in many failed startups is the lack of a product-market fit. This may take some time to develop but it is clearly crucial to success. Startups need to continue to test their theories in the market and find the product that fills a need or desire that someone is willing to pay for. This may mean experimenting in many different directions to find the right fit. Now that it’s open source, perhaps someone could use the source code for ExamBuff and discover a market for it.