The big question: Why did my startup fail?

I never made this explicitly clear in my introductory post although it could have been inferred. Since the day I told my team the end had come and we’d be closing up shop in the coming weeks I’ve spent a lot of time mulling over this question.

The mulling was not in vain; I have the answer. We executed well but we executed on the wrong things at the wrong times. Ultimately it was a flawed macro strategy that did us in. There was nothing inherently wrong with any single piece of the strategy. It was only when taken in whole that it was flawed. Looking back, it would have been difficult to recognize for anyone without prior experience building a company (this isn’t to excuse my mistake but to suggest that my inexperience could have been made-up for with proper guidance from seasoned vets).

tl;dr —

We didn’t have the traction to raise a series A round by the time we needed to. We placed all our sales/marketing efforts into developing partnerships at the expense of building a user base through our own website. We underestimated the amount of time required to get travel companies to integrate our service into theirs. We ran out of resources before we were able to correct for this misstep.

Company Description: First, I think it’s appropriate for me to tell you a bit about the company so you have context when I subsequently describe the execution strategy we developed. I founded a travel company that let people lock in the price of airline tickets before they were ready to book them. A person could pay a small fee and if ticket prices have increased by the time they book theirs, we’d pay the difference. The service worked very much like an insurance service or a financial derivative (stock option). That means that for any given customer that pays us $20, we may be obligated to pay them an undetermined sum in the future, possibly lower than $20 but very often greater. Operating this business at scale would require a sophisticated statistics algorithm, and that’s what we set out to build.

Our overall goal was not to create a successful consumer facing website but to license our technology to existing travel sites (e.g. Kayak, Expedia) for them to offer to their customers. The rational behind this was:

• we were very aware of how expensive it is it build a consumer-facing site in the travel space
• we could leverage their existing large customer base, rather than build our own traffic
• our service was a niche product that only appealed to people at specific times and therefore only made sense in the context of another travel site that sells tickets — “not ready to buy yet? no problem, lock in the fare here and buy later”

The Strategy: Given these beliefs, our overall strategy was to:

1. build a system for collecting airfare data to train our algorithm

2a. build a pricing/statistics algorithm to help us estimate and manage risk

2b. build a website to demo our service to potential partners (i.e. online travel agencies)

3. begin pilot test with small travel agency to show proof of concept and traction

4. raise money so we could expand to serving larger travel companies

The Flaws: This strategy assumed that if we built a service that people undoubtedly were interested in, then we would have no problem finding partners to integrate our service. We had several months of runway with a team of five so we were confident we could make it work.

We spent many months toiling away, building the perfect system to anticipate airfare movement and estimate risk associated with locking in airfare. We did this at the expense of focussing on customer acquisition early on. We didn’t make any significant efforts to gain traction because we felt the time/money spent acquiring these customers would not be refelctive of the cost of acquisition with our technology integrated into partner site. We did generate decent amounts of site traffic and this all came through PR which was very easy for us to get because of the novelty of the service. We accepted this as validation that the market was ready for our product and all we had to do was build it.

After perfecting our algorithm and building an API exposing it to potential partner sites we began our business development efforts. We’d gotten to know people at many of the online travel agencies through conferences and many of them were very interested in what we were building. They understood that variable pricing in airfare was a huge customer pain-point that wasn’t being properly addressed.

What we didn’t anticipate was the length of time these travel companies required to integrate a new service into their own. Some of the larger companies, we found, could take as long as a year to test and fully integrate a new service. We simply didn’t have the runway to support this timeline. At the same time as we got deeper into discussions on integration, it became apparent that many of these companies wanted to see how our service performed on other partner sites, data that we didn’t have. Concerns they had could have probably been partly alleviated with significant traction on our own website. We had not anticipated the importance of this and for that reason didn’t have significant data here either.

As time went by, we continued to build stronger relationships with various travel sites. Many conversations seemed promising but it was clear they required many months to actually make something happen. When it was time to fundraise as planned, we didn’t have the on-site traction to convince investors we were on to something big and we coudln’t prove that any of these conversations would ultimately lead to an executed deal.

Recommendation/Lesson learned: If your goal is to keep a company alive and you aren’t sitting on millions of dollars enabling you to devote all your time to product development alone, focusing on traction and product-market fit should take first priority. We had a hunch (probably a correct one) that people would like our service or something similar to it. Perfecting the algorithm enabling us to scale isn’t really an issue when you don’t have huge amounts of sales.

The ultimate irony here is that a competitor of ours focussed exclusively on building their consumer-facing site and iterating on their service. They raised a modest series A round on that alone. Now they are close to launching a pilot program with a respectable travel agency and will likely see a huge increase in traction. They’ve let me know their pricing engine is pretty flimsy and unsophisticated, but had been good enough up until now. Knowing that we’ve decided to close up shop, the CEO has approached me asking about buying or licensing our technology to bolster their own. There was a time when I didn’t respect their approach but they’re still alive and kicking and we aren’t.