Biopsy of a failed startup

In 2014, in my last year of university myself and two friends co-founded a backpacking app called BackTracker. We graduated in June 2015 and immediately went full-time building the company. That summer we hired two developers bringing our team to a total of five. Fast forward 9 months, almost 2 years after founding it, and we decided to call it a day on BackTracker. We sold the assets and moved on.

I wanted to document what I personally believe went wrong, with a few underlying lessons that I will take away from the experience and apply to my next venture. I originally wrote this ‘post-mortem’ purely for myself and for a few others who had supported us along the way, however several people mentioned that it contain an insight or two that might be helpful to others currently or soon to be in my situation. So I’ve decided to put it out there in the open, in the hope that anyone reading it can avoid the same mistakes that I did. I’m sure much of what I mention is obvious to most, but this is what I learnt.

BUSINESS MODEL

1. An easily understandable and bullet proof business model

We took too long to strip back the business and get to the heart of the business model, the nuts and bolts of the business, in our case the product: what precisely it does, what the value proposition is, what the different components are that absolutely must function for it to succeed.

After 8 months of being live, we finally woke up to the fact that our app was doing too many things. It was feature heavy, which thus made the value proposition confusing: were we an app for blogging your journey? Or a discovery tool? Or a way to connect with other travellers? In the early days we actively defended an all-in-one app, an app that does everything; however our low retention forced us to dive into the product — in doing so we discovered the error of our ways and made the decision to focus on just one aspect, the discovery tool.

The following business model was much simpler, and the value proposition was very clear — an app for backpackers to discover things-to-do in their destination. Even so however, we were never able to really hone the model; questions that still lay unanswered were what precisely did backpackers want to discover, and how were we going to provide those things-to-do. Were we going to have users suggest them? Were we going to source the content ourselves? Were we going to rely on businesses to submit themselves?

These unanswered questions left gaping holes in our business model. True, many businesses begin with one model and then pivot to another; but where we fell short was that we didn’t choose one. We should have picked one, pursued it and had it not worked, pivoted to another. Instead, at no point was it absolutely clear what we were.

2. Market research and/or traction to support the business model

A simple, easily understandable business model can still be fundamentally flawed. Selling sand to the Arabs is straightforward, you have a seller and a buyer. The problem is that Arabs already own a lot of sand and realistically don’t need to buy any more. We made the same mistake with our business model, even after simplifying it. At it’s core it couldn’t have been clearer, the business had two elements: content (suggestions) contribution and content consumption. However by not conducting market research early enough we didn’t answer whether people wanted to contribute content or consume content in the first place.

In this respect we perhaps had an unfortunate piece of good luck: our position on the App Store meant we had a constant stream of downloads. ‘Good luck’ because it meant we had exceptionally low marketing costs, ‘unfortunate’ because it blinded us to the reality. The high acquisition metrics, in our minds, validated the concept; in reality we should have been listening to our low retention metrics and discovering whether people even wanted what we had built.

When we eventually did this, we discovered that backpackers were far less excited by the prospect of a discovery app than we had anticipated. Yes they were looking for suggestions on what-to-do but couldn’t see how we were any better (or different) to Lonely Planet, TripAdvisor and other apps / websites. In itself this wasn’t terminal information, but it did mean that we should have been differentiating ourselves in a way(s) other than simply being backpacking-specific. Had we known earlier that apathy was a large obstacle for us, we would have been able to tackle it head on.

More important however was that we ignored the question of content contribution. “Fake it ’til you make it” is a dangerous tactic because it can also deceive the fakers themselves. Having contributed content ourselves to ensure there were suggestions on the app for new users, we had been able to avoid answering the question of how we were going to source content in the future. This tactic to avoid a blank slate is justified and perhaps even necessary, as long as it is used to kickstart organic content creation, not avoid it. Because we hadn’t felt pushed to answer the question of content creation, we were very late in conducting the appropriate market research: in doing so we discovered that businesses were not at all as cooperative or proactive in submitting their tours and activities to the app as we had thought. The two alternatives were to rely on users to submit suggestions or write all the content ourselves; the former had traditionally been achievable but impossible to monetise and the latter was obviously expensive, time-consuming, difficult to monetise and no different to Lonely Planet’s model. There was clearly a fundamental flaw to our business model that would have been revealed had we properly conducted our market research. It might be impossible to get traction without first ‘faking’ it, but it is essential to ensure that the action you’re trying to encourage is actually a desired action in the first place.

3. Conclusions

The first mistake we made was not paying enough attention to the contradictions and potential risks to our business model. These issues became apparent as soon as we began fundraising because we were forced to answer the difficult questions. Consequently, a valuable exercise in hindsight would have been to meet with a bunch of people in the early days and get them to play devil’s advocate with the business model. The second mistake we made was not to conduct thorough market research. In one month of being in Porto in hostels and talking with local operators our assumptions were falsified. We should have taken a month to do that before we got started. Traction is of course the best way to ensure that your idea has legs, but it’s vital to understand that traction — pure downloads for example isn’t a sign that everything is well.

TEAM

1. Be clear with why each member of the team is there

A defining factor in our failure was that we didn’t have the right team for the job. The reason for this comes down to not knowing precisely what we needed done and who we needed to do it. We may have been five first timers, straight out of university, but ultimately this inexperience was not what let us down. You can learn to run a business by trial and error, and by surrounding yourself with people who can advise you; in a similar way you can learn to manage people. But as individuals we weren’t good enough to do what was required in our respective roles.

In this respect we made three mistakes. Firstly we went into business as friends based on friendship, rather than friends based on skillset. Secondly, we hired people for key roles who simply weren’t good enough. Thirdly and most importantly however, we were not honest enough about which precise skills we needed in order to make the company work. You can open a restaurant with a chef who has never managed a kitchen, but she at least needs to know how to cook.

Had we set out precisely who and what we needed we would have probably thought twice about who we would have brought on, and ensured that those we did bring on could one hundred percent do what was required of them. There was for example perhaps no need for a marketing position at a time when we only needed to concentrate on product development; this lack of clarity regarding who should be doing what, and a sense of ‘too many cooks spoil the broth’, caused tensions between the team. Similarly, the most important thing we needed to have been doing in the last 9 months was gathering data on how and why users used the app, and making the necessary improvements / changes. Not once in 9 months did the app function properly; we never really had a proper opportunity to power forward and this ultimately came down to not having hired someone good enough in arguably the most important role in the company at that time.

2. The co-founders need to have the same levels of commitment

Another factor that caused tensions was around how committed each co-founder was to the company and what each was prepared to sacrifice to make it work. Commitment in itself was not a fatal factor, but it exacerbated existing grievances, and had we got further down the line this question of commitment would probably have caused major problems.

As a team of five 22–24 year olds, we were taking less risk in starting a business than a team of founders in their late twenties / early thirties who had families to support and who perhaps were risking their existing career. It was the perfect time to start a business because we could afford to fail; but this meant that personal development was placed first in people’s list of priorities. Every person worked undeniably hard and went through indisputable levels of stress — there was never any issue with someone not putting in the effort — but this was being rationalized as an opportunity for learning as opposed to for the good of the company.

Consequently, there was no confidence that the founding team was in it for the long haul. Some were prepared to put in more money, others weren’t; some were prepared to make sacrifices, others weren’t. Like any relationship, all parties need to meet each other in the middle; it is absolutely essential that each co-founder has ultimate confidence that the others are in it to the same extent. This may not cause problems in the early days but it’s important to determine before it’s too late.

3. Conclusions

We rushed into building a team. We should have been clear about precisely what we needed and then brought on the right person. In roles that were vital to the success of the business, we didn’t have the right person. It’s necessary to flesh out before approaching anyone precisely what roles are needed at that moment in time; business models and strategy change, so there’s no use predicting what and who you may need in the future, you should instead only choose to bring on those indispensable at that moment. In our case, the only person we should have brought on in the beginning was a developer, all other positions could have been filled subsequently. Furthermore in the early days there’s no time for you to wait for people to learn the necessary execution skills, you need to be sure that from day one they can do what is required of them.

STAGE

Ultimately the fatal mistake we made was not answering these questions soon enough, and the reason for this was that we were never honest with ourselves about the stage we and the company were at. A combination of fortuitous early traction and inexperience resulted in the misguided belief that we had proved the concept and were at a point where we could concentrate fully on growth. The confidence we had that this was where we were at, definitely rubbed off on others that we met and on those who were helping us. This meant that there was also external pressure for us to be growing quickly, so even when doubts creeped in early on we weren’t able to stop and listen.

Had we accepted that we were in fact much earlier in our journey, we could have come to terms with and focused on resolving the issues above. We would have been able to strip back the business model and iron out any contradictions; we would have been more proactive about conducting extensive market research as opposed to justifying actions on our own experience of backpacking and from the few backpackers and users that we spoke to; finally we would have been able reflect on the team and examine precisely who we needed and whether they were capable.

In the end there was pressure both internally and externally for us to be running, when in fact we should have been walking. Being honest and clear as to what stage the business is at, is essential for managing clarity, focus and expectation and the key component for being able to answer all other outstanding questions.

CONCLUSION

18 months after co-founding BackTracker, in February and March 2016, we realised that we were not as far as long as we had thought and that it was necessary for us to take a few steps back to continue forward; unfortunately the steps we took back weren’t large enough and it was only after intense market research and product focus in March that we realised we really needed to go back to basics. Ultimately if we were in September where we were in March then we could have solved the issues that in the end were our downfall. Unfortunately at this point, commitment issues became apparent and it was clear that the team lacked the drive and desire to essentially, as we saw it, start again. BackTracker didn’t fail because of any issues with the market or core concept, but purely because we took too long to get the basics right; and by the time we did realise, it was clear the willingness to do what was necessary was simply no longer there.