Tab: Startup To Shutdown
The mistakes an angel and accelerator-backed startup made and what we learned along the way.
The idea was simple: customers would open prepaid accounts at local shops and earn bonus credit by doing so. We would eliminate payment fragmentation and use the transactional data to automate retention and marketing for these shops. Best of all, customers did not need a smartphone, as we used their mobile number to create their account.
We raised seed money from a local angel investor early on, joined an accelerator and started to grow the team. We encountered a lot of issues that on their own we could have tackled, but together set us on a path to failure that we struggled with.
1. Building a Random Team
My previous startup did not see the light of day, even though I spent almost six months working on it, because I did not build a team. I thought I was a master of my trade and that I could do everything myself. It turned out that I was wrong about that.
This time around, I was adamant that I would build a team from day one — and that is exactly what I did. I found someone that I could work with, that had similar experience as me and most importantly was able to commit all of their time right away. Not long after that, I persuaded a German developer I knew to stay in London and not return home (after working at another failed startup), but to join us on our journey, as we felt we were executing quickly.
It was not until we started talking to institutional investors several months later, that we realised they viewed us as having no credibility in the space we were attacking and that no early traction, innovative approach or growing metrics would save us.
Investors wanted to see former payments, daily deal and retail executives as the team behind Tab—not three random guys trying their hand at “disrupting a crowded space”. It was not that this was a deal breaker for investors, but because of some of the other mistakes we made, it played a major part in the slow demise of Tab.
Key learning: be from the space or build a team of experts in the space you are attacking; if you can not find anyone to join you, your vision and plan may need rethinking or you need metrics that just can not be ignored.
2. Raising Too Little Too Early
As co-founders, we all had different (short) personal runways, which made money a real concern. Early on, we were lucky enough to meet an angel investor from the finance space that wanted to put the first money in to Tab. We thought this initial seed money would not only solve our cash-flow issues, but would also be a great signal for other investors and would allow us to build our first product.
It was clear that if we did not take this money, Tab would have never been more than a paper prototype — so we took the money. This started a proverbial time bomb that we would never be able to stop: we started taking small salaries that we relied on each month and exhausted our personal savings for anything else we needed. It also made fundraising and cash-flow a priority from day one.
Looking back, we should have continued working part-time until we had a simple digital prototype and early revenue before we raised any external funds. Further, the funds should have only been used for running experiments, marketing expenses and outsourcing the development of things we could not do ourselves.
While our first angel investor helped us get started, it was not a good enough signal for institutional investors, as they were looking for industry experts putting money in. It was also evident that we either needed to be a killer team that has done it before and raise a lot early and quickly, or have crazy traction and growth to get their attention.
Key learning: try to avoid raising a single penny until you have built a working prototype and have some (any) early revenue — and in a best case, revenue that can at least pay your overheads, so you can have the upper hand when negotiating with early investors. Also, raise more money than you think you’ll need and pinch pennies the whole way.
3. Building a Not So Minimum Viable Product
In September of last year, we had lunch with a friend. He suggested that we should find a way to test our idea without writing any code. We were always talking “lean” and this was the opportunity for us to walk the walk. We took their advice and instead of building an app, we printed numbered business cards that we used to identify customers in shop and we printed a simple ledger that our first pilot locations used to keep track of balances and transactions—our paper prototype.
It went very well and in a few weeks we had more than 30 consumers using Tab to prepay for their coffees and pastries, and they used their identification card to pay for future purchases. It became apparent that it was very hard for shops to handle transactions this way and they almost begged us to build a digital version to make it quicker at the point of sale and reduce human error (which was appallingly high).
The problem was that we wanted all the bells and whistles from day one: payment processing, account profiles, multiple locations, transaction histories, email receipts and more. The result was that we ended up building something completely different than what we tested and we built too many features without validating they were needed. We had to quickly start customer development again and ended up rebuilding Tab from the ground-up, which took us another 8 weeks.
We should have quickly turned our paper prototype into a simple digital prototype — nothing more, nothing less. We could have then quickly iterated, skipping features that were not needed and building features that shops actually wanted (and would pay for).
Key learning: build early and as little as possible initially; talk to your customers and iterate your product. Each iteration should be the minimum input needed to generate the minimum output needed.
4. Focusing On an Accelerator Too Early
One of the General Partners at Seedcamp happened to be the first consumer to use Tab at our very first pilot location, the cafe at Google Campus in Shoreditch, East London. We got excited about the idea of being a part of Seedcamp and what it could do for us as a company: open doors, keep our ticking time bomb going and provide us with a stamp of approval for future fundraising.
We tried to join to Seedcamp twice (once traveling all the way to Portugal and once on home turf in London) and we were accepted the second time around. Each event needed our attention for around three weeks: one week to get prepared, one week of mentoring and investor demos, and another week following-up with everyone we met. In total, we spent at least 6 weeks of time devoted to getting in to Seedcamp that could have been used to talk to our customers and iterate our product (or lack of product in the first case).
We launched our first product the day we finally got accepted into Seedcamp, so we had no product validation at this point; we were not ready to start scaling in any shape or form. In hindsight, we should have spent those weeks solely on product development and iteration and applied to an accelerator only when we felt we were close to product-market fit and were ready to start scaling.
Seedcamp has changed me as an individual, and for that I am ever grateful for participating, but as a company we were not ready to reap the full benefits the programme had to offer.
Key learning: being in an accelerator is a full-time job. You should apply when you think you have product-market fit, you have early revenue and you are ready to put fuel on your startup’s fire.
5. Going to The USA at the Wrong Time
One of the biggest opportunities Seedcamp offers new portfolio companies is to go on a 3-week long trip to New York, Boston, San Francisco and Silicon Valley. We visited some of the biggest names in technology including Google, Facebook and Foursquare, and pitched some of the best venture capitalists around. This trip really opened my eyes as to how different American founders/investors are compared with the more relaxed and timid European counterparts we were.
This trip came a couple weeks after we joined Seedcamp and released our first product — so it was not the best timing. During my time away on the trip, product and business development dropped to a glacial pace. At a time when we should have been solely focused on talking to customers and iterating the product (this is a common thing you will hear in this post), I was half-way around the world schmoozing with VCs and the tech community trying to build some hype.
It turns out that hype does not build successful companies — successful teams and products do. As tensions started to rise and shortly after my return to London we parted ways with one of our co-founders (whom we luckily remain good friends with today). We should have postponed this trip to save time (and money) so that we could have had a clear objective and reason to travel halfway around the world at time when our product needed our full attention. It is great to see how our experience and other learnings are already being integrated at Seedcamp: new companies now have two opportunities a year to go to the USA, so they can make the trip when it is best for their companies.
Key learning: stay where your market is and only go abroad when you have a valid reason to do so: such as entering that new market or you are ready to play with big boys and are fundraising and ready to scale.
6. Starting Scaling Too Early
As we had already raised some funding and joined Seedcamp, we started regularly interacting with investors as they came in to the office and at other events. The majority of the investors we spoke with warned us that they would need to see proof that we could scale Tab beyond a couple pilot locations. They wanted us to prove that we could execute our ambitious plans on a smaller scale.
Our biggest mistake was listening to these investors too much, and we started focusing our efforts on how we could make Tab more investable rather than talking to customers and iterating the product. If we spent more time working on the product, the product itself would have made the company investible, rather than us jumping the gun. Seedcamp has since started their Academy to help those teams that have not yet achieved product-market fit, which is a great idea.
Once we started scaling-up shops and building a sales team, we were no longer raising money on a vision — we were raising money on our metrics. As we only had several months of runway left, we did not have the cash to keep our early growth levels and quickly picked all the low hanging fruit we could. We should have resisted investor pressure to start scaling, especially as our product was so young and untested.
Key learning: once you start scaling you can not put on the brakes and everyone will be looking at your metrics—your grand vision goes out the door at that point. If you do not have the money to continue growing at the rate you need to, do not start until you have that cash and enough runway to raise your next round if needed.
7. Overvaluing Qualitative vs. Quantitative
At the end of the day, shops wanted two things from Tab: to make more money and to save more money. Anecdotally, Tab did just that and increased the spending of customers and saved shops money by bundling card payments. The problem was that we were not able to show this to shops.
We spent a lot of time building features that made Tab qualitatively better than incumbents, but that were not necessarily quantitatively helping shops. This made it extremely hard to convince shops to pay for Tab, even if they had processed several thousand transactions and had hundreds of customers using the service.
For example, we required no app to download or barcode to scan for customers to open an account and we passed the transaction onus on to the shop, which meant customers did not need to show anything to pay. As much as both shops and customers loved this, shops could not put a price on the value.
There were potential features that we knew could provide quantifiable value, such as the ability to communicate with customers, to view analytical information and to send rewards or perks to customers to bring them back in the door. All of these features could present a quantifiable return for shops, but we were late to deliver them.
Key learning: find out early on what your customer values the most and is able to quantify, and deliver that single thing. When you have done that, find out the second, the third and so on. If it does not make a dent to their bottom line, save it until later.
8. Not Generating Any Revenue
As we were using shops as guinea pigs for customer and product development, we offered Tab for free to shops. This made it very easy for us to sign up new shops, especially when they saw their competitors in the local area were using Tab.
We assumed that if shops piloted Tab, they would see the value and would convert later on when we had a fully-baked product. This hurt us in a few ways: we did not have any revenue being generated, which was a red flag for investors and the feedback we were getting from shops was different than if they were paying.
We should have charged shops a small flat fee from the beginning. This would have not only extended our runway by a few months, but would have allowed us to iterate the financial model, just as we were iterating the product.
Key learning: if no one wants to pay you, it is not a business. Do whatever you can to get your customers to pay you from day one—whether it is 0.99 or 99.99, find out what they will pay now, so you can up-sell as you add more value for them or charge new customers more.
9. Not Building a Financial Model Early Enough
We did not build any real financial models until one of our existing investor asked us for one. I mean, we wrote on the back of napkins and that sort of thing, but nothing that looked at how the company would scale in the coming years.
It was a shocking experience when we realised how much capital we really needed to scale Tab and how little the return was after subtracting acquisition, support costs and other overheads.
If we had built a simple financial model earlier on, we would have seen the flaws in our business model and we could have attempted to fix them before it was too late; we could have easily shifted focus to features that could have generated more revenue for us.
Key learning: as you build the product, build a simple financial model that you can iterate as the product changes. This will help you make product decisions that make a difference to your bottom line.
10. Putting All Our Eggs in One Basket
We quickly realised that direct sales was the best approach to convert shops in to piloting Tab, so we put all of our eggs in one basket and focused solely on direct sales. This meant that we never really tested online sales or teaming up with a distribution partner.
This was a big mistake, as when it came time to fundraise, we could not show that one or the other would help lower our blended acquisition costs or that is was viable for us to pivot from direct sales to online sales if needed—and potential raise money for this pivot.
Key learning: your assumptions will not always be right. You should have a Plan B ready to go, in case Plan A fails at the last minute.
What’s Next For Us?
Tab will be turned off on December 31st, 2013—until then shops will be able to debit accounts, but not accept new top ups or new signups.
We want to to thank Abdul Nusrat, Daniel Siemaszkiewicz and Gabor Valuch for helping build Tab from the early days. We are extremely grateful that James Phillips at Dose Espresso and David Abramovich at Shoreditch Grind helped kickstart Tab by allowing us to work with them and pilot Tab at their busy coffee shops.
We also want to thank Nishul Saperia for believing in us as our first angel investor, and Carlos Espinal, Reshma Sohoni & Philipp Moehring at Seedcamp for giving us a chance. Last, but not least, we want to thank all the mentors, advisors and investors that met with us along our journey—your feedback and advice gave us a real fighting chance.
P.S. Talk to your customers and iterate your product. :)
Co-Founder & CEO, Tab