HitMeUp was originally conceived as a deal publishing platform for businesses. The idea was that businesses could react to lulls in demand by instantly creating promotional deals and immediately pushing that deal to as wide an audience as possible.
The below is my post-mortem that I put together for my investors analysing not just what I thought went wrong, but also pointing out some things that investors should look out for next time so hopefully they can avoid making the same mistake twice.
We broadly went through 3 iterations:
Version 1: Web-only where suppliers could create group buying or flash sales deals and publish them to our website, their website, their Facebook page and to banner ads for affiliate distribution.
Version 2: Version 1 plus a mobile app for businesses but also with a separate customer-facing mobile app for aggregating and displaying all deals.
Version 3: Mobile-only app that allowed user-generated content with the idea of allowing people to tell others nearby of what was going on nearby, into which location-based ads could be dropped.
In theory the idea that a business can instantly adjust its price-point to react to supply and demand is a good one. In fact, we know from the speed at which businesses signed up for this idea (on Version 2, when we got the user-experience right) that it is something very attractive to businesses - but somehow we didn't manage to harness this demand. We went through these three different versions of the business and what I will do is sum up why I think each iteration failed, and then draw out some more general lessons learnt.
I hope by doing so I can pass on some of the things I have learnt so that you may know what to avoid if/when you decide to invest in a start up again.
Version 1 failed for 3 reasons:
1.) Our hypothesis that F-commerce (selling on Facebook) would be the next big thing was wrong.
2.) The time taken to build the application was so long that by the time it was built the market had moved on and deal platforms had fallen out of favour.
3.) Sellers are rubbish at selling themselves.
We were the first UK company to have an application approved by Facebook that allowed E-commerce within the Facebook ecosystem. There were some in the US but our agreement with HSBC allowed us to take payments on behalf of third parties and this really was the USP of the original platform. At that time Facebook and 'social shopping' was seen as the new frontier of E-commerce. As the first investors may remember, we were lucky enough to test our business model early on with the handbag deal promoted by Plan (a charity) alongside PR in the national newspapers. Over 1.5 million impressions converted to just 12 sales. Put simply, we did something no one else had done before, tested it, and found we had our numbers wrong. The conversion rate was dramatically lower than traditional E-commerce channels. It turns out people are on Facebook to look at what their friends are up to, not to shop. You don't really hear of F-commerce any more - that's why.
The second thing was the time taken. I had no tech knowledge when I started HitMeUp, so not only could I not build a simple MPV myself, but also I did not even know how to get someone else to. That meant twice I paid for developers to build Version 1 and twice I got ripped off and had to start again. When I met Alex and Piers things sped up, but our MVP was still far too complicated. We stripped it down but it could have been far simpler. When we made our first sale in May 2012, I had already spent a year and half on it and way too much money - but that wasn't the worst bit, the worst bit was that just as we were getting started Groupon had had its IPO and the stock was rapidly tanking. 'Deals' was a dirty word and the reviewers who I had originally signed up to scale the business were put off by doing deals and walked away.
On the plus side we were getting interest from businesses wanting to sell stuff. However, the big flaw in the business model at this point was the self-publishing. This was supposed to remove a large cost-layer from the traditional deals platforms, but it turns out those expensive sales teams who write copy are needed as businesses are just terrible at selling their own products. Businesses would put stuff on HitMeUp with terrible copy and photos, thereby making their products look totally unappealing. Also, often when they created a promotion they did not tell anyone. They did not want to promote a deal to their existing customers, they expected us to provide new customers via our own audience. This lead to Version 2...
Version 2 failed because we did not manage the balance between the number of customers Vs the number of businesses.
Our seed round was aimed at getting us off the ground for Version 1. It didn't work, we had very little money left and so with the remaining budget we took what we learnt and stripped down to the very basics of what you needed for a deal: what it is, where it is available, and how long it is available for - What, When, Where. The focus here was on user experience. You could create a deal and publish it in under a minute (including Twitter now), the customer app showed live deals, the vouchers were unique and trackable, and the business app scanned the customer app's voucher thereby solving the age-old problem of voucher redemption. This is was a neat bit of technology and what got us on to Seedcamp and then Collider.
The problem was that it did not have a defined marketplace. The underlying technology is exactly the same as Hailo - which has an app for the drivers and a corresponding app for customers to hail the drivers. The difference though is that it is limited to taxis - so everyone knows what it's for. HitMeUp had hundreds of different businesses sign up - restaurants, bars, shops, etc - it was great but that meant it was chaotic. It was un-curated and the content was generated by the businesses (unlike YPlan that launched the same week) so as a customer you arrived on this app and did not know what the app was really useful for. This made the customer-acquisition troublesome.
That leads to another problem - market liquidity. In the first month we signed up 400 businesses and about 4000 customers. However, because this was location-based deals you need an awful lot of stars to align for:
- A business to create a deal
- The customer to like the deal and get it
Through some complex calculation based on residential square miles in London, iPhone penetration and usage and some other social inputs we worked out you needed 225 customer downloads to every business download to make the app useful. The problem was that the app was too popular with businesses and that meant we mismanaged the liquidity. What
quickly happens is that not enough vouchers are being redeemed so businesses fail to get the benefit and then drop off. If you are going to invest in a marketplace, have a think about how many market participants are needed in order to make the market function reasonably efficiently and then ask how the business is going to get from zero to that number.
Returning to that defined marketplace issue, in hindsight I wish we had restricted it to just street food to start off with. We had all the traders behind us, they could have done the promotion of the app and the location-based technology suits the mobile nature of heir business. I decided not to restrict it as I didn't think street food was a big enough market and I ambitiously wanted it to be for 'everything' which was stupid, unrealistic and what ultimately made Version 2 fail.
Version 3 failed because the user experience was not fit for purpose.
Apps are extremely competitive, particularly consumer ones, and when it comes to it you need to spend money on advertising. We just didn't realise this and arguably when we got the Collider 12 money we should have ironed out a few things with the Version 2 app and spent the rest of the time and money marketing it. We didn't though because we were afraid of the scale of the challenge (because the marketplace was undefined). Instead we tried to increase the chances of success by looking at the stars that needed to align (above) and work out how we could help that. We needed something that would make the users frequently return to the app. That something had to be around time and place for the deals part to work and so we set upon allowing customers to add to the map by taking pictures of things and sharing them with people around them. The idea was to create an app that showed what was going on around you right now. Into which could be dropped geo-tagged ads and deals.
This is quite a bold idea, and I met with Facebook and Google both of whom had looked at something like this but not cracked it. This year many others have failed too. One, Circle, launched in the last couple of months shows early indications that they may have cracked it. What have they done differently? I would say design. Not just that it looks beautiful, which is a huge thing in itself, but also the way in which the key elements are displayed. The key pieces of information you need for an app that tells you what is going on around you now:
- What is it (text description)
- Who provided this information
These bits of information make up the core elements of the database. Those are the exact things HitMeUp V.3 had and exactly what Circle has. I know several of you thought we were crazy to think anyone would care about a social network for people nearby. I am please Circle has done so well as it proves me and the team weren't mad! The problem is that we did not have the right user experience and the design focused on the map as opposed to the content. We did this because we wanted to introduce the geo-tagged deals therefore the precise location was important. History relates that we were wrong and that precision is secondary to other things (photo and text). We were rushed due to thinking we had to get something for Unilever by the end of the Collider programme so went in to overdrive building the app based on the concept without really thinking through the use-case properly.
The first member of the team was myself. I came from a strong business background, had plenty of passion, energy and enthusiasm and I also had cash to fund the business through the early days. However, what I really lacked was tech knowledge. This was a huge hindrance: not only could I not code so could not build the MVP, I did not know the market well, I was not well connected in the space and also I did not know how best to get things done. This meant I was dependent on other people - something I can not stand. The really hard thing was that I did not really know what skills I needed from others, and there a plenty of cowboys in tech, so a lot of time was spent learning and I got ripped off along the way. Now I have this experience my advice would be to never invest in a tech team with a single non-tech founder.
That leads on to my second problem - I lacked a co-founder. Starting a business is bloody tough - there is just so much to do and my answer was to just work harder. In hindsight working 7 days a week isn't that clever. I should have had a co- founder who could fill some of those skills I lacked, been a sounding-board, a friend, a person to share the ups and downs with. I would strongly advise only investing in start ups with at least two founders.
Once I had the funding to hire Chris as CTO things completely changed. Chris is the best developer I have ever met. He was expensive, but in my mind that expense was justified by the sheer pace with which we got things done at after he joined.
However, he was our biggest expense and so again I would say if you could invest in a founding team that has the skills to build the MVP then that is a huge advantage. We were constantly being told to be a 'lean startup'. Note that it is undeniable fact that a team that has to pay for any skills required to build the MVP can never be a truly lean startup, so expect a higher burn rate for those teams and factor that in to your investment decision.
Tom and Max joined after we had build Version 1 as salesmen. My first email from Lord Stafford gave me some advice he had learnt in the dot-com bubble which was not to hire expensive sales teams early on. I was very conscious of this and now with hindsight have something to add to that. Tom and Max were not expensive (in fact the poor guys were paid well below minimum wage) and were the most dedicated, loyal and enthusiastic advocates of HitMeUp we could have ever have wished for - and as a pair two of the best sales people I have ever met - but the lesson learnt is that you need to have a product before you can start selling it. Thank god we did not have an expensive sales team because what we really needed was a designer and a front-end developer to get the product right before we got the sales team. If you don't have the skills in-house to constantly test and tweak the product in the early-days then I'm afraid your chances of success
Lessons learnt - Product
As those of you who have followed our progress on my blog will know, I kept a record of any things that came up which I wish I had thought of before I had started. As I started to gather these thoughts in some form of structure to make some sense from it all it occurred to me that there were recurring themes around the notion of 'definition'.
Essentially, and somewhat predictably, the lesson learnt is to have a crystal-clear explanation for what it is you are actually doing. I know that sounds obvious but clearly articulating what your product does is the nucleus of the entire business.
Failure to be able to clearly define the product results in total failure.
I go on a fair amount about this on the blog, but there is one section I believe could be relevant for an investor: the post on Facebook Ads. This post is written for entrepreneurs, but you may well turn it on its head and ask a Founder you are thinking of investing in to provide you with the three elements required for a Facebook Ad.
Ask them for:
- A one line description of the business
- An image they would use for an ad
- A description of the target audience
As a start up you need to do one thing extremely well. Without generalising too much, regardless of what it is they do, if they can not answer those questions with confidence I would say walk away.
Our final product was not thought out. We were trying to build something we thought could fit with brands. We rushed and it should have been better designed/planned/thought out. In hindsight this was plain stupid as we should have stuck to our own guns. Building for a corporate is a ridiculous idea as they don’t know what they want, don’t take risks and can not make a decision without asking their boss, who has to ask their boss, who has to ask their boss. No one is capable of thinking for themselves – as can be seen from Bauer taking eight months to make what ought to be a relatively simple decision – they move on a different timeline to a startup and you can not rely on them.
Lessons learnt - Apps
I would never try and build a business around an app again! I would discourage any investor from investing in a business which has an app at its core. I do not have a problem with websites that have a mobile version, but an app on it's own is infinitely tougher:
1.) Getting the app installed
People don't go online and download apps there and then, they download apps directly from their phone. This is critical as your website receives ow traffic comparative to businesses based around websites. If they do find you they do not become a customer easily - they have to get out their phone, go into the app store, search for you and then install the app. This makes the conversion particularly tough.
Apps compete for a very small shop window in the app store; if you are not promoted by the App Store the chances of anyone discovering you 'in the shop' are very slim. A significant number of apps never get triple digit downloads.
2.) Competing on your phone
On average people only regularly use 5 non-native apps on their phone. To try and get your app into someone's Top 5 apps that they use every day means you are likely to be a household name. Those that are not used regularly yet not deleted are very good at doing something for a specific purpose (e.g. Hailo) - again the theme of defined marketplace is critical.
3.) Apps are hard to share
The tools / infrastructure available for sharing apps is not as streamlined as web. There are not like buttons everywhere and when you share something it tends to be shared to a web-based platform. If the app has inherent virality (i.e. requires more than one person to make it useful) that helps, but if it is one that gets more useful the more people on it then you need to ask at what point is there some critical mass and what are you going to do until you reach that point?
4.) Apps are easy to delete
If the user has a poor experience of an app they are unlikely to return and very likely to delete it. It takes seconds to delete an app and is unlikely to be reinstalled.
As we know, our team was long on marketers. In terms of guerrilla marketing ideas I think the team excelled - the partnership with a huge international charityfor V.1 and the V.2 launch party (EatMeUp, the pop-up Christmas market) getting Time Out's Pick of the Week were examples of our skill on that front - but as we are doing a critical appraisal for investors I think the criticism would be the lack of digital marketing knowledge, as opposed to skills.
The arty stuff - creating content, distributing it on social media and driving the inbound marketing is the (relatively) easy bit as the channels are well known. The hard bit is the scientific stuff - the measuring, the analysis and the optimization. The tools for gathering and displaying the data are virtually unknown outside the tech community. They require specialist knowledge to set up and run and require a totally different skillset to interpret and act on.
As it turned out, once we became aware of these tools we quickly became good at gathering, interpreting and acting on data. This largely became Tom's role, but thinking from an investor point of view, I would take it as a strong positive any company that has a team member who knows this area - as it is these data-informed decisions that help fine tune a digital product. A team that understands the data process will save a huge amount of time and money.
Lessons learnt – Sector-specific knowledge
On the more creative side, a good blog can go a long way. I wouldn't claim to be the world's thought leader on group-buying, but in the online world there was a time when I arguably was. This was mainly because I wrote a lot about it: on my blog, commenting on other websites and answering questions on Quora. At one point my blog had over 10,000 followers. I became the thought-leader on this subject just because I had written so extensively about it. If I were to invest in a business I would look very hard at the Founder's passion for tackling the problem, and my first place to look would be Google - for a quick search for their expertise in the public domain. Thought leadership and passion for the subject are huge plus points.
At this point it is worth pointing out that I was successful in getting the HSBC merchant account. Although it was not used much it is worth noting that another company, who asked not to be named but are one of the most progressive technology companies in the country, attempted to build something similar to HitMeUp Version 1. They had better resources available and, except for the payment gateway, they built this in just a few weeks compared to my 1.5 years (albeit at three times the budget!). However they never got a merchant account which meant they could not take payments so having built 99% of it they never got it complete. l got the merchant account because I worked with HSBC to mitigate the risks through a self-insured structure. Although they failed fast which has it's benefits, my point here was that ultimately I got further than one of the best tech teams in London, simply because I understood the banking side. If you're going to invest in a business that requires very sector-specific knowledge, make sure the team has that knowledge.
Lessons learnt - Design
A big lesson for us was that design is everything for a consumer-facing business.
A good product with bad design will go no where. We saw this with Version 3. Version 2 looked great and everybody thought it was - the lesson there being good design hides faults!
Design is not just how a product looks but also how the user interacts with it. Every touch point of the customer experience (website, app, emails, business cards, poster, etc) has to be designed with the customer in mind. The fact is you will not be taken seriously if any of those touch-points fail. You are building a brand - that is what sells.
We understood this, but as I mentioned before, the team lacked a designer. In the early days I think you need a full time designer. Preferably one who understands interaction design as well as graphic. The reason being that you spend so much time refining and tweaking the site/app that unless you have a design resource on tap you are totally restricted to make those changes. Given a preference as an investor I'd say if you are going to have two Founders with different skills you should pick a developer/designer duo over a developer/marketer duo for the simple reason that you can do the most amount of work done creating and refining the product before bringing in outside (i.e. paid for) help.
Lessons learnt – Investing in tech
One of the things you may not expect me to say is my advice would be to not make direct investments in pure tech start ups unless you understand the tech.
The London start up ecosystem is pretty good, and it if you have got a good product it really isn't hard to raise money. The seed funds do a great job of getting their names out there, they are approachable and given their high profile investments, they give off the impression they are the gatekeepers to fast-track success. As a result they are first port-of-call for start ups. It makes no sense to be a VC if you're good at finance, but to be a good VC in early-stage tech requires no finance knowledge, it just requires knowing the tech - can the team build the product properly and is there a market for this? The good seed funds are made up of ex-techies, not finance people and they will know that answer in seconds. They snap up nearly all the low-hanging fruit - and by that I mean the pure tech: the ones that require no physical product, can be quickly proto-typed, tested and tweaked then scaled rapidly with small, geographically-agnostic teams.
If you're looking for the next Facebook and don't work in tech then there is very little chance you'll find it, but there is a chance one of these guys might, so I'd say invest in one of these funds.
We just sent you an email. Please click the link in the email to confirm your subscription!
OKSubscriptions powered by Strikingly