CBInsights analysed 101 startup post mortems and found the following reasons for failure:
Many of these can be prevented with discipline. My friend Stephen Allott who was CEO of Micromuse, a UK startup that peaked with a $3bn valuation on NASDAQ, once described managing a startup as a process of identifying problems, putting a box round them and then finding and implementing solutions. Taking that approach it is possible to avoid failing for many of the reasons on this list.
E.g. failing because the team isn’t right is preventable in most cases. It’s difficult, because it takes discipline to look at team questions thoroughly and real courage to address issues when they arise, but discipline and courage are two of the things that separate great entrepreneurs from the rest.
Going further down the list, poor marketing, ignoring customers and losing focus are all also questions of discipline and execution.
Going back to the top, even failing because there is ‘no market need’ shouldn’t really happen. Taking the ‘identifying problems’ approach I described above you would make establishing market need the number one priority. That puts you in the mindset of testing demand before you put much effort into building a company, and if it turns out there is no market lead it feels more like an experiment that didn’t yield the result you hoped for than a failed company.
The big take away here is to be structured and deliberate about the way you build your business and to face the hard problems first.
In my fifteen years as a VC I must have seen well over a thousand pitches and advised over a hundred entrepreneurs on their pitch decks. Here at Forward Partners we work with very early stage companies which has meant more time spent helping craft stories to raise investment than when I was at DFJ. In short, ‘how to pitch well’ is an important topic for us and one we spend a lot of time reading and thinking about.
Tomasz Tunguz of Redpoint Ventures writes a good blog, and given the above when I saw his headline The secret ingredient to the best startup pitches I clicked straight through.
The secret, which I’ve already given away in the headline, is to cultivate a sense of inevitability. Tunguz put it like this:
The most successful pitches argue the market will unfold inexorably in the way the founders envision on a relevant time scale. And, that this startup in particular will dominate share in that new world.
That’s great advice. When I think back to the investments that I’ve been most excited about it’s when I’ve been certain that the world will develop in a given direction and that other investors haven’t yet woken up to the new reality. I haven’t always been right, but at the point of investment I have felt strongly that a certain market outcome is inevitable.
There are a myriad of techniques for making a story convincing, but everything starts with the conviction of the storyteller. Entrepreneurs following this advice should make sure they really believe, and then be clear about the inevitability they are predicting and why it’s happening. Simply pointing to trends is much less powerful..
I wrote the post below for Crowdcube, one of the UK’s leading equity crowdfuding sites. It went up on their blog yesterday.
Two developments have changed the face of startup investing in the UK in recent years.
The first development is increased capital efficiency. Entrepreneurs can now achieve an awful lot with very little money. We see this all the time at Forward Partners where we invest right from the idea stage and most of the companies get a first version of their product live for less than £30k (that generally includes founder salaries and time spent doing customer research). When startups can do more with less money the returns from investing small amounts of money go up, and that can be seen in rising Series A valuations and declines in the average amount of money raised before exit.
The second development is SEIS and EIS. Along with many other governments around the world ours believes that a healthy startup ecosystem is critical to the future long term economic health of the country. That has resulted in a number of policies designed to stimulate startup activity of which the most important is the Enterprise Investment Scheme (EIS) and it’s younger brother the Seed Enterprise Investment Scheme (SEIS). Both of these schemes use the tax code to make it more attractive for high-net-worth individuals to invest in startups.
These two developments combine to make startup investing much more exciting than it ever used to be and we have seen a massive increase in the number of individuals who want to be angel investors.
However, finding and assessing investment opportunities is still a difficult and time consuming business. I’ve been a venture capitalist since the first internet bubble and what I’ve learned is that it’s hard to assess investment opportunities unless you have a thorough understanding of the company’s product and market and unless you’ve spent some serious time doing due diligence on the business. Similarly, from the other side of the table fundraising is still a tough thing for entrepreneurs to pull off and building a company remains as hard as ever.
Forward Partners and Crowdcube are both attacking elements of these problems.
The great thing about Crowdcube is that it makes it easier for people to invest in startups. For me the most important things they do are select start-ups they think are likely to succeed and give investors a standardised investment process that’s almost as easy as buying a movie on iTunes. That makes it possible for investors to cost-effectively invest small amounts and make lots of small bets on different startups.
Rather than help individuals to make investments Forward Partners focuses on making it easier for entrepreneurs to build great companies (it’s still hard though…). I could give lots of examples but perhaps the best is the way we work with solo-founders. Rather than pitching their powerpoint idea to potential technical co-founders, solo-founders who take investment from us pair with our developers and customer development team to quickly validate their idea and launch a site. Then, after a couple of months they can work with our Head of Talent to find a co-founder from a position of strength. We recently published a case study of how that played out for our portfolio company Snaptrip.
Whilst we have different models both Crowdcube and Forward Partners work to make investments in startups, and whilst we haven’t done a deal together yet I hope that we might do in the future. We add value in different areas and should compliment each other nicely.
Vinod Khosla wrote on Techcrunch yesterday about the reinvention of medicine and whilst he didn’t mention artificial intelligence (AI) by name his underlying message was that within a decade AI will transform the way patients are diagnosed and prescribed. Medicine is becoming big data and predictive and learning algorithms are the tools that will drive insight from that data.
The current system isn’t very scientific at all. Khosla describes it thus:
Much of the current practice is driven by conclusions derived from partial information about a patient’s history and symptoms, incomplete medical understanding based as much on opinions as validated science, and interacting subjectively with known and unknown biases of the physician, hospital and healthcare system.
And in the future:
Hundreds of thousands or even millions of data points will go into diagnosing a condition and, equally important, the continual monitoring of a therapy or prescription. Companies like Quanttus are proposing 10,000 vital sign readings per hour, not to mention Applied Proteomics, which already gets 300,000 biomarkers from a blood sample, in addition to thousands of genes, their epigenome, microbiome and more.
During the next decade, we will see systems providing “bionic assistance” to physicians and other healthcare professionals, allowing them to perform at substantially improved levels of expertise like the very best specialists in multiple domains. Inevitably, over 20 years, the majority of physicians’ diagnostic, prescription and monitoring functions will be replaced by smart hardware, software and testing.
Humans simply aren’t good at analysing the volumes of data we will shortly be working with. There will probably still be a role for doctors but it will most likely shift to the empathetic and ethical side of medicine.
Important to note, however, is that early versions of these future systems will most likely seem silly. Khosla makes an analogy with early versions of mobile phones which were attached to the floor of the car and connected with a silly cable. These early systems will be used by far-sighted doctors to improve their patient care and get better over time. It was maybe twenty years after those laughable mobile phones that billions of people had smartphones and there was a vibrant app ecosystem.
All this should lead to big increases in the accuracy and timeliness of diagnoses and big reductions in the cost of healthcare.
I think we will see similar developments in many other spheres. In step one the AI is laughably silly in some aspects and simply serves to augment the human delivery of service and then in step two the AI starts to replace the human altogether in parts of the service delivery.
At different times over the last year or so I’ve been writing about combining design thinking with the lean startup and the importance of doing great customer development, and it has just come home to me that the underlying point of it all is to add customer empathy to the lean start up process. It is all too easy to have an intuition about a customer problem and then jump onto the build-measure-learn loop without ever having enough understanding about the emotions that drive your customers. Everything is then too scientific.
Steve Blank famously advised founders to get out of the room and meet customers. That’s great advice, but it’s not sufficient. It’s imperative to not only meet with customers, but to understand them and that requires more than just interviews and surveys, it requires ethnographic techniques.
Full blown ethnographic studies are too slow and expensive for the startup world, but simple lightweight ethnographic techniques such as asking customers about how they experience the problem you are trying to solve and observing them using prototypes and later your product lead are still powerful. They will give you a much deeper understanding of your customer and better intuition about which features to build for the first version of the product and as you iterate. The goal is to be able to see the world as your customers see it rather than as you see it. If you have target multiple personas you will need a separate understanding for each one.
All this has come clearer for me after reading The service startup :: Design gets lean by Tenny Pinheiro which was passed to me by my Partner Dharmesh a couple of weeks ago. It’s not the most accessible of books but it has a wealth of practical tips for empathy and designing services.
No 2,001,926 (55%)
Yes: 1,617,989 (45%)
I don’t write much about politics on this blog but I’m making an exception today because it’s great that Scotland has voted to stay part of the United Kingdom. We have enjoyed remarkable political stability as a single country but as separate countries we could have found ourselves in turbulent times. The risks to Scotland centred around financial uncertainties and the balance of the two to three party system would have tilted uncomfortably to the right in the rest of the UK. Much better the devil we know.
Also worthy of note is the peaceful way this debate has played out. I imagine many parts of the world are looking on wishing that their governments would allow a similarly threatening debate to play out without violence.
Hopefully attention can now turn to building the future in a similarly constructive manner.
I just read this on a great post about the paradoxical traits of entrepreneurs:
Creating something from nothing and watching other people use it, love it, and share it is simultaneously the most empowering and humbling feeling. The satisfaction that comes from watching numbers climb as you tweak and improve your product is indescribable. The exhilarating triumph you feel each time you solve an unsolvable problem is your very own version of winning the lottery. And the relief that floods over you when a risky decision pays off mirrors a brush with death and results in similar feelings of invincibility and extreme gratefulness.
In short, the highs are HIGH.
The lows are, of course, low, but for many it’s the highs that make it, or make being an entrepreneur so addictive. If you’re into rapid learning and seeking new experiences there are few options like it.
The article goes on to list paradoxical traits that successful entrepreneurs exhibit – e.g. stubborn yet flexible and both big picture and detail oriented. For anyone thinking of being an entrepreneur it’s a good list to evaluate yourself against.
Commerce is moving mobile and mobile is moving towards apps. M-commerce apps are, therefore, the way forward. We have investments in a couple (Top10 for hotels and Stylect for shoes) and I expect we will do more going forward. We are taking a serious look at another one right now.
However, maintaining user engagement can be challenging in m-commerce. Becoming part of their users’ daily routine, one of their ‘home screen apps’, if you will, is difficult because the customers of most m-commerce apps don’t purchase that often. Top10 and Stylect have lots of engaged users, but nobody is booking hotels or buying shoes every day.
Notifications are one of the most obvious tools for improving engagement, but they need to be used carefully. New research from Kahuna published on Andrew Chen’s blog found that only 12% of users engage with “ecommerce and retail” notifications, implying that up to 88% of users find them annoying. “Utility and financial services” notifications top the chart with 40% engagement, perhaps reflecting their daily usage patterns.
The trick for m-commerce apps is to use notifications sparingly and doing the work to make sure they are timely and relevant to the user. Unfortunately that means work – there’s no quick and easy way to use notifications to drive a sustained improvement in engagement.
The Kahuna research makes three recommendations:
Find the appropriate cadence: Users may not want to hear about shoes every day, but once a week might be the sweet spot. Look at the stats and see what works.
Make it personal: Don’t send the same notifications to every user. It’s much better to send users personalised sale notifications for shoes on their wish lists than to send everybody a 10% off coupon for a retailer who wants to do a blanket promotion.
Get the timing right: users engage more with notifications at times when they use the app, so figure out when that is and send notifications then. Urgent notifications are different. If a pair of shoes from my wishlist has gone on sale then I should know about it now, before they sell out. Don’t make the mistake of waking people up though (see below).
The best practice book for m-commerce is still being written. A small number of companies are making it work, e.g. Amazon and Uber, but e-tailers with less frequent purchase patterns are still figuring it out. I’m excited for Forward Partners to watch and help them do that.
Below this are some examples of well and badly executed notifications along with customer responses on social media. They are from Andrew Chen’s post. If you are involved with m-commerce you should go read it…
Social recruitment business Branchout raised $49m and grew to 33m users. Today Techrunch is reporting that they are in a firesale situation.
I think there are two lessons here:
- Growth without engagement is fragile. Branchout did an amazing job of attracting users to the service, but they used spammy tactics and too few of those users became active. Rapid growth without a user story that resonates widely spells danger for investors, particularly when acquisition methods are on the edge.
- Platform dependency is a killer. Branchout collapsed when Facebook banned the Wall Post method they were using and many other companies which depended on a single platform have suffered the same fate (Google’s algorithm changes have claimed many casualties). Caveat: dependency can be ok when there is a clear alignment of interests between the platform and the startup. In this case there was no alignment. Branchout was hurting the Facebook experience rather than helping it.
Of these two lessons avoiding platform dependency is the easier for investors to observe. Growth’s intoxicating promise is that the business can be MUCH MUCH bigger tomorrow and that makes it easy to suspend disbelief when areas of the business aren’t working (yet…). Fortune generally favours the brave. In venture capital that usually means those that make risky investments, but in cases like this it means having the courage to stand aside.
Dag Kittlaus, founder of Siri, wrote on Techcrunch yesterday to predict that A Cambrian explosion in AI is coming. He notes that there has been “massive uptake of assistant services spurred by Apple’s Siri, Google’s Now and Microsoft’s Cortana” but says these services are still in their infancy. The missing piece is an ecosystem of services that work with these assistants. That would enable the Holy Grail of an assistant that intelligently finds and uses whatever apps or services we need for a given task – just like in the movie Her from earlier this year.
Need a babysitter for tomorrow night in a new location? The assistant should know the age of your kids, find a service that matches your price profile and then make a reservation, maybe asking you to confirm first.
I have just been looking for a babysitting service in a new location, and it wasn’t easy, so that that sounds very exciting, and I very much look forward to it becoming a reality.
However, it’s not clear to me how the discovery process will work. Healthy ecosystems have some way for quality to float to the top. Examples include ranking or voting systems as we see in Product Hunt and Stack Overflow, algorithms that incorporate user signals e.g. Page Rank, and manual system that takes user feedback as a core input e.g. the app stores.
The whole point of assistant services is that they choose services for us. In the babysitting example above I don’t want the assistant to come back with three options and make me choose, I want one option that I’m happy with. Before I trust the assistant I might want to hear about second and third options to make sure I’m getting the best, but I imagine I would stop bothering with that pretty quickly.
However, if assistant systems make the choice for us then user signals are limited to feedback given on the service. New services, by definition have little if any feedback, making it unlikely that assistants will recommend them and that innovation would suffer. That’s one nightmare scenario. The other is that the assistants only recommend services which have a relationship with whoever wrote the assistant – i.e. Siri only recommends services that have built a relationship with Apple.
In my view app store owners already have too much influence over which apps we use, and that runs counter to the original promise of an open internet. My fear is that assistants, wonderful though they will be, will worsen this problem.