Monthly Archives

December 2015

Three holiday tips for happiness and productivity

By | Startup general interest | No Comments

There’s been a lot of productivity and happiness in my feeds this morning. I guess it might be because we are heading into the holidays and a lot of people are looking forward to spending time with their families. Certainly I’m thinking about how we can all work smarter at Forward Partners next year.

Here are three tips that struck a chord with me:

  1. Effective prioritisation – focusing time on the most important things is so important. That can’t be overstated. There are lots of hacks to help. The one I use is to be clear on the two or three priorities that I’m working on at any given time and pay attention to progressing each of them every day. That means not filling my schedule with too many meetings and being prepared to let email go to hell from time to time (although that does increase my stress – see 2.). Some people find it helpful to write the 2-3 priorities on a piece of paper every morning. Another hack from Warren Buffet is to list your top 25 priorities, identify the top five and force focus by doing absolutely nothing on items 6-25.
  2. Calm your mind – the goal is to not get irritated by or worrying about things that are outside our control. Again there are lots of techniques. Meditation works for me but I know lots of people find that difficult. I read a tip from Tim Metz this morning which is to focus exclusively on what you can change (emotions, judgments, creativity, attitude, perspective, desires, decisions, determination) and not be sidetracked by the weather, what other people want or say (including on email), circumstances etc.
  3. Choose the right people to work with – StartupLJackson put it like this: “Focus on good people/culture. Above all else, my observation is that when you find good people (high-integrity, smart, hard working, etc) and a compatible culture, you end up happy, even if the company fails. When you ignore suboptimal people fit because you think the product is sexy or you’ll make money, you end up sad.”

None of these are easy, particularly 1 and 2, but then we have to work for many of the best things in life. The magic comes in the reward when you get it right.

Happy Holidays everyone, and thanks for reading!

Early stage startups should have a financial model

By | Startup general interest | 3 Comments

I met recently with an early investor in Deliveroo, a current darling of the UK tech scene, who cited their financial model as one of the things that made her invest. The model she said was well thought through and covered all the important elements of the business without being too detailed. And it didn’t have any mistakes.

That struck a chord with me mostly because what we’ve seen over the last two years of working with idea stage companies is it works best when they have a financial model from a very early stage (we’re sufficiently convinced of this point that we are going to add it to The Path Forward). The other part of the story is that I’m a bit of a modelling geek myself. I love a bit of Excel 🙂

All of this is a bit counter the ‘modelling is something that bad investors make founders do’ meme which is prevalent in some parts of the startup community. I think the difference comes down to what’s in the model and why it’s built. What I’m talking about is model for the next twelve months which charts the path from today to where the company wants to be. Done well a monthly model like this does three important things:

  • Forces the company to define it’s goals in hard metrics
  • Shows how feasible the goals are by breaking the path there into monthly increments
  • Gives a plan which can be used to set targets and monitor progress

The other, bad, type of model that some investors ask for (and I think fewer and fewer these days) is a five year plan with lots of detail in the out years. Having a notion of where you expect to be in five years is a good thing, although it should be done fairly quickly, but trying to work out lots of detail is premature because everything will change.

It’s also true that a lot will change during the twelve months of a one year model, but in this context the model is helpful because you know earlier when you are off track and have more to time adjust. If the business gets off track to the point that it won’t get back on track then it makes sense to update the model. That should only happen every couple of months at most and shouldn’t be a big exercise because unless something has gone dramatically wrong most of the logic will still hold and the changes will mostly be changes to assumptions.

This post explains why it’s important to have a model. In the New Year I will follow up with a post explaining how to do it.

Digitisation, productivity growth and opportunity areas

By | Startup general interest | No Comments

The chart below is taken from a McKinsey Global Institute report into digitisation of the American economy (it’s page 5 of the executive summary). The major conclusion of the report is that large swathes of the economy still haven’t digitised, and they estimate digitisation could add $2.2tn to US GDP by 2025. That’s 13.4% of the $17.4tn GDP recorded in the US last year.

I like this chart for two reasons. Firstly the sectors towards the top with more green because they are highly digitised are showing greater productivity growth than the lesser digitised sectors. Admittedly this is a correlation and doesn’t imply causation, but it strongly suggests to me that startups which force sectors to digitise are doing some good in the world, and so therefore, are their investors.

Secondly it gives a framework for thinking about our investment strategy. We describe our focus as ecommerce, marketplaces and related software which maps best onto the education, retail trade, entertainment and recreation, and personal and professional services sectors below. They are just starting to digitise, which feels like a good place for an early stage VC like Forward Partners to be investing. Looking forward to the next fund then we should be heading lower down the table, and healthcare and hospitality stand out as sectors where large numbers of highly valuable companies might be created next. We are already seeing increasing VC interest in healthcare (including form us), but market timing is a big question for many of the startups in this space.


Screen Shot 2015-12-22 at 13.39.19

‘Must have’ products

By | Venture Capital | One Comment

This weekend I was reading posts on A Founders Notebook post about Must have products and navigated to Clayton Christensen’s super interesting Job to be done framework for evaluating products.

As I’ve written many times before I believe that great product is the key to building a successful these days, but there were 30,000 consumer products launched in 2011 and the failure rate is c95%.

Investors have to find ways of identifying which products are going to be in the 5%.

The first place that many of us look is customer engagement. It’s a great sign if lots of people are using a product regularly. It’s better still if they are telling their friends. These are the reasons investors obsess over cohort data, engagement stats and NPS scores.

If you invest in pre-product companies, as we do at Forward Partners, then there’s no customer data and it gets more difficult. A common refrain is to look for products which offer a ’10x improvement’ – that works for some products, but there are many others which succeed where there’s no obvious 10x metric – gmail is a good example. We think a lot about use cases and many of the companies we work with use customer interviews and prototypes to find the features and benefits which will make their products ‘must have’.

The ‘job to be done’ framework takes the use case and places it in a real world context asking ‘what job is the customer hiring this product or service for?’. The language of ‘hiring a product’ is a little clunky, but as Christensen notes people only buy products and use services because they have a job that needs doing. This real world context is powerful because it forces you to make the analysis from the customer’s perspective and inherently factors in competing alternatives. The risk with thinking only about use cases is that they can sound cool but not work in practice because they miss an important facet of the customer’s context.

In the post I linked to above (and again here) Carmen Nobel describes research that Christensen did into why people buy milkshakes from fast food restaurants. It turns out that 40% of milkshake purchases are made by consumers on the way to work and drunk in the car.

From the customer’s perspective (and that’s what counts) they are buying the milkshake to make their commute more interesting and to stave off 10am hunger pangs. Thick milkshakes are great because they take longer to drink and do the ‘make the commute more interesting’ job for longer. Following the research the fast food chain made their milkshakes thicker.

Another major job that milkshakes are bought for is to treat children. In this case thicker milkshakes are unhelpful because parents have to wait longer for the kids to finish. Following the research the fast food chain made thinner milkshakes just for kids.

When I think through the successful companies in our portfolio it’s clear what job they are helping customers to do. With the less successful ones, not so much. That’s why I’m drawn to this framework. The most obvious application for us is in the analysis of individual investment opportunities, but I’m also thinking that it can be used to evaluate new market sectors that we might focus on, asking the question ‘what job will products in this new sector do for consumers?’

Driverless cars and rules vs principles

By | Startup general interest | One Comment


The folks at the General Motors-Carnegie Mellon Autonomous Driving Collaborative Research Lab are now debating whether they should teach their self driving cars to commit minor infractions from time to time. These minor infractions would mimic human behaviour to help keep the cars out of trouble, for example by nudging out at busy junctions.

Raj Rajkumar, the lab co-director, recently said:

It’s a constant debate inside our group. And we have basically decided to stick to the speed limit. But when you go out and drive the speed limit on the highway, pretty much everybody on the road is just zipping past you.

The rules, as codified in law, say that the speed limit should be obeyed, and that’s what Google, General Motors and other autonomous car companies are coding in their software. Human drivers on the other hand, and to a large extent the law enforcement agencies that control them, operate a more complex system that blends rules and principles. There are speed limits, and clear rules about right of way and so on, but bending the rules in certain well understood cases is expected. Nudging out in front of cars at busy junctions when they have right of way is a good example.

This blend of rules and principles has worked pretty well until now because humans are well equipped to judge when to break the rules and what principles to apply when they do. Also important is that we’ve developed a complex system of minor punishments to guide people to sensible behaviour.

Now we have to code that in software which requires us to be explicit about exactly when it’s acceptable to break the rules.

The complicated piece of that is getting agreement across society where views on these topics vary by geography. I remember when I moved to London that I had to start driving more aggressively to get around town. Aggressive driving is the norm here and if you drive in the way I learned in my home town of Gerrards Cross you surprise other drivers and it ends up being more dangerous. Then when I go back to Gerrards Cross I have to remember adjust my driving style.

Getting national agreement on how and when self driving cars should commit infractions basically requires re-writing the law, or at least agreeing how the law should be re-written. That will be tough to achieve, yet I suspect we won’t see the full potential of self driving cars without it.

More interestingly, this debate on how self driving cars should behave foreshadows a much wider debate about how artificial intelligences should behave. We have this rules vs principles tension all through society, and we currently get by with human judgement and systems of punishment.  Because they are software AIs will require us to be explicit about the trade-offs.



Victoria’s Secret’s success came from a deep understanding of customers

By | Startup general interest, Uncategorized | No Comments

In 1982 Victoria’s Secret was a four store chain headed for bankruptcy. Their strategy of selling lingerie to men wasn’t working. Leslie Wexner bought the company and sales now top $7bn, with analysts seeing $20bn in global potential. In the speciality retail world they have market leading sales per square foot and profits. Victoria’s Secret is now one of two primary brands at L Brands, a $28bn company where Wexner is CEO, Chairman and Founder. He’s now 78.

BCG Perspectives made a thorough analysis of what made Wexner and Victoria’s Secret a success. Interestingly, much of the insight that made the business work is the sort of insight that comes from customer development. Customer development only emerged as a thing in the last 5-10 years, so in one sense they were ahead of their time. In another important sense they weren’t though, because whilst Wexner’s instincts took him to the right place he relied on the magic of his own intuition rather than understanding that appropriately structured customer interviews are a powerful tool for developing and checking gut feel.

The following are quotes from the BC Perspectives post (emphasis mine):

  • Consumers cannot think in abstractions. They cannot envision a new concept. They cannot predict their own behavior.
  • A curious mind asks the questions that open up the consumer to talk about her latent dissatisfactions, hopes, wishes, and dreams. A curious mind knows that functional goods sold en masse earn a good return, but breakthrough profits come from satisfying emotional needs. A curious mind does not jump to conclusions but tests carefully and thoroughly.
  • You need to get into the heads of consumers and be able to tell their stories. It is both art and science.
  • Winning solutions respond to the distinct and specific needs of a group of consumers.
  • A curious mind does not jump to conclusions but tests carefully and thoroughly.
  • Completely understand your customers by seeking an intense, complete, and empathetic view of their lives and the context for their purchases.
  • Deliver infinite growth by having your customers talk about you, exclaim about you, and tell their friends and colleagues about you.

This chimes with a lot of our thinking, as captured on The Path Forward. As mentioned above, the best way to glean these insights is through structured customer interviews. My partner Dharmesh wrote a guide to doing great customer interviews here.

Lessons and stories from Waterstones’ resurgence

By | Startup general interest | 2 Comments

I was surprised to read this morning that Waterstones, the UK’s largest chain of bookstores, is undergoing a resurgence. At a time when all their competitors and brethren around the world are going bankrupt they have posted a profit for the first time since the financial crisis and are opening new stores.

There’s lots of things I love about this story.

  1. I once thought it inevitable that ebooks would wipe out paper books. I thought the reading experience would become equivalent and the convenience of ebooks and networked comments would win. As things have panned out, what I thought of as inevitable has not come to be – people do still like to read books and this article about Waterstones’ recent successes have made me pause and take note. It’s important for investors to have a clear and confident view of the future, but also to recognise when mistakes are made. As Marc Andreessen says we should have strong views weakly held. (As a side note, I have abandoned my Kindle and Kindle app to go back to books, I can’t put my finger on exactly why, but think it’s partly because books are easier to read, partly because it’s easier to flick through the pages of a book than an ebook, and partly because I can give books to people afterwards.)
  2. The Waterstones turnaround is a great story. In 2011 they were on the cusp of bankruptcy when Russia’s ‘least famous oligarch’ Alexander Mamut rescued the business. He brought James Daunt, founder of Daunt Books, a small chain of idiosyncratic books stores, in as CEO who’s booksellers perspective has made the difference. Contrast his background with Ron Boire, CEO of struggling Barnes & Noble who is a generalist retailer with Sears and Toys R Us on his CV. Daunt took the job because he loves books and knew that if Waterstones went bust many of his favourite mid-sized publishers would lose 60% of their sales and also go under.
  3. Daunt’s turnaround plan was very brave. When he started publishers were paying Waterstones £30m per year for prominent placement of their books. He threw that revenue away, focused on stocking titles people wanted and gave local store managers complete control over where they put their books. After these changes the percentage of books returned to publishers unsold fell from 20% to 4%.
  4. This is another example of the power of decentralising control – see 3.
  5. We have another reminder that pricing is often counter-intuitive. Daunt changed the discount policy from global to locally set noting that when it comes to books many people are price insensitive. The purchase price is not that significant compared to the value of the time they will spend reading the book.


Mismeasurement may be at the heart of the productivity paradox

By | Startup general interest | One Comment


My brother Rob is an economist by background and we have an ongoing debate that is repeated between techno-optimists and economists around the world. Every day I see improvements in technology helping us to live fuller lives and do more of the things we want to do but Rob and I struggle to reconcile that with the meagre growth and stagnant productivity statistics that he looks at.

Getting more specific, I believe that advances in robotics and artificial intelligence will most likely displace a significant percentage of human labour over the next 20-30 years (the Bank of England is saying up to 50%), but if robots are replacing humans in the workforce the productivity of the remaining humans should be rising, and that isn’t happening.

This is the productivity paradox.

One explanation for the productivity paradox is that economic output and productivity statistics are broken. Many people on the techno-optimist side, myself included, have pointed to the rise of free services from companies like Google and Facebook as an example of valuable goods that aren’t caught in the stats. That explanation has always been a little unsatisfactory though, because to my knowledge the free services dividend hasn’t been quantified and it’s hard to be confident that it will add up to enough to explain what will eventually be a massive discrepancy between labour replacement and productivity stats.

The good news is that I’ve just read another explanation, which is that improvements in quality of products aren’t being caught in the stats. This is a view held by Larry Summers, amongst others. As you can see from the chart above official statistics (i.e. the PPI) overstate the real price of computation by 27x. That’s a massive difference and it’s quantified. If this overstatement of prices is repeated across enough other categories then we could be misunderstanding today’s world due to mismeasurement and we have the basis of a calculation which could solve the productivity paradox.

A product people love is key to scaling

By | Startup general interest | No Comments

Chis McCann of Greylock took the Stanford course on Blitzscaling over the last four months. The course is taught by Eric Schmidt, Reid Hoffman, Marissa Mayer, Brian Chesky, Diane Greene, Jeff Weiner, and more.

Chris wrote up the 16 top lessons here.


I’m going to call out a sometimes counter-intuitive point which Chris mentions in his 5th and 6th lessons.

Brian Chesky said:

Paul Graham gave us a series of advice that changed our trajectory. The most important of this advice was that it was better to have 100 people who loved us vs. 1M people who liked us.

And Reid Hoffman said:

analytics, dashboards, and data management will not matter at all if no one cares about your product. Getting people to care about your product is much more important than metrics [when you are just getting started – 1-10 people]

I’m writing this because whilst every founder wants people to love their product most don’t get there. The natural instinct of many is to build a first version which matches their vision of what customers want and then iterate. The problem with this approach is that the initial vision is rarely right and iterating from a bad starting point often doesn’t get you very far. Technically it will get you to a local maxima. And that’s if you do it well.

Much better to focus on building a product people love. The way to do that is to talk with customers. Find the thing they care about and the feature(s) that make their eyes light up. Then build. We go on about this endlessly at Forward Partners and it’s a key component of The Path Forward.

Thinking fast and slow at a VC fund

By | Startup general interest | 2 Comments

For the second time this week I’ve been thinking about my early career as a VC. This time I was remembering how I would find interesting deals only to have them shot down in seconds by the partners I was working with. Usually because they didn’t fit with some pattern he was working with, or were in an area where the fund had had bad experiences.

For those of you that have read Thinking Fast and Slow, That’s Kahneman’s Type 1, or fast thinking in action. The partner was subconsciously using his (and they were always men…) experience to pattern match and make a quick decision – to go by gut instinct.

It was frustrating for me because my thinking was Kahneman’s Type 2, i.e. of the slow, considered variety. I had spent time thinking about the prospect and had logical reasons for believing we should at least consider investing. The partner’s Type 1 thinking didn’t want to engage in a Type 2 debate though, so my arguments were quickly dismissed.

What I realise now is that the partners were right, or at least mostly right.

Nine times out of ten when you ask a VC how they are doing they say they are busy. That’s because if they have got any presence in the market they are constantly overwhelmed by entrepreneurs looking for investment. I’m fond of saying that any VC who knows what she’s doing should be permanently busy.

It’s this busyness that explains why Type 1 thinking is appropriate when evaluating deals. Good pipeline management is all about quickly working through the prospects that aren’t going to turn into investments and only spending time on the ones that are going to close. That means crude filters and quick decisions at the top of the funnel.

The challenge for people new to analysing startup investment opportunities is that it takes time to build up the experience to be an effective Type 1 thinker, but it takes too long to analyse every opportunity deeply in a Type 2 style. My solution to this challenge was to work really hard and try to learn from investors with more experience. John Taysom and Simon Cook taught me a lot.

The challenge for partners is to build strong instincts and be an effective Type 1 thinker. The most difficult piece of that is to keep improving. Type 1 thinking doesn’t listen to reason and it takes conscious effort to know when to slow down, analyse a question more deeply and make the switch to Type 2 thinking. That switch comes with a cost because it means taking time from something else that now won’t get done, or not done so well. I make a big effort to take notice of opinions contrary to my own from credible sources, it takes time, and much of it is wasted, but it keeps me moving forward – and that’s essential in this fast changing world of ours.

The other challenge for partners is to help their associates develop good instincts. That means understanding the underlying analysis or pattern matching that drives instinctual or Type 1 decision making and then taking the time to explain it.

This is a post about analysing tech startups for investment purposes, but I think the point holds true for many other domains as well. Product management, design, and software sales spring to mind.