How To Have Growth AND Profits? –part2

Transactional-based startups allow investors to forecast both growth AND profits (check out the part 1) more easily than non-transactional models, where the value created is not captured by direct monetization, such as in social network, IP-driven companies, ad-tech network, certain marketplaces, etc.

Dare to Predict Revenue

These non-transactional models are especially hard to predict.

Just consider the following assertion of David Heinemeier Hansson, creator of Ruby on Rails and partner at 37signals (Highrise, Basecamp, etc.). He said in 2010:

Facebook has been around for seven years. It has 500 million users. If you can’t figure out how to make money off half a billion people in seven years, I’m going to go out on a limb and say you’re unlikely to ever do.

These models are especially hard to predict because their monetization not only often comes by third parties (which one? around which business model?), but they will come in the future, in a different environment (who will be the players? the devices? the usages?)
For instance, nowadays, most of Facebook’s revenues are coming from the mobile, something David Heinemeier Hansson couldn’t have predicted, since the first iPhone was released on June 29, 2007.

Find The Right Proxies for Growth And Profits

When monetization can’t be part of the equation, we need to find “proxies”, indicators that are not in themselves directly relevant, but they are useful in place of an unobservable or immeasurable variable. Regarding value for non-transactional startups, our proxies are metrics that accounts for users’ engagement (eg: visits, usage behaviors, referrals, subscriptions, etc.).

To give you three examples for social networks such as Twitter, you can measure DAUs (counted as the average number of days active per MAU), number of timeline views per MAU or link sharings.

Then you make your quick ratio by just replacing the money by what you measure. Reminder: quick ratio = (new + expansion + resurrected) / (cancelled + contraction))

For example, if you measure link sharings:

new = you check new people sharing link
expansion = people sharing more links on average
resurrected = people sharing links that stopped sharing the previous month
cancelled = people that stopped sharing link
contraction = people sharing less links

Finding Proxies For Engagement

Then, you can use exactly the same method that we saw for transactional models, using cohorts and engagement distribution. Let’s just take some times considering engagement distribution (this could be done with revenue as well).

The best way is to use Cumulative Distribution Function, distinguishing the CDF of users and the CDF of the engagement you are measuring (could be revenue, # links shared or as below, days of use).

The interpretations of CDF charts are closely linked with the type of business you are considering.

Two example using the same chart above. The chart shows the days active in the month which was dubbed L28 at Facebook. For instance, a user of L28=10 was active 10 days of the last 28. If you sum up L28 across all users in a month (as measured on the last day of the month) you get the total sum of DAU across the month. You could compare the total sum of L28 in one month and compare it to the next month and do growth accounting on it. (source of the chart and methodology)

Eg1: an mobile application that provides informations upgraded only once a week (eg IMDb new movies).
You consider to have hit product-market fit when your MAUs open it at least 4 days a month, which means that you are then looking for L28 >= 4.
In the chart, L28>=4 represents above 25% of your MAUs (check the blue line: 100–75 = 25).
25% of users being truly committed is pretty good.
Note that these 25% of MAUs are contributing to 72% of DAUs (check the green line: 100–32).

Eg2: an application to send email for business, which should be done on a daily basis during work time.
You consider to have hit a product-market fit when your MAUs opens it 5 days a week, which means that you are then looking for L28 >= 20. The chart tells you that only 8% (blue line, 100–92) of your MAUs are engaged enough to be considered as truly active. This is not good…
Note that these 8% of MAUs are contributing to 40% of DAUs (check the green line: 100–60).

[A quick note: if this chart would measure revenue (in blue CDF of users and in green CDF of revenue), it would say that 8% of your user represent 40% of your revenue. Business with low scalability, such as service providers charging man-hour, could chose to focus on this segment to keep almost half their revenue for potentially 10x times less work.
In a scalable environment, such as in startups, when you don’t want to reduce your client base, this kind of users are also on the average the most resilient ones, who are in love with your product.]

we look at engagement metrics and the behavior of different kind of users overtime

It is not useful to go deeper in measurement and analytic methods. We will do so later on if you ask us to. The main question here was to know how we can estimate and predict the value of a company that doesn’t make money and which will monetize in an unknown future. Our answer: we look at engagement metrics and the behavior of different kind of users overtime.

Create The Value First, Capture The Value After

But in the end, it is always a question of fundamental user needs and the way startups tackle them. Investing in startup is always looking at a product (a certain experience at a certain price) in a market (a specific user need), with a specific team (with specific values, goals and experience). In non-transactional models, the decision will also rely on other variables, such as evolution of usages, market structures, technologies, etc.
The moto should be: “build the value, monetization will come eventually” — either by youself (monetize directly (eg: BlaBlaCar), monetize through ad/lead generation (pinterest, facebook, twitter, etc.)) or by a third party, who will capture the value and make it monetizable (eg: facebook with instagram, which needed to enter the mobile)

And this is hard, which leads us to the next part. These difficulties explain why investors fail often (but not fast). But before digging into these failures, let’s make quick remarks about late stage valuation for private companies.

Want to know more about startup funding? Read our articles:

Part 1 — Startup Funding: Growth Is The Only Way
Part 2 — The Jedi Trick: You Have to Choose Between Growth And Profits
Part 3 — If Growth Was Easy To Forecast, There Would Be Only One VC fund: Nostradamus Partners
Part 4 — You Need to Lose Money, But Some Loss Are A Really Bad Idea
Part 5 — How To Have Growth AND Profits? (Part1: Transactional models)
Part 6 — How To Have Growth AND Profits? (Part2: Non-Transactional Models)
Part 7 — What About Valuation For Late Stage Startups?
Part 8 — Fail Often, Fail Fast. Investors Do Half of That
Part 9 — What daphni will not invest in
Reminder: daphni investment thesis