A few #fitness hypotheses I want to test with @MyFitnessPal data

I have been using the MyFitnessPal app for the past 14 or so months – using the app everyday to meticulously log every food intake and exercise. It is almost a habit now to do the data entry just before digging in or right after.

The app has been tremendous help in giving me awareness of how easy it is to overeat – there are calories rich food everywhere and the calorie density  (calories per volume) is so high for the food you least expect and most like – like those Specialty’s cookies and Panera sandwiches. Having to log every bite I eat gave me visibility into these foods and eventually led me to reduce and void them. Note that I did not say anything about good vs. bad calories, this app does not help you yet and I will write more on that later.

And the result you have been anxiously waiting for? I did lose somewhere between 15-20lbs since I started using the App. But I am not assigning it causation nor should you. My journey stated six months before I started using MyFitnessPal. The initial motivation, drive and discipline existed before the app. You could say that precondition was the one that led me to search for and use an app like MyFitnessPal. Besides using the app I made several changes to my daily routine, cuisine etc. Having clarified that I still believe this is a service I would gladly pay a monthly subscription price for.

I am not the only entering data of intake, millions of people are doing the same. Everyone entering, almost everyday,

  1. Mix of breakfast, lunch, snacks and dinner foods
  2. Exercises they do
  3. Their weekly weights

So MyFitnessPal is sitting on a treasure trove of data (dare I say #bigdata?) that I believe can be put to some good use in the name of fitness science. I do not mean go data dredging into these to tease out interesting correlations but start with specific informed hypotheses and test them with random sampling. All I need is 300-400 data profile samples to test the following hypotheses

Fitness Hypotheses to Test

  1. When people exceed their calorie limit they had set themselves for the day they exceed by significant margin (20-40%) and definitely not 1-10%
  2. There exists a limited specific set of food categories that constitute significant portion of daily budget
  3. For those who exercise regularly (as shows by their logs) on days they skip are also the days they eat extremely poorly
  4. Eating within the limits is a far better predictor of weight loss than exercise
  5. Among those who lost weight, larger portion of calorie intake consists of homemade food and less of packaged processed food.

That is it for now. Next I will write about a few product enhancements their product management team should consider that will lead them to monetization through a subscription service. I sincerely hope their business model involves adding value to the users and getting paid for it in the form of subscription than selling their data for marketers.

 

 

If you most aggressively value @Uber from its only known revenue model

Today there is news that Uber is doing a whopping $400 million funding at a reported valuation of $12B. Previously I wrote about those who were saying even that is very low and Uber is actually worth $100 billion.

Let us consider the only known source of revenue and its current lines of business – 20% cut from taxi fares. Let us do a very aggressive (overly optimistic) market size and share estimate.

  1. 2013 US market size of taxi cabs is $11B
  2. Assume rest of the world is another $11B – unlikely just by sheer sizes of economies, standard of living and purchasing power parity
  3. Assume a 5 year cumulative annual growth rate (CAGR) of 5% – aggressive because the GDP is not growing as much and income is not growing as much.
  4. Uber’s share of this revenue grows from current 10% to 50%
  5. Uber’s net profit margin grows from 20% to 30%
  6. Say current P/E is 136 (matching $12 billion valuation)
  7. But its 5 year P/E can’t be more than 20 – since all the growth is factored in and share of market is frozen its P/E cannot be any more than historical S&P P/E ratio

That is in five years, if this most optimistic scenario plays out, Uber will be worth $16.85 billion. A nice 40% return on your $400 million investment today.

uber-model

But …

  • is the current P/E is so high, isn’t that already building in all future growth?
  • how likely is to gain and maintain 50% market share – that is every other taxi ride is through Uber

If the valuation is based on potential of the platform, on not yet demonstrated revenue model, then why stop at $100 billion, give going up and make it the size of world’s economy.

A popular darling startup is worth $1 Trillion

This is not against any startups or businesses that are focused on serving the unmet needs of customers with a product that is far superior to alternatives. This is on those armchair analysts and social media cheerleaders who make up things to tell us that those startups are larger than they really are.  The business leaders of these ventures will correctly point out where they stand in reality with respect to blogger or market expectations. See for example Elon Musk (CEO of Tesla),

“The stock price that we have is more than we have any right to deserve,”

For those not actually running the business – breaking new market, building new channels, acquiring customers, investing in  R&D and market development etc. – it is just making up things to grab page views. One such article is on valuation of Uber. It is titled, Uber is going to be the next $100 billion company.

I do not know, it very well could be. But as I have said before, if the methods by which the answer was arrived is wrong we should reject the answer. Let me point out the flaws in the argument for even justifying $10 billion valuation let alone $100 billion.

Input: The article cites sources that Uber had a gross revenue of $1 billion growing to $3 billion this year. And a EBITDA of $400 million.

Method: Then it assigns a EBITDA multiple of 8-15X, stating just that “it is reasonable”.

Output: Then jumps to a valuation of $6 billion (400 times 15). It does not stop there – oh by the way since the E is growing $6 billion to $10 billion is child’s play.

There is no big problem with input. Other leaked reports have stated Uber gross revenue of $20 million a week. At 20% cut Uber’s real revenue is about $200 million a year.

There is a problem with the method. It uses a multiple that has no basis or fails to tell us what the basis is. Using EBITDA multiples to value companies is a fair method except that the multiple applied has to be rooted in some logic. The multiple should be based on

  1. Current market size and market growth
  2. Current firm revenue and past growth
  3. Future growth rate
  4. Market dynamics, shifts, trends, competition
  5. Firm’s ability to consistently execute to capture the said growth given market dynamics

To assign a multiple of 15X based only on past growth (when the product is in its exponential part of growth curve) is at best overly optimistic  and at worst downright wrong.

Even if we accept that the mistake is trivial the silliest of all mistakes is how it jumps from $6 billion to $10 billion. It once again applies the high growth rate of EBITDA to bump up to 10. That is it not only applied the high growth rate to assign 15X, the high end of its 8X-15X estimate, it does it again to get to a number it wants. That is like a EBITDA multiple of 25X.

Since the method by which the article arrived at the answer is wrong we must reject the answer.

Let us not even go to how it jumps from $10 billion to $100 billion. It bases that argument on potential on how the platform could be used to do lot more things than just hailing cabs. What about the uncertainties?

So how would you do this valuation right? You surely won’t attempt that in a blog post. But here is a method with more rigor.

  1. Current taxi market size in US is $11 billion
  2. Past growth rate has been 3.5%
  3. Likely future growth is close, but let us say it is 3.5% to 7% (90% confidence estimate)
  4.  Let us say the total world market size (including US) is $15 to $22 billion (90% confident estimate)
  5. Uber’s current share of US market is about 10% ($1 billion in $11 billion)
  6. With its growth let us say the future share is 20% to 60% (again 90% confident estimate)
  7. Then you find out what is the most likely scenario in next five years.
  8. Finally apply a multiple based on comparable businesses, noting that the market size is capped, and give a range for the valuation

If you want to model the most optimistic number – say Uber captures 100% of world market on taxi cab in 5 years. Let us say the market grows at 5% from its current $22 billion and they still make 20% of fare as their real revenue. That comes to Uber revenue in 5 years to be $5.6 billion. And that is it with no where to go in Taxi revenue. Would that justify $10 billion valuation. Sure if indeed all the optimistic conditions I laid out are true.
But you do not just look at most optimistic outcomes regardless of their likelihoods to make valuation prediction but look at all possible outcomes and their likelihoods to make an estimate under uncertainties.

What is your take?

Keep walking past those creativity studies

walking-signThere is a recent buzz in the echo chamber about a Stanford study on walking and creativity. The derivative blog post based on the said walking study takes several steps further (pun intended) to say that explains the creativity of Steve Jobs and Mark Zuckerberg,

Steve Jobs, the late co-founder of Apple, was known for his walking meetings. Facebook’s Mark Zuckerberg has also been seen holding meetings on foot. And perhaps you’ve paced back and forth on occasion to drum up ideas.

new study by Stanford researchers provides an explanation for this.

If you do a simple google search you will find several such studies – all published in reputed magazines with editorial oversight or in peer reviewed journals telling us about a recipe to increase creativity. Here are some samples,

  1. We are highly creative in showers  (this blog post says based on 3 research reports)
  2. Seeing Apple logo makes you more creative (Journal of Consumer Research)
  3. Self imposed restrictions increased creativity (Cognitive Psychology)
  4. Imagining spatial distance between you and the problem increased creativity (Journal of Experimental Social Psychology)
  5. Dim light and ambient noise increase creativity (my take on these two studies here)

All academically rigorous, all found statistically significant difference in increase of creativity between control and treatment groups. Some like the Stanford study found 60%+ increase in creativity.

So if you combined all these studies

Weighing yourself down with iron boots while holding Apple logo in front of you, with dark shades and ambient noise piped through earphones to take a walk in rain and thinking your problem came from a galaxy far far away, should increase your creativity by 480%.

Seriously!

There is too much of these non-sense out there. While the studies themselves may be more cautious in their reporting the regurgitated media posts get too far ahead.

Most of the studies are not reproducible. Even if they are, the study conditions, problems and the metric they use for creativity do not reflect any of the real world scenarios. And statistical significance does not mean managerial relevance. There are just too many such non-sense studies and definitely their effects are not additive. In fact you and I could design an experiment that could should the effects cancel each other out.

Your best bet is to keep walking past these silly studies.

Hat tip to Shlomo Argamon for the walking in rain suggestion

 

Help! My Customers are Choosing the Wrong Version

Whether you are a low tech business like a hair salon or a super awesome tech startup, whether you are a large established enterprise or a startup disrupting status quo you likely seen this problem (or have not done enough data collection to realize you have this problem):

You offer your customers a choice of three (to five) product versions at different price points allowing your customers to self select. You expect customers pick the version that is right for them. But when you do the revenue analysis you find to your dismay that your customers are overwhelmingly choosing the cheaper (if not the cheapest) product version.

37signals-pricing-page1

You likely are thinking whether your pricing page design is wrong, it lacks specific nudges or you need more A/B split testing. But all that are red-herrings compared to the true underlying situation.

There are two significant ill-effects from a scenario where your customers are choosing the wrong cheaper version. First you miss out on price realization – since your profits are going to be higher with Plus and Premium versions it affects even more. Second you are doing a disservice to your customers because you are not enabling them get better value and grow their business by utilizing Plus or Premium version.

That is correct I said you are doing a disservice to your customers when then don’t choose the right version for their business. When a customer understands the value and selects the right product version both you and your customers win. There is more value created to be shared between you two. On the flip side when your customers do not understand the value they get from your versions and choose based only on price they miss out and you suffer.

To take it to extreme, it is like you offer a third class train car with no roof and second class with roof but your customers didn’t know the value from a second class car. So even though they would have preferred to pay premium to travel comfortably with a roof over their head they chose your Basic third class car because of your failure to communicate value.

When customers do not understand the value messaging of your offerings they end up choosing just on price while they would have happily paid more for better options. No one benefits from this situation.

When designing multiple versions – Basic, Plus and Premium – do not just cobble together a set of levers and levels and hope your customers will understand. If you have not done the value allocation correctly across versions and made that value difference clear to your customers, no amount of pricing page design excellence will help you fix the mess.

 

 

When demand exceeds supply – Uber worries

Last weekend there was an article in Quartz about downside of Uber drivers rating passengers.

I had a rude shock recently when trying to hail an Uber cab. It showed up alright, but only barely. The driver told me he almost didn’t pick me up “because of your low score.”

 Low score? For what?!

The author frets about the inversion of power and how customers have no control over this and other ratings on us by all kinds of entities. The fundamental question here is, why does the rating of the customer matter when they’re handing you over cash? Especially in a transactional vs. relationship driven business like cab rides?

When drivers have the option to not serve a customer it means one very clear thing – demand exceeds supply. The irony is, that was the very reason why Uber and such services seek to address – by better matching demand with supply. And when drivers know the opportunity cost of declining a customer with lower rating is acceptable compared to expected value of serving a better rated customer they will do so.

But this assumes that the drivers somehow rationally compute the negative impact of serving poorly rated customer and incremental benefit from serving a better rated customer. In reality both the cost and the benefits are negligible and even out over large numbers. The drivers simply choose to decline because they know there are enough customers waiting for a ride.

The supply-demand mismatch is the driver for Uber dynamic pricing as well.

If you think about it, the customer and driver ratings should not matter to each other if this were a truly efficient market that Uber wants it to be. Imagine for a second if Uber pricing is done in the same as Adwords bidding.

  1. Customers state their destination and a maximum price they are willing to pay for.
  2. Drivers reply with a price same as bid price or a lower price based on their view of demand – in one variation they only see customer willingness to pay and bid only that and the destination
  3. In another variation the drivers also see customer ratings and can take that into account in the bid price
  4. If customers do not find any takers at their bid price they keep increasing it until their true willingness to pay or when they find takers (whichever is lower).
  5. In summary this becomes a simple two sided market decided purely based on price customers willing to pay and drivers who believe they can make a profit at those prices

All these driver and customer and driver ratings are really distractions to create an efficient market.

While we are there let us also get rid of the tips. At market clearing prices tips make no sense.