Product Analytics Newsletter – Edition #6

  • In Strategy Bracadabra, we share a video on experiment-driven products – take me there
  • In Not Fake News, we look at the war between bloated unicorns and why unprofitable companies go public – take me there
  • In Let’s Get Technical, we’re watching a quick video on the future of data warehousing – take me there
  • In Who You Looking At, we look at how to conform to GDPR for newsletter owners – take me there
  • In My Own Product Analytics Journey, we look at my anemic click rate metric – take me there

Experiment-Driven Products
Experiment-Driven Products: Learning From 1M+ Experiments – Presentation by Jon Noronha, Director of Product Management at Optimizely (at the time)What is your North Star metric and what are your biggest assumptions in how to improve that metric? How can you test your assumptions through an hypothesis that will be tested by an experiment? And most importantly, are you noticing side-effects to those experiments that might indicate you’ve chosen the wrong North Star metric?

It might be somewhat of an oldish presentation, but this encapsulates foundational principles as to why and how to have experiments driving your product’s development. But also the challenges of relying on experiments.

A struggle that caught my interest is the following: choosing the right North Start metric (your most important metric) is crucial as your experiments will make you progress towards improving that metric… but sometimes with unfortunate results.

In his exemple, Jon Noronha talks of how at Bing their North Star metric was the number of searches within one session. That lead to shaping the product towards making what users were searching for… more complicated.

Anyways, good intro and valuable lessons to be aware of if you’re new to driving development through experiments.

The War of Bloated Unicorns
As a product manager and probably even a business owner, what’s your definition of business value? Is it to become a fat unicorn or add value to society?In the ride-sharing war, capital is where battles are won and lost. Lyft raised US$4.9B over 19 rounds, while Uber raised US$24.2B over 22 rounds. It’s definitely a war between bloated unicorns.

To step up its game, Lyft decided to go public earlier than Uber. For a company that has lost $911.3 millions last year, having a valuation of US$24B and raising US$2.3B at IPO , that’s not too bad.
Now this things has been crumbling down ever since. It started at US$72 a share on March 29th, finished the day at US$78.29 and is now at a low US$62.09.

So why do unprofitable companies take their companies public? Recode’s take on that question is that…

“Investors are willing to buy in now in order to subsidize and grow a company that could make lots of money later. They believe that the companies’ future profits will eclipse these current losses.”

Was that the right gamble from Lyft? They did raise more money through that IPO, but will they have as much freedom to maneuver in the future?

My own take on this? Who cares. Those are fat unicorn’s problems. I think all of us are most interested in building products that our users (and even society) as whole will find valuable. This war of raising ridiculous capital is what we should beware of.

The Future of Data Warehouses
Data warehousing is not a use case, by Jordan TiganiLeveraging data means different things to different persons. It can mean confirming hunches with data. It can mean investigating hypothesis. It can mean raising up insights that hadn’t be noticed before. Or it can mean going predictive on such things as user churn for example. Whatever it is, at core of a data strategy resides a data warehouse.

In this talk, Jordan Tigani first provides a bit of context on data warehouses and how they’re are an essential component in a company’s data infrastructure. And how also they have a role that is different than from other tools such as a data lake, data streaming, processing, analysis components, etc.

It’s through an history of that role that Jordans walks us through. What were the core functions of data warehouses at first (storage and computer) and how that evolved towards today’s DWs (for example how storage is now independent from compute, the rise of serverless infrastructures, etc.).

Jordan then takes us on what the future of DWs is and their newer use cases. For example, having real-time analytical capabilities, security and trust, data sharing and (of course) predictive analytics capabilities.

To have your own analytical infrastructure most often than not requires to have your own data warehouse. This is a quick intro to learning more about this crucial component.

Newsletters and GDPR
Alright, you have a newsletter (at least I do…) and you want to respect your reader’s privacy, what the hell does that imply anyway?Well let’s take that thing at first degree as this is a newsletter and I do have to conform to GDPR. Why, cause some of you might be European and I am collecting personal data (your email). And since I don’t intend to block Europeans from subscribing, I need to conform.

So if you’re also in a situation where you collect email addresses, we’re in the same boat. Luckily, conforming for this scenario is really not that complicated. They key here is being transparent.

For example, let’s talk about consent. Mailchimp’s take on this

“Verifiable consent requires a written record of when and how someone agreed to let you process their personal data. Consent must also be unambiguous and involve a clear affirmative action. This means clear language and no pre-checked consent boxes.”

It also goes further than that. Such as providing clear means to ask what data has been collected, how to access it, how it’s being used, how to delete it, how to unsubscribe, etc. All in all, making individuals the owners of their data.

All in all, if you’re running a newsletter through Mailchimp, you most likely got a clear consent to collect a user’s email address. But it should always be made available to users that they can not only unsubscribe, but make other types of requests in regards to their data.

On that note, to cancel your subscription to this newsletter, the link is below (please don’t). And do not hesitate to ask me if you’re curious as to what data I have on you, how it’s being used, etc 🙂

The Click Rate Metric
Remember that video from Dan Olsen on Pirate Metrics and Upside Potential of a Metric? Well, that one really helped me find out what my newsletter’s OMTM should be at this point – Click Rate.Why? Cause that’s a clear sign of engagement and that I’m starting to share stuff with you guys that hits a nerve.

It’s not growth that matters for the moment, cause it won’t happen until I have a clear indication that this newsletter’s mission is being fulfilled with my current readers. And that relates to the forces at play. This newsletter is actually competing for your attention with the thousands of other emails you’ve received within the last hour.

So how is my Click Rate right now? Well, like last week, let’s take this humiliation thing a bit further… As can be seen in the image below, my click rate is, well, umm, anemic.

But never fear, I won’t give up on you good people. And that’s why I’m trying something new this week – that summary table with one-liners you saw above. Let’s see if that will bring an uptick.

I’ll update you on the next edition.

And on that, see you in 2 weeks!

Was this newsletter forwarded to you?

Subscribe Here

How can you leverage data to strategically grow your digital product? This newsletter’s mission is to share data analytics’ best practices and new ideas, so you can improve your feedback loop and accelerate the development of your product.