Edition #23 – November 18, 2019
Originally sent via Mailchimp
Good morning product owners 👋
Just got back from London to meet colleagues, clients and partners and how enriching the whole experience has been. We are privileged to be of a generation where working with someone at the opposite end of the world is totally possible. But to meet in person, get to know them personally and have face-to-face discussions, that’s still something technology can’t replicate.
This is the most cliché-filled intro to the newsletter, but this past week has been packed with so many genuine connections, and for that I am so grateful for the opportunities this adventure allows.
With that, on with the 23nd edition of the Product Analytics newsletter!
I was reading this article while above the Atlantic, and it might have been because of the altitude, the lack of sleep or both, but I got all excited about it. It might also be because of my background studying political science, but the idea of modelizing individual behaviours and predicting social movements based on events such as natural disasters, is just mind-blowing!
Now imagine if we could do the same for product analytics. Where testing wouldn’t involve A/B testing with real users, but evaluating how a new feature’s version would impact the behaviour of your users based on a model you’ve generated of them.
I suggest you get sleep-deprived a few days, step on a plane, absorb an alcoholic consumption or two and read that one. Thank me later.
Battle of the BI Giants
It’s conference season and it’s fun watching giants battle it out in the BI space. In one corner, Google-owned Looker. And in the other corner, Salesforce-owned Tableau (watching rock-star Marc Benioff during keynote was… ummm… something else). And Microsoft-owned Power BI in the assistance watching the show.
Looker came out with release 7 which focuses on 3 things: building and deploying data experiences (“a variety of ways to harness data — data experiences — to fuel everyday business processes”); reimagined BI experiences out-of-the-box; enterprise-class security, hosting and management features, etc.
On their end, Tableau came out with their 2019.4 release. They are continuing on their path of using machine learning to recommend ways of analysing data, this time with a recommendation engine for vizzes; the improved management of large tables (that seems underwhelming, but dealing with tables in Tableau is painful); support for Webhooks, etc.
And Microsoft couldn’t be left behind so they came out with… a modern ribbon (fancy word to say a menu at the top of the app). Well, it does give the thing a Microsoft vibe for sure (if that’s a good idea).
Here’s a project that’s worth following. Per their own definition, “Meltano is an open source convention over configuration product for the whole data life cycle, all the way from loading data to analyzing it.” What that means is that they are packaging a bunch of open source projects (such as Singer and dbt) to provide an end-to-end experience for users.
I still have questions in regards to that project (who are they targeting exactly, what’s their business model, how much flexibility is there in the components to use?), but I think they’re on to something if the idea is to target tech-inclined analysts who are not data engineers and might not have a team to back them up (I wrote a piece on the democratization of data engineering back in July).
Whatever their aim, this is an herculean project and I like that they’re being so transparent about their progress so far. Whatever the outcome, I think the more players in the modern BI landscape, the better.
Spenser Skates, CEO of Amplitude, talks of the execution gap at the Web Summit.
“There’s a gap between your vision of what it is to be empathic with your users and the reality. And at Amplitude we call this the Execution Gap.”
The user journey is comprised of thousands of touchpoints. Tools such as Amplitude helps product managers understand the whole journey, through user behaviours. How the change in navigation impacts trial conversion; what features provides value over time; etc. It’s those questions that can be answered through “product intelligence”.
The whole idea here is really about once you’ve deployed analytics throughout an organization, how do you creatively act on those insights – how to decrease frictions on your user’s journey and fill lost opportunities.
Another one of those BIG TRENDS report with bold statements such as “you will grow your revenues by 10% vs 5% if you combine creativity with data”. Fine. But beyond those statements, there are a few interesting ideas here for product people.
“Combining the power of human ingenuity and the insights gleaned from data analytics is a good start. But the best marketers are going a step further and integrating this power combo into all functions across the marketing value chain—from brand strategy and consumer insights, to customer experience, product, and pricing to content and creative development, media—even measurement.”
A good follow-up read after watching the presentation above from Spenser Skates.
This is pretty much a follow up to last edition’s top pick which explored collaboration within product teams, as well as the handbook to guide towards the structuring of a data team. Well this post, from the fine folks at GitLab again, describes the 3 levels of maturity in an organization. It kinda gives you a roadmap of the value you can reap from investing in data.
I like the “realistic” spin of this piece, where it’s not all about machine learning and artificial intelligence, but really getting your data in shape and building your analytics muscles first. As stated, “A mature data organization, first and foremost, is a mature analytics organization.”
On the same wave as the story before, this one also focuses on the fundamentals – that there are steps to be taken before even considering introducing machine learning into an organization.
“there are 3 things that I believe it takes for a business to derive any meaningful value from machine learning (or AI, data science or whatever they brand it as) –
- They must have raw data that’s valuable.
- They must have a specific, strategic outcome in mind that is attainable with machine learning.
- They must be willing to invest in making it successful.”
In our last edition, we went over the BI landscape with the help of the “Forrester Wave™: Enterprise BI Platforms” report. This article is probably gonna be even more valuable as it provides some guiding principles to choose the solution that is right for you.
This includes cloud availability and cloud connections; ease of use and interface; automated reporting and notifications; ability to import and export data; speed, performance and responsiveness; modeling layer with version control and development mode; visualization library and support for custom visualizations; and pricing, of course.
It’s interesting how what transpires here is not totally aligned with last week’s report. Although this article does not declare winners and losers, it does spend more time explaining interesting features from BI solutions that were not clear leaders in the other report. It’s probably important to point out that last week’s report was more focused on enterprise solutions, whereas this article from Fivetran clearly targets the startups as they are a B2B SaaS company targeting startups themselves.