User-Focused Gems

Bite-sized thoughts on product strategy, product analytics and UX.

All Posts:

New Podcast!
Why I’m working with DAGsHub
Honing the tool - a comprehensive guide to sharpen your SQL skills
Throwing the baby with the bathwater
Keeping score during Corona
Small increments that delight users
Automating banking in the UK
DataScience @ Gartner Innovation Center


New Podcast

Posted on 17-12-2020.

I’m starting a pocast!
I’ve realized that I’ve been having a ton of intersting conversation with friends about their careers and how they planed and strategized around career-decisions and so - here it is now available both live and offline. Tune in here:

Why I'm working with DAGsHub

Posted on 01-12-2020.

The past two years have taught me two lessons that I can (and should) not ignore:

  1. It is easy for so-called “hard tech companies” to underestimate the importance of marketing.
  2. It is easy and common to underestimate the importance of collaboration in data-science projects (and software engineering, generally).

Regarding the first point - just this past month I had two different and unrelated conversations with mentors about how R&D managers tend to overlook marketing and sales efforts (of course - generally and broadly speaking). Apparently, this is a cross-sector and very wide issue.

Regarding team work on technical projects, and especially data-science: from impact perspective, it is all the more obviuos that working on the right task becomes more important than ever today. We see a very strong trend to outsource big chunks of code-writing, mostly with teams in Eastern Europe. I’m thinking this will only grow even more. Therefore, communication and clarity becomes key, and importance of those will likely increase even more over time.

DAGsHub tackles exactly those two aspects. The marketing challenge is right on the nose - DAGs is looking to grow its community as a northstart KPI; that is the main focus and it is an extremely unique opportunity to help this effort. Regarding collaboration on technical projects - that is exactly the produt. The platform lets teams work better together and DAGs is laser-focused on solving exactly this challenge.

While I did not plan to commit to any other project this year, this is just too relevant, too important and too interesting to say no to.

Honing the tool - a comprehensive guide to sharpen your SQL skills

Posted on 23-09-2020.

HackerNoon kindly posted the game plan I created for folks preparing for product analytics interviews. Check it out here.

Throwing the baby with the bathwater

Posted on 22-09-2020.

The Israeli biggest TV&Internet provider came out with their own Netflix.
It has no search option. Really, no search option. I’m not even kidding.

hot next

This gives a whole new interpretation to Reid Hoffman’s saying: “if you’re not embarrassed by the first version of your product, you’ve launched too late”.
Speed is key, but has to be balanced by user’s expectations. Users don’t care if you’re in beta. Users expect some great experience.

If you’d like to make sure that you’re not throwing the baby with the bathwater when you launch, here are some resources I recommend starting with:

Roman Pichler product strategy post

ProductPlan on feature prioritization

foldingburritos on prioritization frameworks

Keeping score during Corona

Posted on 01-06-2020.

Users care about thier score. That might be obvious but apparently Google’s fit team missed it this time.

Product’s core-followers would care a lot about their score, especially when it comes to personal benchmarks, and of course when it comes to such an intimate and important data like one’s personal progress towards achieving better physical fit.

Alternative metrics instead of # of steps could have been:

  • allow users manual logging (so could be pushups, pull ups etc)
  • alllow users to take videos of themsels (automatically count pushups for example using neural nets)
  • allow users to keep accountable to each other so users log for other users.
  • just to name a few options.

Here’s an example for what a small startup probably would have done better - being agile and keeping up with daily changes is much easier as a tiny company. It is really hard to maintain such culture when a company grows really fast, and here’s one example shown.

Small increments that delight users

Posted on 25-04-2020.

With the immense promise that AI brings, it is easy to forget how small increments could sometimes make a difference. Here I share a delightful moment for me as a user - when Google Keep added autocomplete that works really well for groceries.

google keep autocomplete

A quick drill-down explaining why autocomplete is an important UX component:

Considering that algorithmically the suggestions provided by the autocomplete are good:

  • Users would feel like the porduct “gets them”.
  • Users can type less.Obviously, especially when it comes to forms.
  • Autocomplete guides users - allowing both higher degree of assurance regarding a search query, and also some degree of serendipity.
  • Autocopmlete reduces visual competition allowing users to focus on the search box when designed correctly.

For PMs looking into more details I would recommend the following resources:

  1. Great autocomplete UX keypoints
  2. Google’s design guide including autocomplete chapter


Automating banking in the UK

Posted on 01-12-2019.

I was immensly lucky to work with Lloyds Banking, one of the biggest banks in the UK during 2018-2019. One of the projects I was part of was aiming to automate mortgage application over the blockchain.

During this super interesting project I worked with representatives from all 10 biggest banks in the UK as well as the FCA and the Bank of England - all in the same room for a couple of months to come up with a solution.

I co-authored a whitepaper presenting the solution proposed.


DataScience @ Gartner Innovation Center

Posted on 03-11-2017.

This past summer I was lucky to work for Gartner, at their innovation center in Tel-Aviv, Israel (called “GICI”). I worked as a datascientist on three different projects: one was a BI-style project, one was on a recommendation system, and one focused on ways to leverage big-data.

GICI is a really unique place to work at. Decisions are all backed with data, all RD is driven by insights well-thought and structurally-considered. Above all of that, the place feels more like home than an office.

GICI is very open to employee’s-led initiatives, so I was asked to give a short lecture regarding my focus on Datascience for Social Good. You can find here the lecture (in 3 parts). I’m drawing on famous case-studies without going too much into technical details. Enjoy!


Video: Part 1 ,Part 2, Part 3