A quick thread on how @Amplitude_HQ is using @MiroHQ to help customers instrument product analytics.

We use unique sticky colors to describe Events, Actors, Event Properties, and Actor Properties.

A player from the Oakland As swings a bat and gets a strike (1/n)
When helpful, we think in terms of timelines.

This can help visualize how properties change. In this example a music fan on the free plan plays a bunch of songs, and then upgrades their plan type. (2/n)
With workflows, we use a slightly different pattern.

Here, an Admin starts and finishes the configuration workflow, and we indicate all of the intermediary steps ... which often can happen out of order (humans are weird)

Note we include the narrative and interface (3/n)
Your measurement approach should focus on key questions and key decisions. e.g:

"What config options predict fast reconciliation" (maybe error free/fast reconciliation is our North Star)

You start to see how tracking "basic" events can unlock a long-tail of insights (4/n)
Putting it together, we've been creating these "maps" that link the customer narrative, with key events, to key decisions and questions.

With this info, we can target where we should spend more time instrumenting. @nrose calls this "T-shaped instrumentation" (5/n)
Does this look at all interesting? Want to learn more?

If so, fill out this simple Google Form, and we'll email you once we have a date set for a workshop.


Hope you found this useful (6/end)

• • •

Missing some Tweet in this thread? You can try to force a refresh

Keep Current with John Cutler

John Cutler Profile picture

Stay in touch and get notified when new unrolls are available from this author!

Read all threads

This Thread may be Removed Anytime!


Twitter may remove this content at anytime! Save it as PDF for later use!

Try unrolling a thread yourself!

how to unroll video
  1. Follow @ThreadReaderApp to mention us!

  2. From a Twitter thread mention us with a keyword "unroll"
@threadreaderapp unroll

Practice here first or read more on our help page!

More from @johncutlefish

12 Oct
Here’s a basic (but overlooked) aspect of how product teams work ... and that is how they scope missions.

Say you scope your work in missions of 1-3 months ... (1/n)
That gives you enough time, as a team, to truly “start together” ... do research together, everyone get involved, etc ... and still have plenty of time to “do work”.

But ... (2/n)
Say teams typically scope/shape work in the 1-3w range.

Well of course that is going to require upfront research, handoffs, and “landing” the work on a team. 1-3d is simply too short for a meaningful “start together” type approach.

Read 4 tweets
3 Oct
Back story.

Way back I worked in a PPT factory at an investment bank. Our job was to assemble “decks” for investment bankers. The operation ran 24hrs a day with dozens of “operators”.

Young bankers would work CRAZY hours.

Reflecting (1/n)
...though it involved tons of specialized work, (the bankers) work was primarily research, models, content, etc. Their job was to deliver the M&A doc, the roadshow, etc. Our/my job was to assemble.

Some slides too up to an hour (2pt font, 200 logos, 20 footnotes, charts) ..(2/n)
...other slides took 5 minutes. And with some experience you could — as an operator or a manager, which I did both of — estimate how long a deck would take. 120 slides? 60hrs.

Why? How? Repetition. Repetition. Repetition. And low complexity.

Read 11 tweets
27 Sep
Have been inspired by @shreyas linking to older threads. Very cool. My filing system is not great, but here are a couple ...

1/n: A thread of common loops:

2/n ... the wicked downward cycle of debt, distrust, lack of resources, and finger pointing:

3/n ... on the value of sprints (and some common misconceptions):

Read 8 tweets
26 Sep
(1/n) product developer tip:

You have to build a sense for when you are juggling too many dependencies, too much coupling, too many drivers, not enough floats. A self-inflicted wicked problem.

The trouble is that this is hard to teach/learn....
...when you haven’t seen enough things go wrong.

It is natural to be over confident that you can somehow thread the needle, and brute force the optimization problem. If you adjust the levers just right, things will work!

Do this enough times, and fail, and you...
...build the ability to sense the warning signs. The instinct.

But you need to be in the thick of it. Observing — from a distance — a team in the throes, and you are liable to believe that the levers could have been moved just so.

But they can’t. Alternately...
Read 9 tweets
18 Sep
I've been thinking a ton about disagree & commit

...and I realized what was making me uncomfortable.

Most teams -- even with the best intentions -- don't create an environment where...

1/6: people with diverse perspectives are actually in the room
2/6: people have time to actually process the decision, the data, the context, the implications, etc.

3/6: safety exists to challenge the fundamental underpinnings of the decision and/or "problem" and or question
4/6: multiple promising options are explored. The "decision" is basically "should we do X", which is a decision, yes, but it isn't a choice between options.

5/n: the actual commitment is understood! Which ties in to some of the items above.
Read 4 tweets
16 Sep
"How do I get the developers on my team to take more of an interest in outcomes and our customers?" PdM

This is an interesting one. When developers seem disinterested, it is often for a couple reasons.

1/n: They are overwhelmed with an output-based incentive structure...
...so even if they do care, there is a lot of pressure from their department to "just write code" and "finish projects".

2/n: When there is a ton of debt, getting *anything* done can be a challenge, and this occupies 95% of their emotional/mental energy

3/n: They used to care -- maybe at a prior job -- but got told to stay in their lane. They asked a question, got shot down. Their product manager didn't share context and/or engage them in the problem. So they mentally shut down on getting involved. Too many hassles.

Read 5 tweets

Did Thread Reader help you today?

Support us! We are indie developers!

This site is made by just two indie developers on a laptop doing marketing, support and development! Read more about the story.

Become a Premium Member ($3/month or $30/year) and get exclusive features!

Become Premium

Too expensive? Make a small donation by buying us coffee ($5) or help with server cost ($10)

Donate via Paypal Become our Patreon

Thank you for your support!

Follow Us on Twitter!