# product-management

Ryan Grimard

06/08/2022, 2:00 PM
I'm also looking for any guidance folks have on surfacing the value of Platform Engineering initiatives to the business side of a company to help answer questions like "Why is this tech initiative more important than this business initiative that is going to produce X $ per day?"

Andrew Fong

06/08/2022, 2:30 PM
Personally I don’t think thats the right framing. Its a short vs long term projection. Platform needs to be able to understand the business need and design a financial model for either revenue or COGS reduction. If the company cares about revenue then it has to fit into that story. The problem you’re going to run into is if the company isn’t really a growth stage company but is trying to get into growth (again) then reduction of COGS isn’t super important short term bc its about growth. I’d push leadership right now to explain if you’re chasing efficiencies or growth - macro economic climate seems to indicate caring about efficiencies(?). YMMV though depending the market segment etc
Existential problems mean there’s real trade offs and it could mean that platform isn’t important. Knowing this is important.

Alan Barr

06/08/2022, 2:45 PM
I work on related insurance/home loan/ long lived products in this business. My focus is just the platforms that keep using the latest. There is a humanitec webinar on it if it is interesting

Ryan Grimard

06/08/2022, 3:07 PM
Thanks, Andrew!
Alan, are you saying there IS a webinar?

Ryan Grimard

06/08/2022, 3:22 PM
I'll definitely watch as it's highly relevant to what we're doing, but it doesn't seem to address the question. Thank you for the link!

Schuyler Bishop

06/09/2022, 10:29 AM
Happy to connect if you’re interested. My employer is in fintech and we have a 60 person dev team my PE org serves!
👀 1

David Schweizer

06/09/2022, 5:17 PM
The most compelling arguments are the ones that convincingly supply dollar values for platform initiatives. There’s material on the impact on organizational performance in chapter 2 of Accelerate and in the McKinsey report on Developer Velocity. I learned a long time ago that getting the business side of the house to believe cost reduction claims requires careful attention to the counterfactual in advance.

Cruise Hall

06/23/2022, 1:10 AM
@Ryan Grimard I find Jean-Michel Lemieux’s “3 buckets” mental model helpful - essentially he posits that eng investments generally fall into one of 3 buckets - experiments, features or platforms. He recommends targeting a certain investment ratio among the 3 buckets (10%, 40%, 50%, respectively) to maximize the flow of value and feedback among assets in each of the 3 categories. He offers more commentary in this podcast at `25:41` (+ transcript). ^^ I think an implication of embracing this 3-bucket model - regardless of your target ratio - is that platforms don't go toe-to-toe with features for $$$. Maybe the org’s ratio is way off, but negotiating the ratio gives platforms a fighting chance vs features.
Admittedly, the 3-bucket model doesn't answer the main question directly - but in the podcast Lemieux does hint at an interesting platform selling point:
in the platform I'd say, five [of] the most important features of every product are owned by the platform teams, uptime, scalability, performance, extensibility, and maintainability
^^ even if these “non-functionals” can't be measured, their value can often be appraised (“What’s the per-minute cost of an outage?”, “What would you pay to reduce the runway to production from days to hours?”) In some cases you can even calculate a platform vs feature break-even - e.g. platform investment X will outperform feature investment Y soon as X shaves 50 cumulative minutes off of our time-to-restore (TTR).

David Schweizer

06/23/2022, 3:42 AM
@Cruise Hall Not only can anything be measured*, but “appraisal” as you describe it is a measurement. *That anything can be measured is a literal consequence of the definition of measurement, “[a] set of operations having the object of determining a value of a quantity”, which means that guesstimating based on prior experience qualifies. Importantly, and perhaps surprisingly, anything can be practically and usefully measured. See Doug Hubbard’s book How to Measure Anything.
🙌 2