Weeknotes 474
I did:
Jumped between delivery work in the now (yeah, ok, it was kinda fun) and big vague possible future stuff quite a lot this week. Also…
- Stopped more work coming our way that we really weren’t the best team to handle. Chalk up another success for being clear about purpose.
- Talked about delivery metrics, the calculations behind them, and how we use them to make improvements.
- Did some opportunity mapping which gives us about a hundred years of work.
- Reflected on how I completely missed that a piece of work had gone astray.
- Registered for an MBA. I’m really looking forward to being a student again.
I read:
Secret product skills
I agree about these skills being essential for a product manager but I like to add a layer of ‘how’ to the mix. I think product managers are at their strongest in using skills like negotiation when they approach it dialectically, bringing together opposing ideas through reasoned argument. It’s the philosophy that underpins being data-driven, evidence-backed, etc.
It changes things from being about defending a position to being about finding the best option. Of course product managers need to have a position on things, but they should be open to changing it based on new information and not be swayed by opinion. And they should be encouraging others to think dialectically too.
Design and AI
Also Design methods in the age of AI and Prompting Is A Design Act: How To Brief, Guide And Iterate With AI
I’ve definitely seen a shift in how AI is being talked about. Seems like people are sharing more real lessons rather than being on the extremes of for or against AI.
Who’s copying who?
While reading about LoRA (Low-rank adaption), I noticed that IBM and Cloudflare have extremely similar explanations. I wonder how that happened. Did they both use GenAI to write content for their website and not check that no one else had done the same? How does that kind of duplication affect search rankings (especially given Google uses GenAI to generate summaries of web pages)? Who owns the copyright? Does any of this matter?
Anyway, this is the original paper that introduced the concept of low-rank adaption of large language models.
This is how to make change happen
Katie Andrews is my hero. A manager who prioritises spending time with people, mentoring and coaching them. Plus extra cool points for dining alone.
I thought:
What product managers care about
Because a product is fundamentally a way to exchange value between an organisation and a user, product managers are concerned with maximising and balancing the value exchanged.
Maximising means making sure users get as much of the value they are looking for as possible, and the organisation gets as much of the value its looking for (these are almost always different things).
Balancing means making sure the value exchange is worth it for users and the organisation so that neither is giving more value than they receive.
To achieve this, product managers need to care about three things:
- Opportunities – potential to exchange unrealised value.
- Outcomes – changing user behaviour in ways that get business results so the value exchange happens.
- ROI – cost and benefits realisation. The economics of the product, how much does it cost to run the product, do those costs scale linearly with the number of users and increased usage, etc., so that it’s worth having the value exchange.
Build-run-own-adopt
Should platform teams and internal product teams be responsible for the adoption of their platform/product?
As always, it depends on what problem you’re trying to solve. If there’s loads of demand for it, if other teams are banging on the platform team’s door asking for it, then there’s no reason for them to responsibility for adoption. If other teams don’t know about their platform/product, don’t see the value, don’t understand the wider benefits to the organisation from teams adopting a common platform, then the platform/product team really does need to have full build-run-own-adopt responsibility.
Alex’s ownership trio is a really interesting idea. It says a team needs knowledge, mandate and responsibility in order to own their platform/product. I guess they’d need the same for an adoption trio.
A value proposition for a product operating model for distance learning
“The student experience of the products is so good that they are a marketable differentiator, helping them choose one university over another.” Or something snappier, but you get the point.
Following on from thinking about what products are in universities, I’d love to see products positioned inside universities as needing to be so good that they affect student’s choices. Wouldn’t that be an ultimate outcome for product teams!
AI-assisted essay writing
The academic debate about whether students should use GenAI in writing essays continues.
The counter-logic test to banning GenAI is that banning any new tech would have included ballpoint pens and laptops, and so students would still be writing essays using quill and parchment. Does that seem reasonable? If not, then there’s no logic to banning the new tech of GenAI.
As a product person the answer is obvious. Build a tool that helps students write essays using GenAI. The tool would take them through good practice for essay structure, suggest text for each part, encourage them to question the assumptions in the text, test their knowledge, etc. It could be the only way to submit an assignment (with the usual alternatives universities love) which would ensure adoption and reduce the use of other GenAI, and it would be auditable by the university so there would be clarity about how much input there was from students. The grading would be a combination of the quality of thinking expressed in the essay and the advanced use of the tool.