Weeknotes 472
This week I did:
One ring
I’ve never seen lord of the rings but I get the meme of ‘one ring to rule them all’ and how it represents the desire to make complicated things seem simple by ‘having everything in one place’. It’s a false sense of simplicity, but I get it. The opposite is ‘right tool for the job’ and abstracting a layer so that the browser is that one place, which I tend to argue for. Other stuff:
- Did another workshop, this time on product strategy, north star metric and OKRs. I really like how those three things fit together.
- Discussed new ways of working, how they align with measuring progress and how we’ll use our tools. It’s interesting to think about how to design a system of work where team rituals and practices, processes and tools work together.
- Started a new piece of work using some of those new practices, processes and tools, and paying attention to how long it takes us to do things. Early indication is that we’re faster because we’re being clearer about things like what problems we’re trying to solve, so our shared understanding coalesces sooner.
- Got excited about the idea of creating a perfect abstract mirror of our product in our work management tool, and then had my hopes and dreams dashed on the rocks of pragmatism.
- Talked more about our emerging north star metric to stress-test it and see what appetite there is for it. Want to know how I found our north star? I listened to lots of people talk about what’s important to them and looked for themes. Ya know, user research.
I read/watched/listened to:
Law of complexity
Following on from last week’s thought on complexity, I watched this video which includes a description of how and why complex things are complex. The law of functional information attempts to explain the naturally occurring process of how systems evolve to become more complex. It counters the idea of entropy, which explains why systems breakdown, but there’s no force acting on systems to simplify them. It just doesn’t happen. Systems get more complex, or they break down, but they never get simpler.
Teresa Torres’ Step-by-Step Guide to AI Product Discovery
Brilliant podcast from the brilliant Teresa Torres about discovery and AI, including the how to interview users to understand their problems and how to structure discovery for AI features.
Agentic AI
I wasted a few minutes of my life reading PwC’s executive playbook on agentic AI. Of course it’s all sales pitch about the possible future not the reality today, but when it gets to the “real world success stories”, they are mostly about machine learning, not GenAI and certainly not agentic AI.
Conflating established, proven technologies with emerging, unproven ones makes for a pretty poor sales pitch in my book. It even has a section laughingly called, ‘avoiding the GenAI hype’.
It also reminded me of something said on the Practical AI, that when most people ask for AI what they really want is workflow automation. It’s one big product management lesson on the difference between what users say they want and what users need.
Ambiguity in behavioural design
This paper explores how ambiguity in motivation, capabilities, and opportunities leads to unexpected outcomes. It talks about what every product person knows; outcomes are unpredictable. We can never be sure we’re going to change user behaviour in ways that get business results until after it’s happened, and we never know how long it’ll be until the behaviour changes away from our business result. It’s the ambiguity which makes it fun.
And I thought:
University products
I’ve been thinking about something Saul Couzens said at UKEduCamp; “Products are the things the university provides, not the things provided to the university by IT teams”. That resonated with me because I’m really against the idea that a piece of tech is ‘the product’. The products are things like admissions, support, funding, complaints & appeals, and libraries. So I started mapping out some of the products universities provide and might turn it into a blog post.
Marginal gains
Small improvements, compounded over time, lead to big results. That’s the promise of marginal gains.
A focus on this kind of optimisation has a big drawback. Over time, the results get less and less from the same amount of effort. Then we have a choice, increase our efforts to get the same return, or accept that the gains run out.
Marginal gains works, until it doesn’t.
So, a product strategy based on achieving marginal gains might look good in the short-term but it’s going to fail over the long-term as it isn’t going to create a lasting competitive advantage.
RANCID
Make your stakeholder engagement RANCID.
- Responsible
- Accountable
- Not involved
- Consulted
- Informed
- Decider