Weeknotes 394

This week I did:

Not very much

I was on leave, and apart from one small, but I feel justified, work thing, I managed to stay away from work. That doesn’t happen often, usually I work when on leave, but it gave me more time for thinking.

I read:

UK charity camp

Francis Bacon’s insightful reflections on the state of digital in the charity sector from Charity Camp shows how charities are struggling to use digital beyond fundraising, must work out how it will fund enduring digital services, and are not open or sharing enough. Sector-wide digital transformation looks like it’s becoming more and more of a challenge. Things are changing faster and the charity sector is falling further behind.

Joy of agility

I started ready Joshua Kerievsky’s Joy of agility. The thing I’m taking away most is that to be able to respond to change quickly takes lots of preparation and even more practice.

Team work is broken

Mural’s team work research says “66% of knowledge workers aren’t very happy with how their team works together.” They don’t have any answers (other than using their product, of course) but it’s a fascinating area organisations should be experimenting in.

Digital Innovation for Student Success: Research & Development Insights 23/24

This presentation about digital innovation for student success from King’s Collage, London talks about revitalising digital strategy by going from dead projects to living products and harnessing AI to enhance student success. It feels like a revealing slice of ‘where we are right now’ in the digital transformation of education.

Social learning systems and communities of practice

I’m a regular reader of Doug’s posts, but this one on social learning systems stood out as super interesting.

Why you need to fail

We should all make more mistake.

And I thought about:

The product problem for different sectors

The product problem facing higher education is clearer than that facing charities.

Charities tackle wicked problems and there isn’t (yet) the competitive impetus for digital transformation. So, not only is digital transformation not a pressing existential issue, it isn’t clear how to tackle the issue.

But in the higher education sector, the product problem is clearer. Every university is racing to figure out how to provide effective digital/online learning, engaging user experiences, etc. Those that get there first and stay ahead will win.

Parts and wholes

Thinking about complicated systems and our tendency to pick out the parts that seem simple and ignore the whole system, my mind wandered to Kanban. David Anderson, writing in 2010, defined five core properties of Kanban:

  1. Visualize the workflow
  2. Limit Work in Progress (WIP)
  3. Manage flow
  4. Make process policies explicit
  5. Improve collaboratively

Of all those, the one that seems most obvious and intuitive to us is ‘limit work in progress’. It makes sense and is immediately actionable.

So, my question is, does limiting work in progress alone create enough benefit for the system or does it take all five properties working together to actually improve the system?

The organisation paradox

People are wonderful, organisations are awful. How can this be?

Talk to someone face-to-face and they go out of their way to be helpful. But put policies and procedures, incentives and measures, in the way, and suddenly dealing with an organisation becomes difficult.

Organisations are designed by those very same people who are fundamentally good, so how does this paradox come about? Why can’t wonderful people create wonderful organisations?

(I should make clear my bias towards the prime directive and believing that people are fundamental good, and that Dalberg-Acton was right when he said “Power tends to corrupt, and absolute power corrupts absolutely.” We have dictators and sociopathic CEO’s because we’ve created the social systems that enables and empowers those people. The fault lies with the system, but the system is designed by people. This is the organisation paradox.)

Weeknotes 393

This week I did:

Opportunity is the mother of innovation

Necessity isn’t the only mother of innovation, opportunity is too (Greenbaum, et al. 2019). Sometimes an opportunity comes around that is worth dropping everything else for. So I spent Friday investigating and spec-ing a new product. This is good in so many ways; it’s great to be able to reprioritise work quickly and jump on new opportunities, there’s short and long-term financial benefit, and although I’m the only one who is interested in this part, there’s a shift towards digital business models. I’m on leave next week, but I’m keen to get this built pretty quickly the week after.


I completed 36 tasks this week. That’s an average of 7.2 a day and sticking below my target of 10 a day.

I completed 47% of my weekly goals, bringing my average goal completion to 44%. I might have done better, had opportunity not come knocking.

I interacted with 24 people 58 times.


I started setting up my second brain spreadsheet for my next role. It’ll include all the good stuff I’ve learned about task tracking, linking to other files and websites, etc. But I’m most interested in how I can set it up to help me document my mental model of how things work and what things interact with what.

Played with Wandy

I saw Wand.ai on a list of LLM’s so signed up to play with it. It allows you to do things like specify websites or datasets for the AI to use, but it isn’t clear how much it uses that in its answers. And it allows you to create a chatbot and drop it into a website pretty easily, which I didn’t finish playing with so can’t say what it’s like (but of course, all the usual safety, privacy, ethical and sustainability considerations apply).

And I read:

Transforming your organisation for AI

You don’t need AI transformation: you need to transform your organisation for AI. I mean, yes! Two years ago, every company was a technology company even if they didn’t know it yet. In two years time, every company will be an AI company even if they don’t know it yet.

What’s interesting about the analogy with electrification that the article uses is how long it took for electricity to change business models. The first generation of factory managers used electricity to light their stream-driven factories but didn’t do anything to fundamentally shift their business model. It took future generations of managers to see the potential of electricity to change how factory machines were powered, that it made smaller machines possible which changed the layout of the factories, and changed the skills people needed. Electricity created wholesale change in manufacturing, but it took decades. Today, we’re those first generation of factory managers applying AI to how we currently work to be a bit more efficient. It’ll be the next generation who create entirely new and as yet unthought of ways of using AI.

This pattern tells us that changing people changes organisation (more on this thought below). So, if you wanted to optimise for change, when hiring you might ask the interview question, “What did you do differently in your current role that you didn’t do in your role before that?”. That might help to identify people who make change happen. For me, the answer would be, “Talk to more people.” In my previous role, I pretty much only interacted with the people on the project team, so maybe twenty. But since I started tracking in my current role, I’ve spoken to 53 people (about a third of all the people in the organisation), and many more from before I started tracking. As an autistic introvert, this isn’t the easiest thing for me, but it’s the most impactful.

Neuroinclusive workplace

Neurodiversity is a feature not a bug. That seems to be the message in this article about creating a neurodiverse workplace. It talks about some of the benefits from having neurodiverse people, including enhanced productivity, better overall management practices and increased innovation. Then, unfortunately, it goes into the usual boring change management stuff. The disappointing narrative is that all the leaders already in organisations with the power to make change aren’t neurodiverse (myth, although obviously a small minority), but if they can be convinced of the benefits of hiring neurodiverse people then they can reap the rewards (how heroic of them).

Acknowledging the feelings

Sam’s weeknotes are some of the bravest you’ll ever read. I can only aspire to this level of openness and honesty. Sam talks about acknowledging her feelings about some difficult changes at work but, for me, the message is simple: leaders, responsible for people’s health and wellbeing, careers and livelihoods, do better.


IF’s catalogue to help teams design trustworthy services that work for people is pretty great. And so is Rob Whiting’s bookmarks for things like error messages and mobile accessibility. One day someone is going to create the ultimate catalogue of these kinds of catalogues (because intermediation).

And I thought about:

IOOAI game

Mapping impact, outcomes, outputs, activities, inputs is one of my favourite techniques for creating shared understanding (second only to user journey mapping). I really like how it reinforces a causal connection from the impact you want to achieve, which outcomes might do that, what outputs will achieve the outcomes, etc. But, without the rigorous, rational thinking it can easily be misused to justify doing what you wanted to do in the first place.

So I started thinking about how a team game might work to develop more rigorous thinking for connecting impact all the way down to inputs. It could be like a card game where the team shuffles and picks an impact card, and eight outcome cards. Then they have to discuss and justify which of the outcomes would lead to the impact. Once they have their outcomes, they shuffle and pick some output cards and have to figure out which ones will achieve the outcomes. And so on down to inputs.

Maybe the last round is with blank cards and the team has to create their own causal connection from an impact they define down to the inputs needed to achieve it.

If any of those card-maker people want to steal this idea, please do. I’ll never get around to doing anything with it.

Understanding the internet-era

What would you teach someone to help them understand the internet-era? Maybe the transition from the mechanical to the information age, feedback loops, agile and lean, kaizen (continuous improvement) and kaikaku (radical change), the information goods problem and the economics of digital goods, Schumpeter and the first mover advantage approach to innovation, network effects and lock-in, algorithms and long tails. What else?

Mechanisms of change

There are only two mechanisms of change in organisations; people and process.

From memory of the strategic HR management module in my MSc, there are only two ways to change people. The first I’ll call, “Who’s involved?”. In HR terms this is hiring and firing, but for a product team it makes us ask who do we need to make this work a success. The second is, “Who knows what?”. This is learning and development in HR, but for a team it covers the competence, character and confidence of those who are involved (team, stakeholders, subject matter experts, etc.).

And from ISO90001, there are three parts to any process; inputs, activities and outputs, or “What do we put in? what do we do with it? what do we get out?”.

So, if you want to change an organisation, changing process inputs is the lowest impact change, and changing people’s knowledge and skills is the highest impact change. So, why then, do we focus more on making process changes? What’s going on there? Does it just seem easier than doing the messy work with people? It’s a genuinely fascinating question.

Weeknotes 392

This week I did:

Tough week

I resigned. As the project management approach to managing products becomes more embedded, it’s clear there’s no place for product management and so no place for me. It’s a shame. It feels like a big step backwards for an organisation that wants to use digital products to create social change. I firmly believe that we could and should be applying good product management practices in making products successful, but I was unable to convince anyone else. I failed. I’m really sad to be leaving, and I’ll miss all the fantastic people I’ve worked with over the past couple of years. Oh well, that’s life.


I did the GitLabs TeamOps course. Doing more online courses is one of my goals for this year, and this was a interesting one to start with. TeamOps was “Developed, practiced, and refined by GitLab, it’s a framework grounded in actionable principles that transform how teams work and relate.” Although the course isn’t very well structured it’s clear the ideas in there have been well thought through and tested with real teams.


I completed 38 tasks over five days, an average of 7.6 a day.

I interacted with 22 people 53 times.

I only managed to achieve 13% of my weekly goals, but to be fair, it wasn’t an ordinary week.

I read/listened/watched:

It’s groundhog day… again


So, I watched Groundhog Day.

Scopus AI

This is interesting. This is how Gen AI becomes ubiquitous. Build a really good tool, and once people are using it, take away the words AI. “Scopus AI is an intuitive and intelligent search tool powered by generative AI (GenAI) that enhances your understanding and enriches your insights with unprecedented speed and clarity. Built in close collaboration with the academic community, it serves as your trusted guide through the vast expanse of human knowledge found on Scopus, the world’s largest multidisciplinary and trusted abstract and citation database.”

Loneliness and suicide mitigation for students using GPT3-enabled chatbots

This research into Intelligent Social Agents shows how it halted their suicidal ideation in 3% of the students in the survey.

I thought about:


I thought about how deductive reasoning applies to planning, because its about going from the general to the specific and inductive reasoning applies to retros because it’s about drawing conclusions by going from the specific to the general, and as it creates a loop for how to think.

Mechanisms for change

I’ve been slightly involved in some work that had the glimmer of potential to create some impactful change but then tailed off into mediocracy. I wonder what makes this happen. I remember listening to a podcast a while ago about social experiments that tried to explain why people in social groups do and don’t act to things like fire alarms and screams for help. It takes one person to be brave and go against the social pressures of the group

Minimum viable change

The MVP for organisational change is a conversation. It allows for testing out ideas in a quick, low-stakes way. It provides a sense of the doubt, the inspiration, the commitment, or lack of.

Weeknotes 391

This week I did:

It’s all about the feedback loops

Feedback loops are one of the most fundamental first principles of digital work. It’s what makes the digital mindset and ways of working so different to traditional linear way of working. This week had lots of good examples of us building in feedback loops so the team can learn. Retro’s, playback, audience validation and data analysis. They still feel a bit inconsistent, but it’s really great to see then getting established.


I did it. I got my weekly average under 10 tasks a day. This week it was 9.6 as I completed 48 tasks over 5 days. The target of under 10 is part of me trying to spend more time on bigger tasks. I know everyone says they want to do that, but it was only possible for me because I have the data.

Completed 60% of my weekly goals, twice as good as last week and most definitely connected with me reducing the number of tasks in favour of working on bigger things.

I read/listen to:

Systems Practice Toolkit

This, from NPC, is awesome. “The problems we face in the social sector are complex. But we often approach them as if change is linear and predictable. We are meeting complexity with simplicity. To be more effective, we need to think and work more systemically. If the goal is to change the system, then systems practice is how we get there.”

Understanding LLMs

Torchbox’s AI resources for non-profits developing an AI strategy (which they all should be) are brilliant. “All nonprofit leaders – whether they want to or not – are going to need to develop a strategy for these changes. There are opportunities for those who take them.”

Now let’s figure out how we mix AI and systems change.

The state of product management

Jason Knight, on the No Nonsense Agile podcast, talks about “The over technicalisation of product management is one of the biggest barriers to product management”, and how product management is broad in scope and responsibility.

Why do Agile Initiatives fail?

Because, really, what problem does agile solve for organisations? And even if you can answer that (and I bet you can’t), is that problem worth solving to the amount of investment in change that it requires? Almost always not.

And I thought about:


You know the dangerous animals of product management? I randomly thought of PIG’s, Prematurely Initiated Guesswork, this week. A PIG is any product work that lacks validation. It’s purely a guess, but it’s going ahead anyway.

Modern task management

Usual task management is cumulative, it assumes that if you do enough of the right things then you’ll achieve the goal. My approach is subtractive, it assumes that if you remove enough blockers (each task is a blocker) then you free up the flow of value. It takes away the need for upfront planning as ‘the obstacle is the way’ and the next task reveals itself as something blocking you from achieving the goal. It feels like a more modern, digital way to approach task management. It’s based in lean and theory of constraints, it avoids upfront planning, it uses data.


Maybe it isn’t that things change that is the problem, it’s how they change. Too much, not enough. Too fast, too slow.

Weeknotes 390

This week I did:

Web page design isn’t easy

We did some interesting work on learning how to create high-performing pages. We don’t know what we need to learn, and we can’t learn everything we need to in one go. So we need to approach it by enabling intelligent failure, uncovering new barriers and challenges, trying things and knowing they might not work. Over time, this knowledge will coalesce into some good practice which we’ll be able to apply more generally, but for now it’s a very uncertain space.


I completed 52 tasks, an average of 10.4 a day.

I had two goals for this week, and achieved 30% of them (60% for one, 0 for the other).

I had 45 interactions with 23 people, the lowest of both since I started tracking. That’s suggestive of the direction my work is heading.

I had a realisation about how the small, individual tasks connect to big objectives. It’s kind of ‘the obstacle is the way’ thinking where I view a task not as a building block that contributes to the goal, but as removing a blocker to the flow of value. The more tasks you complete, the more blockers you remove, the easier value flows.

Graph showing the distribution of tasks across eight projects. The most important project has the most tasks completed.

This graph shows the distribution of tasks across eight projects. The more tasks completed, the more value flows. So the most valuable product should have the most tasks completed, which it does. This is how small tasks connect to big objectives.

2024 goals

Think I’m settling on the things I want to focus on this year. I’ve had three long-standing goals on my roadmap for a few years now, and each year I pick some opportunities to move towards the goals.

Contributing to the digital transformation of the non-profit sector

Continually developing my knowledge, skills and practice

  • Reflective practice: Writing Daynotes and Weeknotes regularly. It isn’t so much the writing that is useful but it creates a space for reflecting on what I’m doing and learning.
  • Formal learning: Microsoft Learn, GitLab Remote Working. I want to do a bigger course but want to make sure I’ll get into the habit of online learning again before I commit.
  • Informal learning: Reading blog posts, articles and books.

Leading an intentional life

I read:

Product Model Concepts

Marty Cagan’s product model concepts describe five shifts across culture, strategy, team, discovery and delivery that contribute to an organisation consistently achieving outcomes. As a body of thinking, it blows all the product management maturity models out of the water, and I think it’ll form the basis of the solidification of the function and discipline over the next few years.

Generative AI Framework for HMG

The government published it’s Generative AI Framework. It’s good to see statements like this: “Like all technology, using generative AI is a means to an end, not an objective in itself. Whether planning your first use of generative AI or a broader transformation programme, you should be clear on the goals you want to achieve…”

Talking about… AI and the charitable sector

Dr Clare Mills, Zoe Amar and Rhodri Davies talk about generative artificial intelligence. Obviously, it’s huge topic but the interesting part for me was about how AI is causing a shift in ways of working, and leading to the stealth introduction of AI in organisations. This pattern is almost regardless of the fact it’s AI, it’s just another piece of tech, but as we try to learn from the Post Office Horizon scandal one of the important lessons is that people without tech knowledge shouldn’t make decisions about tech in an organisation.

21 small thoughts (and counting?) about information architecture

I think this is really interesting idea. A collection of small thoughts about a topic (in this case information architecture). I have a bunch of paper record cards that I use (quite unsuccessfully) to collect ideas together. I wonder what small thoughts I have about charity product management?

I thought about:

Working to deadlines

Parkinson’s Law says that work expands to fill the time allowed for its completion. When a team is trying to work to a deadline, this law means that everyone’s work expands and the work carries on past the deadline. Some seemingly useful rules that might help to achieve fixed timeframe / flexible scope work:

  • Front load the schedule – Get more done sooner rather than spreading work equally across the time available.
  • Don’t estimate the work – Don’t ask how long something will take to do, ask when it’s needed by and then scope the work to complete it by the deadline.
  • Flex scope up rather than down – Start with the smallest, simplest version and add to it later if there’s time. That’s easier and more efficient than descoping.


Thought more about impact/effort mapping, impact/reach mapping and merging the two. I’m thinking that if impact and reach show the value then effort shows the drag factor to achieving that value. So that’ll be three blog posts I haven’t finished about this topic.

Random links

One of the funny things about the internet is how linking creates such information asymmetry. I got traffic to my website from ohpen.atlassian.net. Ohpen is a banking platform, which uses Jira, and which someone shared one of my blog posts. I don’t know which post or whether they found it useful. We put stuff out there but never know who sees it. The internet puts the power in the hands of the user. Obvious when you think about all the business models built on that idea, but interesting when you see it working at even the smallest level.

Weeknotes 389

This week I did:


Among lots of others things I:

  • Welcomed our new director of digital & innovation. I’m really looking forward to seeing how our team shapes up over the next few months.
  • Kicked-off a short, sharp project to learn how to build high-performing campaign landing pages and generate new leads.
  • Chatted about how we create familiar mental models for ambiguous problems, and apply it to a new product we’re building.
  • Did a bit more audience validation work for that new product.


Completed 55 tasks, an average of 11 a day.

Set five goals this week and achieved 20% of them. This takes my total weekly goal performance to 45%.

Using Google Tasks as a personal to do list is working pretty well so I’m trying out using it for work to do’s. I’m still of the opinion that an effective to do list is just about having reminders in one place to reduce cognitive load. The trick is in building the habit of remembering to add things people say, emails, thoughts you have, etc.


Started updating my roadmap to reflect the things I want to focus on this year. Still need more thinking, and I’m tempted to divide each of the columns with the smaller long-standing goals I have. Then it would show things like ‘Do Microsoft Learn courses… because it contributes to… formal learning… because it contributes to… continually developing my knowledge, skills and practice.’

Impact/effort mapping

Started writing a blog post on some thoughts on how to do impact/effort mapping in a way that reduces game-playing of putting the things where you know it means it’ll get them done and make it more adaptable to each team’s specific situation. Might even finish it one day (I also need to finish other posts about why impact/effort mapping is a bad idea because it prioritises work that is convenient for the organisation and product management maturity models).

And I read:

Getting user-centred design work into agile planning increments

I agree, “Working within the team tooling means being able to act as one team”. But it takes knowledge, experience and most of all, discipline, from everyone to use tools appropriately, and that’s more essential than what tool is used. I feel like we don’t talk enough about the discipline to stick with a well-defined process long enough to learn what needs to be improved.

How High-Performing Teams Build Trust

Collaboration, communication, credit and conflict. Four things for improving trust in teams.

The Chartability Workbook

This is awesome. A playbook for making data visualisations accessible.

The Strategic Foresight Book

I haven’t read it yet, but this book on strategic foresight from Ben Holt looks great. “The decisions we make now will impact the future, and that future will hold new challenges and unexpected opportunities. Strategic foresight allows us to engage with uncertainty, explore possibilities, and turn our insights into action today.” I’m really interested in how product managers do more thinking about uncertainty and what future possibilities might look like. But why a pdf?

I thought about:

It’s ok to do things badly

Maybe one of the consequences of Agile, Lean, start-up thinking, continuous everything, etc., is the idea that organisations and teams should always be getting better. Improvements must be made, everywhere, by everyone, for every process and product. Maybe not. Maybe it’s ok to be bad at what we do. Or mediocre, Or just good enough. Maybe we don’t need that kind of constant pressure.

Problem Solving perspectives

After an interesting chat, I thought about how the different functions of an organisation (finance, HR, design, product) are different perspectives on how to solve organisational problems. Perhaps finance sees solutions in terms of resources and infrastructure, maybe HR sees solutions as requiring people and skills, and product sees solutions as being about changing user behaviour. So, the goal of effective management is to get these perspectives to fit together and complement each other in ways that enables the organisation to successfully tackle all the problems it faces.

But it’s far more likely that those different perspectives are actually in conflict because we don’t have the language to explain them or the desire to challenge such deeply held beliefs. I wonder why it seems like more difficult in some organisations than others. And whether it has something to do with hierarchical, command-and-control structured organisations being clearer about the mental models they impose, whereas flatter organisations with more equally-powered leaders not providing that clarity. Or maybe it’s just one of those team dysfunctions at the leadership level regardless of structure.

Weeknotes 388

This week I did:

Where are the users

One of the interesting things I worked on this week was around audience validation and user involvement in a new product. The hope is that we can set up a small group of beta testers from people who are actually interested in the product. If it works, it’ll be the first time we’ve carried the thread of user involvement through from audience validation to MVP and perhaps into continuous improvement work later.

Some other things I did this week:

  • Moved some survey forms from one platform to another. It sounds a bit dull but I actually really enjoyed figuring out the logic to display the right questions depending on previous answers.
  • Chatted about user journeys and whether somethings (like web pages) make sense only as part of a journey, and that to make them make sense they can only be thought of as a journey, not in isolation.


Four day week, so I completed 42 tasks, which is an average of 10.5 a day. As per the usual trend, the start of the week had more tasks and the end of the week had fewer, bigger tasks.

I had four goals this week and completed 83%. I think two things contributed to me being almost twice as successful as usual; the goals were very independent (I didn’t need much input from others) and I’m involved in fewer things (less distraction as fewer people around).

Had 19 interactions with 12 people. Before Christmas it was three and half times the number of interactions and double the number of people.

The task/goal relationship is interesting. The number of tasks is pretty much right on the overall average, but the goal score is the highest I’ve ever achieved. It must be because of being involved in fewer things and having more time to do bigger things, which must be because of the week off at Christmas and some people taking this week off. I wonder if creating artificial breaks would have the same effect on goals.

And I read:

5 dysfunctions of the team

I started reading the 5 dysfunctions of the team and am pleasantly surprised. I don’t know what I was expecting but as a fable it’s less about the model and the concept of a team and more about the people and their relationships. I think it might be a book I actually finish.

So you think you work in a team

Michele Sollecito’s post about what makes a team and how the way the work is handled tells you whether you’re in a team or a group of people, is really interesting. Good that it says not all groups have to be teams, that creating a team is only necessary for certain types of work. For linear work, it’s actually not worth the effort of creating a team.

A closer look at cross-functional collaboration

Good cross-functional collaboration has higher levels of decision autonomy and shared responsibility (structural factors) and social interaction, trust, and goal congruence (relational factors). If that’s generally applicable for teams trying to work more collaboratively then within the team they need to ensure they spend time together, trust each other, and share the same goal. And outside the team in their working context, they need to ensure they are given the authority to make decisions and given responsibility for what they’re working on.

The WebAIM Million

The 2023 report on the accessibility of the top 1,000,000 home pages, highlights:

  • Across the one million home pages, 49,991,225 distinct accessibility errors were detected—an average of 50.0 errors per page.
  • 96.3% of home pages had detected WCAG 2 failures.
  • 83.6% had low contrast and 58.2% had missing alternative text for images.

Another task trackerer

Sarah tracks her tasks. You should too.

And I thought about:

Wicked and weak

I read Jeff Gothelf’s post about product management being about navigating uncertainty again. And I thought that the uncertainty comes from operating in Hogarth’s (2001) idea of wicked learning environments where some information is hidden and feedback is often delayed, infrequent, non-existent or inaccurate. This means that when a product manager does something, from entering a new market to launching a new feature, there is no way to know quickly and with certainty if it’s been successful. So, product managers have to look for information about emerging developments with likely future impact which become stronger over time as more information becomes known (Ansoff, 1979; Mintzberg & Waters, 1982; Molitor, 1977).

Chicken ‘n’ egg

I thought about chicken ‘n’ egg scenarios a bit this week. There are a few situations I can see that could be made much better if only I could kick-start a different way of approaching it, but there just isn’t a way to get things going.

Three word definitions

I like three word definitions of concepts people argue about:

  • Agile – Uncovering better ways. ‘Uncovering’ because what we’re looking for isn’t obvious, ‘better’ because it’s about improvement, and ‘ways’ because it is a practice that must be followed repeatedly. This phrase is also in the first line of the agile manifesto.
  • Innovation – Creating new value. ‘Creating’ because it’s a process of making something come into being, ‘new’ because what was created didn’t exist before, and ‘value’ because it must be worth something to someone.
  • Psychology safety – Comfortable being uncomfortable. ‘Comfortable’ because that’s how you know you’re getting it right, ‘being’ because it’s very real feelings, and ‘uncomfortable’ because that’s how to know you’re in the right space. (This one is new and needs some refinement.)

I should think of some more.


Now that ChatGPT can answer any question in natural language, I would what happens to all the old chatbots with predefined steps and buttons to navigate the journey. Are they obsolete, or might they find a new niche?

Weeknotes 387

This week I did:

Riskiest assumption testing

I spent some time defining an MVP to help us validate our two riskiest assumptions; can we get the right audience to the product (acquisition), and can we get people to the outcome they are looking for (result, the other ‘R’ I added to the pirate metrics). The first is fairly easy to validate for one of our two potential audiences, so that’s something I’ll work on next week.

100 days of task tracking

Analysed and wrote up what I’ve learned from tracking my tasks for 100 days. I think the most useful things I learned was how ineffective goal-setting is, how the time available affects achieving goals, and that having the data about what I actually did is better than not having it.


I started writing about some of the things I’ve learned this year but realised that it might be taken the wrong way if someone didn’t have an open working, reflective approach. So, instead I think I’ll look back over my weeknotes and summarise some of the things I did, read and thought about.

I read/watched/listened to:

System dynamics

This lecture by Donella Meadows, the godmother of systems thinking, is amazing.

The year of AI

Apart from the other seventy six years of work that went into the field of artificial intelligence, 2023 was the year that one, very narrow, type of AI that got a lot of attention.

The essence of product management

Probably the best product management podcast of the year, this episode of Lenny’s podcast with Christian Idiodi gives a really clear explanation of what product management is and how product managers figure out how to validate solutions.


This list of canvases is great. It makes me think of a product to help people pick the right visual working tool depending on what they are trying to achieve.


Still reading Better Value Sooner Safer Happier. I’ve moved from reading slowly and thinking about it to reading quickly and skipping over parts. It is quite possibly the worst written book with the most impactful ideas.

I thought about:

Product management maturity models

None of the product management maturity models seem to be based on research and, given the websites they are on, are just lead gen content, which is a shame. So, if “all maps are wrong but some are useful”, then “all maturity models are wrong and most are misleading”. The question then is, could a maturity model for product management be useful? I tend to think not, that an improvement kata approach that allows course correction depending on the context would be better, but I’m interested to see if I can prove myself wrong.


I’ve done better this week at using my daynotes to record what I’m thinking about. Mostly it’s been about how to explain the effects of high levels of work in progress.

Weeknotes 386

This week I did:

Doubling down

Had a couple conversations around the theme of knowing when to try something new or double down on things you think are working. The phrase originates from blackjack, so it kinda fits the idea of placing bets even when things feel pretty certain. When we talk about placing bets we mean deciding how much to invest based on the risk we perceive and our confidence in dealing with the risk. That decision making applies whether the thing is new or proven. Proven things should have fewer risks and greater confidence but they should still be thought of as bets

Finishing off

As the end of the year approaches I’ve been rushing to finish off the things that have been hanging around.

  • Migrated forms from one platform to another.
  • Planned out the work from our SEO/SGE strategy ready for the team to pick up next year.
  • Wrote up some thoughts on how we manage campaigns on the website. I’d really like to turn this into a playbook. And then turn that into an entire wiki for how we manage products.
  • User journey mapping for a new product. I love how it expresses the underlying logic of the product, assumptions about user behaviour, and makes explicit the questions we need to answer.


Consciously tried to do fewer things this week so I could focus on finishing off things, which means I completed 33 tasks, averaging 6.6 a day. That’s 55% fewer tasks than last week.

Achieved 35% of my weekly goals. The pattern I’m starting to notice is that one task gets completed, a couple get a bit done on them, and any others get nothing.

I interacted 68 times with 25 people (which is about 1/6th of the organisation).

I’ve been tracking this stuff for a over a 100 days now, so it seems like a good time to write a blog post about it.

I read:

Why autonomous product teams work better

This article is a few years old now but the point still stands. Collective intelligence is better than individual intelligence. Teams are smarter than people. Autonomous teams given the space to solve problems are best.

The 8 Key Challenges Facing Charities

The digital transformation of charities and the charity sector requires wholesale change to business models and mental paradigms. That comes with a lot of challenges. Michael’s blog post goes into 8 of those key challenges.

Move fast and fix things

I read a bit more of Move fast and fix things. I’m feeling a bit torn about carrying on reading it. On one hand I completely agree with the sentiment, but the way the book is written, as five easy steps, annoys me.

And I thought about:

Product management maturity models

I started looking at product management maturity models. They are all broadly the same, five steps and five or more dimensions across the steps. They define the least and most mature versions of how product management exists in an organisation. The question is, are they useful? What can they tell us about the state of product management in an org and what to do about it?

Hofstede’s Cultural Dimensions

Reflecting on how different team cultures interaction with one another, I remembered Hofstede’s Cultural Dimensions. He identified four value dimensions: Individualist/Collectivist, Power Distance, Uncertainty Avoidance, and Masculinity/Femininity (1967-1973). And later additional research identified a fifth dimension: Long Term/Short Term orientation (Bond, 1991). I wonder if teams were mapped by those dimensions we might see what causes conflict between them. For example, if one is more individualistic, avoids uncertainty and is focus on the short term, but another team they work with is more collectivist, embraces uncertainty and thinks long-term, then it makes sense there would be conflict.

Admitting defeat

Thought about when you should know to give up on something. I’ve been going on about ways of working collaboratively for a couple of months but have failed to convince anyone that it’s even a worthwhile goal, let alone try it. So the end of the year seems like as good a time as any to give up and accept it.

Leadership is about promises

Big and small, explicit and implicit, kept and broken. That’s the sum of leadership.

Weeknotes 384

This week I did:

Lots of stuff


  • Content design for campaign landing pages.
  • Started planning some work to increase organic search traffic.
  • Discussed a partnership agreement to cover intellectual property rights and exclusivity.
  • Post-mortem for a recent issue.
  • Set up a dashboard for reporting on KPI’s.
  • Chatted about team harmony, what gets in the way, and what we can do to make things better.
  • Talked about the difficulties of doing what a role should be vs. what the organisation needs.


Completed 63 tasks, averaging 12.6 a day (over 5 days as I did a couple of things on my day off, over 4 full working days it’s 60 tasks, 15 a day).

This week I hit 1000 completed tasks since I started tracking in August. 21 December will be my hundredth day of tracking so I might write another blog post about it, especially how hard it is to set time-bound goals.


On my day off I went to Longleat Safari Park. We were the last one’s in so pretty much had the place to ourselves. The best bit was watching a caravan of camels running at us with wild-eyed glares and humps wobbling side-to-side. Haven’t laughed that much in a long time.

I read:

Do Personas block Systemic Change?

Yes, they do. Anything that focuses on the individual blocks the real change we need, changing our perception human beings being the apex species and individuals as the fundamental unit of society (of course, when we say individual we almost always mean able-bodied, European male).

Do you really need a repository?

No, you don’t. Managing information in libraries, which need proactive administration to maintain the system, is 20th century thinking. It’s the 21st century. Information should be managed in a web of interconnected but independent locations.

The limits of psychological safety: Nonlinear relationships with performance

This article suggests that “high levels of psychological safety climate can actually harm the performance of routine tasks.”

EU Artificial Intelligence Act

The EU AI Act and A European Strategy for Artificial Intelligence.

And I thought:

A rough set of frameworks for product management

Mental models for product management:

  • Resource-based view of the firm – For figuring out how to create a competitive advantage using the available skills, knowledge, etc., in ways that are valuable (improve efficiency and effectiveness) and rare (not available to competitors), and maintain that advantage by ensuring the resources are non-imitable (not easily implemented by others) and non-substitutable (not able to be replaced by some other non-rare resource).
  • Cynefin – For understanding the space you’re operating in and deciding how to approach the work (mostly agile or lean).
  • Agile – For working in emergent domains that need fast feedback loops for learning and course correcting.
  • Lean – For working in complicated domains that need to identify and remove barriers.
  • Teaming – For thinking about how teams can work well together as part of a learning organisation when they face constant change.
  • Value proposition canvas – For thinking through and communicating what assumptions you have about users and what they’ll get from the product.
  • IOOAI – For agreeing what a product or service seeks to achieve and what it needs to do that.
  • User journey map – For understanding what a user will do when using a product.

Every organisation is an AI organisation, they just don’t know it yet

If your understanding about how people will interact with and get value from your organisation isn’t being challenged and changing quickly, then you don’t yet know you’re an AI organisation.

If your workforce and skill development planning doesn’t include moving to fewer, more highly paid people (Goldin and Katz wrote about the association between introducing technology and increasing employee’s skills and wages in 1998), then you don’t yet know you’re an AI organisation.

If your data strategy doesn’t include massive machine-learning data-sets, then you don’t yet know you’re an AI organisation.