Usability testing isn't enough
Usable, useful and used are three different things - pay attention to usability alone and you’ll miss things that really matter (often more).
What’s your bureaucracy ratio?
Every organisation has people who do things and people who spent a lot of time only talking about doing things. The ratio between the two tells you most of what you need to know about whether change will actually happen here.
Traditional GDS model versus Test and Learn
Government Digital and Spend controls were designed for large programmes and for greenfield service development in low-capability spaces - and so the governance is proportionate for these contexts where risks are higher and delivery is staged. But if you’re taking a lower risk test and learn approach, with assurance embedded and delivery regular, small and incremental - they become a throttle.
Freewheeling
Sometimes there's plenty of motion and no actual movement: the wheels turn but the gears never engage. Momentum and the feeling of momentum are not the same thing, and transformations confuse them constantly.
Where’s our moat on AI?
If a machine can churn out the obvious, what's left for designers? Quite a lot, and most of it sits in judgement rather than generation.
Your risk aversion is riskier
There's no risk-free option, only risks you've decided not to look at. How I size a risk honestly, and why starting small enough to survive being wrong beats trying to avoid it altogether.
Spurious detail, false certainty - we need to get more comfortable with "just enough"
The more detail and certainty you insist on before you start, the lower your chances of reaching the outcome you wanted. I had to learn to set off with just enough of a plan.
But we’re already doing…agile
Daily stand-ups, a Kanban board, Jira, retros, a roadmap, and a lot of complaints. Plenty of teams have every ceremony of agile and none of the results - so what’s missing?
But we’re already doing…multidisciplinary teams
Multidisciplinary team, agile, outcome-focused, user-centred: tick, tick, tick, tick. So why no results? Because those labels are shorthand for conditions you might not actually have in place.
Demanding predictable solutions for uncertain and complex problems
Wanting to know exactly what your money buys is fair enough. But when the problem is genuinely uncertain, the more certainty you demand up front, the worse your odds of actually solving the problem.
Services and content that are useful, usable and used: who wouldn’t want that?
The World Bank published its guidance for years, then checked the analytics and found nearly a third of the PDF reports had never once been opened. Putting something online is not the same as it being used.
The death of design
Lately I've read a lot of articles mourning the death of design and research, apparently eaten alive by product. I'm not buying it.
Services - how to even start?
Before anything clever, the basics: what services do you have, who are they for, what are they meant to achieve, and what do they cost? Surprisingly few organisations can answer all four.
67% of accessibility issues originate in design
This gets quoted all over accessibility circles, and it comes from a single Deque case study that measured something quite different. Worth knowing before you put it in a slide.
If your senior accessibility consultant sits below middle management, your org doesn’t (yet) care about accessibility
Every organisation has to start somewhere. But if your whole accessibility effort rests on one junior person, your org doesn't really care about accessibility yet, whatever it says.
Seeing apples and oranges, not just “fruit” - on build vs buy
We treat all our services as if they're the same thing, and waste a fortune doing it. Some solve tidy, well-defined problems; some really don't. Best not lump them together.
Design backgrounds are overrated
For strategic, systems and service design, a design background is overrated. Insist on it and you shrink your talent pool and your odds of finding people who can actually deliver change.
Digital sucks
Digital sucks, as a word. In one week I heard it used for scanning documents, automating processes, redesigning a business, and buying software. When a word means everything, it means nothing.