What's the WIP limit for your team right now?

If you don't have one, that might be part of the problem.

Know the business: A free 10-day email course on business thinking – Humane Design Learn the business concepts that show up in every strategic conversation, so you can contribute with confidence rather than defer to others.

Know the business covers operations and flow.

Free 10-day course on business thinking for designers:

www.humanedesign.co/courses/bus...

Why starting less gets more done:

If you reduce work in progress (start fewer things simultaneously) average completion time falls. No extra capacity needed.

Most teams resist this. Starting feels like progress. Finishing is what actually matters.

The council planning example:

The team speeds up the submission process. Applications arrive faster. But the assessment team hasn't grown.

Throughput stays the same. Work in progress increases. Average wait time goes up.

Speeding up the wrong step made things worse.

Here's the intuition:

Imagine a restaurant. 10 tables, each seated for 90 minutes. That's 10 tables occupied at any time, with a turnover every 90 minutes.

Add more diners without more tables? Everyone waits longer.

Same law applies to your service backlog.

Little's Law in plain English:

Average completion time = work in progress divided by throughput rate.

It sounds abstract. It explains why backlogs grow even when teams work harder.

Design the service: A free 14-day email course on service design – Humane Design Learn to see how services work as systems, identify where they fail, and design solutions that address root causes rather than symptoms.

If you liked this thread, I have a free 14-day email-delivered course on service design called 'Design the service'.

Learn to see how services work, identify where they fail, and design solutions that address root causes rather than symptoms.

Sign up here: www.humanedesign.co/courses/ser...

hat assumption are you testing next?

Not "what are you researching": what assumption is the research designed to test?

If you don't know the answer, you might end up with insights but no direction.

How to turn research into decisions (not just insights):

1. Start with a hypothesis: "We believe X is true"
2. Design research to test it
3. Record evidence for and against
4. Change the design or the hypothesis based on what you find

Insight decks get shelved. Decisions get implemented.

Research after launch found that residents used smartphones primarily for social media and messaging, not for form completion. Phone completion required a printer, documents, and 40 minutes. Most gave up.

Smartphone ownership ≠ digital capability.

The council housing digital-first assumption.

One council assumed their residents preferred digital services because surveys showed high smartphone ownership. They built a digital-first housing application service.

Jobs-to-be-done thinking helps

Instead of: "our users want an easier form"
Ask: "what are users trying to get done, and what's getting in the way?"

The job might be "prove I'm eligible so I can receive support". The form is just one obstacle. There may be others (before and after) that matter more

There's a hierarchy of evidence quality in user research:

- Anecdote: one person told us
- Pattern: several people show the same thing
- Proof: it's systematic and measurable

Most teams act on anecdotes. Better teams look for patterns. The best teams build towards proof, and know the difference.

The difference between good service design and expensive guessing is evidence.

Not lots of it. The right kind.

Here's what that looks like in practice. 🧵

See the system: A free 7-day email course on systems thinking – Humane Design Learn to spot patterns, anticipate consequences, and find better places to intervene in the systems around you.

If you liked this thread, I have a free 7-day email-delivered course on systems thinking called 'See the system'.

Learn to spot patterns, anticipate consequences, and find better places to intervene in the systems around you.

Sign up here: www.humanedesign.co/courses/sys...

Where are you locally optimising in your work right now?

What to do instead:

Map the end-to-end flow before you optimise any step.

Ask: does improving this step make the whole system better, or does it just move the problem?

The theory of constraints in plain English:

Every system has one binding constraint. Optimising anything that isn't the constraint doesn't improve the system. It might even hide the constraint.

Three signs you're in a local optimisation trap:

One team's metrics improve while another team's worsen.

You solve the same problem repeatedly in slightly different places.

Changes that looked good locally create complaints you didn't expect.

A GP surgery speeds up appointment booking.

Patients can get seen faster. Good.

But the admin team now spends all day processing appointments. Less time for everything else.

Patient experience improves for one step. The overall service gets worse.

Local optimisation: improving one part of a system without considering the whole.

It often creates new problems elsewhere. Here's how to spot it.

Škoda DuoBell: A bicycle bell that outsmarts even smart headphones - Škoda Storyboard Pedestrians wearing headphones are exposed to an increased risk of accidents. In an effort to reduce collisions with cyclists, Škoda Auto, in collaboration with scientists, introduces an innovative bi...

Škoda and University of Salford researchers found a frequency (aka 'safety gap') that noise-cancelling headphone algorithms can't filter out. The result is DuoBell: a bicycle bell that actually gets through. Nice bit of design thinking.

www.skoda-storyboard.com/en/skoda-wor...

What's your service's moat?

The advantage that would be hard for another team or organisation to replicate in the next two years.

Know the business: A free 10-day email course on business thinking – Humane Design Learn the business concepts that show up in every strategic conversation, so you can contribute with confidence rather than defer to others.

Know the business covers competitive positioning.

Free 10-day course on business thinking for designers:

www.humanedesign.co/courses/bus...

Most services have one of these moats without knowing it.

The ones that know it design deliberately to protect it.

The ones that don't often lose it slowly — through fragmentation, inconsistency, or outsourcing the wrong things.

Examples in public services:

The NHS has a data moat — decades of patient history no private provider can replicate.
HMRC has a mandatory-use moat — you don't choose your tax authority.
GOV.UK built a trust moat — one domain, consistent experience, credibility by design.

Four types of moat worth knowing:

Network effects: value increases as more people join (more users = more useful)
Switching costs: changing is harder than staying
Data advantage: accumulated over time, hard to replicate quickly
Regulatory position: mandatory use, licensing, or statutory function

Why do some services become irreplaceable?

It's not usually the design. It's the moat around it.

Moats aren't just for businesses. They exist in public services too.

Design the service: A free 14-day email course on service design – Humane Design Learn to see how services work as systems, identify where they fail, and design solutions that address root causes rather than symptoms.

If you liked this thread, I have a free 14-day email-delivered course on service design called 'Design the service'.

Learn to see how services work, identify where they fail, and design solutions that address root causes rather than symptoms.

Sign up here: www.humanedesign.co/courses/ser...

Before you design anything, surface the assumptions.

Write the problem statement. Show it to someone outside the team. See if they understand it. See if it still makes sense when you read it back.

What's the assumption your team hasn't tested yet?