Design thinking that survives the board pack

Design thinking fails in enterprises for a boring reason: the workshop ends on Thursday and the governance cycle starts on Monday. The sticky notes were never the problem. The problem is that the method was taught as a customer ritual and then asked to survive a legal memo, a risk committee, a procurement threshold, and a comms embargo it was never introduced to.

The teams that keep the method alive treat risk, procurement, and comms as users. They interview them. They prototype the paper trail. They time-box disagreement. They do not wait for the steering pack to discover that the concept requires a privacy impact assessment nobody scoped.

This is a field note from rooms that included Services Australia, the UK Government Digital Service pattern, ANZ product and risk, Allens partners, and NSW Government service owners. Same physics in every building. Different letterhead.

The board pack is a user journey

A board pack has a job to be done. It must let a non-specialist say yes, no, or not yet without pretending they understood a 40-page discovery. If you are a Head of Design, a CDIO, or a programme director at ANZ, Services Australia, or a NSW cluster, you already know the pack will be read in ten minutes on a phone. Design for that.

GDS in the UK made this boring on purpose. Show the thing. Write it in English. Name the user. Name the constraint. Name what you will stop doing. The service standard is not a vibe. It is a set of questions a sceptical senior can ask. Steal the questions even if you are not a department.

Services Australia (the agency that runs Centrelink, Medicare, and related services) lives inside legislation, identity, and volume that would flatten a startup in a week. The useful design-thinking work there is not a persona workshop. It is problem framing that can survive a privacy, security, and payments conversation on the same page. When that framing is missing, you get another discovery that cannot change the service.

NSW Government has Service NSW as the public face of "make it easier" and a thicket of cluster owners behind it. The pattern that works: a framed problem with an operational owner, a 90-day window, and a kill date. The pattern that fails: a lab mural and a business case that pretends the act will be rewritten by Christmas.

ANZ has run design-led and agile work across Australia and New Zealand for long enough that the method is no longer news. The work that still dies is the work that arrives at a risk forum as a surprise. The work that lives is the work that brought the risk lead a paper prototype of the control, not only the app.

Allens (and the Linklaters alliance) is the private-sector rhyme. Partners will not fund a journey map. They will fund a change to intake, scoping, or knowledge handoff if a partner and an operator sat in the room and the residue fits a matter. Legal is not allergic to design. Legal is allergic to design that cannot survive a file.

If your prototype cannot survive a one-page risk note, it was a sketch, not a bet.

Frame the constraint as part of the brief

Most briefs arrive as a customer pain and a hoped-for feature. Useful briefs arrive as a customer pain and the constraint that usually kills the idea.

Write it down on day one:

• Legislation or policy. What section actually binds you? Services Australia and NHS-style services die when the team designs against a rumour of the act rather than the act.
• Risk appetite. What will the accountable executive sign? ANZ-style forums will move if you name the residual risk in a sentence.
• Procurement. Can you run a concierge or a paper test under an existing panel, or are you accidentally standing up a new vendor?
• Comms and industrial relations. Who has to see this before a customer does? NSW clusters and UK departments lose months here.
• Partner or panel rules. Allens-style matters have client and conflict constraints. Design them in.

A constraint is not a wet blanket. A constraint is the brief becoming honest. Design thinking without constraints is arts and crafts.

Prototype the decision, not just the interface

Clickable is not approvable. The artefact that dies in the board pack is usually a Figma file with no accompanying one-pager that a risk lead, a lawyer, and a CFO can all mark.

Prototype, in the same sprint:

1. The customer artefact. Paper, concierge, or clickable. Thin.
2. The risk note. One page. Residual risk, control, owner, expiry.
3. The kill criteria. What evidence stops us? Date and metric.
4. The owner sentence. Who decides on Monday, by name and role.
5. The comms line. What we will say if this leaks or lands.

GDS-style assessment questions are a stealable scaffold: who is the user, what is the need, what have you learned, what will you do next, what will you stop? ANZ and Services Australia rooms that work already speak this language even when they do not cite GDS.

Allens rooms need a sixth artefact: the matter implication. How does this change scoping, the first client email, or the knowledge handoff? If you cannot say it, you designed a poster.

Time-box disagreement

Steering committees expand to fill the ambiguity you leave them. Give them a decision, a clock, and a minority report slot.

A working pattern we use in Design Thinking T&E:

• 90 minutes, not a half day. The long session is where people perform.
• Two options and a stop, not twelve concepts. Twelve is how you avoid choosing.
• A pre-read of two pages, circulated the day before, not a 60-page appendix.
• A named dissent. If risk disagrees, they write four lines. They do not get a sequel meeting by default.
• A same-day residue note. Decisions, owners, dates. Sent before people reach the lift.

NSW and UK department patterns that stall are usually sequels. The committee asked for more analysis because the choice was not framed. Frame the choice and the sequel dies.

What the board actually needs to see

Boards and executive committees are not anti-design. They are anti-surprise and anti-open-ended spend. Give them:

The problem in one sentence a customer or a frontline officer would recognise.

The constraint in one sentence the General Counsel or the CRO would recognise.

The thinnest test that constraint allows. Concierge at Services Australia. A paper script at an NHS trust. A shadow process at ANZ. A redesigned first-client email at Allens. A 90-day service test in a NSW cluster.

The money and the stop. Thirty days and a figure, or 90 days and a figure. Expiry is a feature.

The learning even if you kill it. Publish internally. That is how you earn the next window.

This is how GDS-style services earned the right to iterate: not by hiding, by making the learning cheaper than the theatre.

A sequence you can run in two days

Day one, morning. Frame the problem and the constraint. Interview the risk lead, the lawyer, the operator, and two customers (or two frontline staff if you are a service agency). Write the brief on one page.

Day one, afternoon. Diverge on tests, not on visions. Every idea must name its thinnest legal version. Kill the ones that require a platform or an act of parliament to learn.

Day two, morning. Build the customer artefact and the decision artefact. Pair a designer with the person who will sign the risk note.

Day two, afternoon. Put the decision-maker in the room. Two options and a stop. Kill criteria on the page. Residue circulated before close.

That is Design Thinking T&E with a residue. Collective Campus runs this shape because the other shape (three days of personas, a mural, a "next steps" slide) does not survive the board pack.

A one-pager

Copy the headings into your next steering paper.

• Problem (customer words)
• Constraint (legal / risk / ops words)
• Thinnest test
• What we will measure in 30 days
• What will make us stop
• Owner in the room
• What we will stop doing if this works
• What we will tell comms

If Services Australia, a UK GDS-assessed team, an ANZ squad, an Allens practice group, or a NSW cluster cannot fill that page, they are not ready to prototype. They are ready to reframe. Reframing is the work. The mural was never the work.

How legal and risk actually want to be interviewed

Do not send them a calendar invite titled "co-creation." Send a 20-minute brief: here is the customer problem in four lines, here is the test we want to run, here are the two ways we think it might break your world, what did we miss? ANZ operational-risk and credit counterparts will mark that. Allens general-counsel-shaped clients will mark that. NSW privacy and UK IG leads will mark that. Services Australia security and payments owners will mark that.

Take notes as if they were customers. Because they are. Their job to be done is "do not be surprised in a committee." Design for that job. A paper prototype of the control (the check, the disclosure, the expiry, the cohort cap) is more useful to them than a clickable customer path they cannot approve.

If they say no, ask what test they would sign this month. Write that test down. That is still design thinking. The method never promised you the first idea.

What dies in the steering pack, line by line

We have watched the same four deaths:

The concept with no stop. The pack asks for more analysis because the only option is scale. Give them a stop and they can say yes to a test.

The concept with no owner. "The programme" is not an owner. A named SES, a named GM, a named partner. If they were not in the workshop, do not put their name on the pack.

The concept that surprises comms. A leak or a FOI or a journalist will arrive. Prototype the sentence. NSW and UK departments already know this. Banks forget until it is too late.

The concept that requires a platform to learn. Steering packs are right to kill those. Learn on paper.

GDS-style service assessments already punish these deaths. Steal the punishment. Put it in your own checklist before you book the committee.

A 30-day aftercare that is part of the method

Design thinking that ends at the workshop is a course. Design thinking that survives the board pack has aftercare:

• Day 1: residue page circulated.
• Day 7: constraint owner marks the risk note.
• Day 14: thinnest test live or explicitly blocked (block is a result).
• Day 30: kill, iterate, or ask for the next envelope. Publish the learning internally.

ANZ squads can hang this on an existing heartbeat. Services Australia and NSW clusters can hang it on a programme board. Allens can hang it on a practice meeting. UK GDS-shaped teams can hang it on an assessment window. If aftercare has no home, you will get a sequel workshop. Sequels are how theatre bills you twice.

A one-page board annex, not a 40-page discovery

Boards and executive committees at ANZ, Services Australia, NSW clusters, and NHS-shaped trusts do not read discovery. They read for a decision they will not be embarrassed by in a month. Design the annex for a phone, ten minutes, a sceptical chair. GDS already made this boring: show the thing, write it in English, name the user, name the constraint, name what you will stop. Steal the boredom.

Use this page and nothing else as the annex. Discovery lives in a folder for people who ask. People who ask are rare.

1. Problem, customer words. One sentence a frontline officer or a customer would recognise. "People call twice because the plan-change SMS does not match the bill." Not "we have an omnichannel friction opportunity."

2. Constraint, risk or legal words. One sentence a General Counsel, a CRO, or an Information Governance lead would recognise. "We cannot change the system of record in this window. We can test a script and an SMS under the existing panel."

3. Thinnest test. Concierge, paper, shadow, letter, checklist. Name the channel and the number of people it will touch. Name the vulnerability flags. Services Australia and NHS rooms die when this line is vague. ANZ rooms die when this line is a platform.

4. Money and stop. A figure and a date. Thirty days or 90 days. What evidence kills us. Who can pull the stop, by name.

5. Owner who sat in the room. SES, GM, partner, service owner. If they were not in the workshop, do not put them on the annex. Allens partners will not fund a journey they did not mark. Neither will a NSW deputy secretary.

6. Comms line. What we will say if this leaks, lands, or is FOI-ed. Prototype the sentence. UK departments and NSW clusters already know why. Banks forget until a journalist calls.

7. What we will stop doing if it works. One process, one report, or one room. Boards can smell additive programmes. Give them a deletion.

A practical: circulate the annex the day of the workshop. Ask the chair to mark yes, no, or not yet in the margin. If they need a sequel to understand the page, you wrote a pack, not an annex. Rewrite it. Reframing is the work.

Procurement as a user

Procurement is treated as a late villain in half the design-thinking programmes we see. That is malpractice. Procurement is a user with a job to be done: buy something defensible without creating a new vendor accident or a probity story. Interview them on day one, the same way you interview the customer. Prototype the paper trail they will have to live with.

In Services Australia, NSW, and UK department rooms, procurement decides whether your "concierge test" is a harmless use of an existing panel or an accidental standing-up of a new supplier. In ANZ-scale and NAB-scale banking, procurement plus operational risk decides whether a shadow process is a test or an unapproved outsourcing. In Allens and Linklaters-shaped firms, panel rules and conflicts are procurement's cousins. Design them in.

Ask procurement four questions in 20 minutes. Do not call it co-creation.

1. Can we run this thinnest test under an existing panel, a current SOW, or an internal team, without a new tender?
2. What spend threshold turns this test into a process we cannot finish in 30 days?
3. What evidence will you need if we later want to scale: a residue page, a kill note, a service owner letter?
4. What would make you stop us this month, in a sentence we can put on the annex?

Write their answers on the brief. A constraint is not a wet blanket. A constraint is the brief becoming honest. Design thinking without procurement is arts and crafts that dies in week nine, which is when the surprise arrives.

NHS-shaped buyers already live inside frameworks and clinical safety. GDS-shaped buyers already live inside service assessments and commercial gates. Steal their habit: make the buying path part of the prototype. A clickable demo that nobody can purchase is still a sketch.

A practical sentence for the workshop invite:

Procurement is a user of this test. Please send the category lead who can tell us whether a 30-day concierge fits an existing panel. We will give them a one-page test licence to mark. We will not ask them to mural.