BCG · Design Studios — Foundation v10.5
Live · 06 May 2026
at Design Studios

Creative AI.

A three-layer architecture — Lab, Practice, All DS — for how Design Studios builds, validates, and scales an AI-augmented craft.

BCG Design Studios BCG Design Studios
Layers 3
Pillars 4
Fluency steps 7
Festival 1min.ai
01 · Context

AI raised the floor. DS raises the ceiling.

Generative AI lifted the baseline of what anyone can produce. The new differentiator isn't access to tools — it's institutional craft. Studios that codify taste, build workflows others can't, and treat AI as material — not magic — will define the next decade.

The three audits we ran

12
Studios audited
A snapshot across regions of where DS sits today on AI integration, fluency, and creative workflow.
37
Practitioners interviewed
From IC designers to design directors — the people doing the work and the people leading it.
9
Pilot workflows
Live experiments that informed the model — what scaled, what didn't, and why.

The job isn't to use AI. The job is to make work that couldn't exist without it.

02 · The Model

Three layers, one circulatory system.

The Lab invents and validates. The Practice translates and operates. All DS receives, applies, and feeds back. Each layer has a different speed, a different tolerance for risk, and a different metric for success — but they share one closed loop.

Three concentric layers: All DS outermost, Practice middle, Lab innermost — with bidirectional flow All DS · Apply Practice · Translate Lab · Invent Creative AI at DS
Lab Invent · validate · package

~20 people. The R&D node. High tolerance for risk. Ships craft into the rest of DS as patterns and playbooks.

Size ~20 designated Cadence Bi-monthly sprints Output Patterns & playbooks KPI Adoption beyond Lab
Practice Translate · operate · spread

Regional Practice — AMR, EMESA, AP+ME. Adapts what the Lab makes for local craft, language, and case context.

Regions AMR · EMESA · AP+ME Cadence Bi-weekly Pillars Explore · Learn · Apply · Share KPI Active workflows
All DS Apply · feed back · raise the floor

Every designer in every studio. Self-paced fluency, monthly briefs, an annual festival. The user base and the feedback engine.

Reach All studios Cadence Self-paced + monthly Ladder 7 fluency steps KPI Fluency progression
03 · Lab

A small node, doing uneven work.

~20 people. Three designations: Members, Affiliates, Friends — concentric circles of involvement, each with a different commitment and reach. Six focus areas define what the Lab actually works on.

01Inner

Members

Core operators. Drive sprints, own patterns and playbooks, set the Lab's quarterly direction. Significant time commitment — a real percentage of capacity, not a side project.

02Mid

Affiliates

Specialists who join sprints by topic. Brought in for their craft — motion, type, code, sound — when a sprint needs that skill. Time-boxed contribution, not standing membership.

03Outer

Friends

External orbit. Critics, collaborators, sounding boards from the wider design and AI world. Show up when relevant, leave when not. Keep the Lab honest and outward-facing.

Six focus areas

The Lab's beat. Each focus area has a lead, a sprint cycle, and a measurable output.

F.01

Generative ideation

Concept exploration at high volume — text-to-image, text-to-3D, mood and direction tools. Where the broadest space gets surveyed quickly.

F.02

Workflow tooling

Pipelines, scripts, custom GPTs, and internal tools that compress repetitive work — letting designers spend more time on the parts that matter.

F.03

Craft amplification

Using AI to push existing craft further — refined typography, cinematic motion, generative systems that feel hand-tuned, not template-output.

F.04

Editorial & narrative

Long-form content, story arcs, scripts, briefs — where AI assists structure but the editorial point of view stays firmly human.

F.05

Real-time & live

Live event visuals, real-time generative content, streaming-era formats. Fast, reactive, and built for moments that can't be pre-rendered.

F.06

AI & the team

Studying how AI changes the way designers work together — pairing, review, handoff, mentorship. The org-design adjacent question.

04 · Practice

Four pillars, one operating rhythm.

The Practice does the connective work. It exists in the regions, runs on a bi-weekly heartbeat, and structures itself around four pillars — each with its own ritual, output, and accountability.

P.01
Explore
— survey the edge

Track what's emerging — tools, techniques, players. Surface signal, ignore noise. Bring the most interesting findings into the rest of the system.

P.02
Learn
— build fluency, fast

Foundations programme, Jam Sessions (Ignite · Make · Share), workshops. Designed to move a person up the Fluency Ladder by one rung at a time.

P.03
Apply
— ship into real work

Move patterns into live cases. The point at which research becomes craft. Tracked by adoption — workflows running, not just decks read.

P.04
Share
— close the loop

Editorial board, monthly brief, the festival. The artefacts that move learning across studios and back into the Lab as feedback.

1min.ai Festival

An annual rhythm. May launches the brief. August judges and ships. Every studio enters. The whole community watches.

May
Brief drops
Annual theme released. Open to every designer in every studio. Submissions open through July.
June
Make
Studios run their own jams. Office hours from the Lab. Drafts shared inside the regional Practice.
July
Submit
Final pieces submitted across categories. Peer review. Editorial board reads everything.
August
Ship & celebrate
Winners announced. Highlights published externally. The artefacts of the year, preserved.
05 · All DS

A ladder, not a certification.

Seven steps from "AI-curious" to "AI-leading." Self-paced. Public. Every designer can see the ladder, see where they are, and see what the next rung needs. No tests, no badges — just visible progression and shared language.

Step 01

Curious

— I've heard the terms.

You've watched the demos, you've read the threads. Maybe you've tried a prompt or two. AI is on your radar but not in your work yet.

Step 02

Trying

— I've used it on something real.

First real attempts inside actual work. Mostly gimmick or filler stage — but something has shipped that wouldn't have without AI in the loop.

Step 03

Working

— It's part of how I deliver.

AI is now in your pipeline, not on top of it. You know which tools fit which stages. Output quality has visibly improved.

Step 04

Crafting

— I get results others can't.

The taste differential shows. You produce work that's recognisably AI-augmented but still distinctly yours — pattern, voice, polish that survives scrutiny.

Step 05

Building

— I make tools, not just outputs.

You've built a workflow, a custom assistant, a script — something other designers can pick up and use. You've shifted from consumer to producer of leverage.

Step 06

Teaching

— I move others up the ladder.

You run sessions, mentor, write up patterns. Your impact is now indirect — measured by how many people you've taken with you.

Step 07

Leading

— I shape what comes next.

You define direction at studio or practice level. Decisions about which tools, which patterns, which work to take on — these run through you.

The disciplines, in their own words

Six discipline areas across DS — each thinking about Creative AI from its own vantage point.

Visual & Brand

Identity systems, visual language, brand expression. AI changes ideation speed and asset breadth — but the editorial point of view is still where the value lives.

Motion

Animation, video, generative motion. The discipline most directly transformed by AI — text-to-video and runway-class tooling have collapsed entire production pipelines.

Editorial

Long-form writing, scripts, narrative structure. AI is great at first drafts and structure — but the voice, the angle, the right omission still come from a human editor.

Product & Web

Product UX, interfaces, websites. Pair-programming with AI is now baseline. Component generation, content drafting, prototyping speed — all transformed.

Experience

Events, environments, installations, live work. The most physical of the disciplines — but real-time generative content is reshaping what's possible inside live moments.

Marketing & Comms

Campaigns, social, internal comms. AI is heavily used — sometimes too much. The challenge here is editorial restraint, not output volume.

06 · Governance & Coordination

Two groups, different jobs.

Governance sets direction — what we work on, what we don't, how we measure. Coordination runs the system — sprints, schedules, blockers, hand-offs. Confusing the two is the most common failure mode.

Coordination operational decisions resolve here strategic ones escalate to Governance
Governance Group · Strategic
  • Sets direction. Decides which focus areas matter, which don't. Approves the annual brief.
  • Allocates capacity. Who's in the Lab. Who's a Member vs Affiliate. How much time the Practice gets.
  • Owns the KPIs. Reviews fluency progression, adoption, festival outcomes. Adjusts the model when something isn't working.
  • Cadence: Monthly review, quarterly direction-setting.
Coordination Group · Operational
  • Runs the rhythm. Bi-weekly sync. Sprint kick-offs and reviews. Editorial calendar.
  • Manages handoffs. Patterns from Lab into Practice. Workflows from Practice into All DS. Catches the seams.
  • Surfaces blockers. Reports up the things that need a strategic call. Doesn't make them.
  • Cadence: Bi-weekly meetings. Always-on Slack channel.
07 · Setup

The plumbing — so the work flows.

Tools, accounts, channels, repos. Boring on the surface but decisive in practice — most stalled rollouts die here, not in the strategy.

Tooling stack

A standard kit — image, video, code, copy. Procured centrally. Approved for client work. The default, not the exception.

Pattern library

A living repo of validated patterns and playbooks — written by the Lab, used by the Practice, contributed to by All DS.

Showcase

Where finished work lands — internal-facing, browseable, searchable. The library that proves the patterns work.

Channels

Slack first. Studio-level channels, regional Practice channels, a single global #creative-ai. Where the texture of daily work lives.

08 · Comms

Four flows. Each with a job.

Bi-weekly Coordination keeps things moving. The monthly Brief reaches everyone. The annual Festival is the public moment. Slack carries the texture in between. Each has a clear sender, audience, and call to action.

Bi-weekly Coordination

Operational pulse. Sprint reviews, blockers, upcoming hand-offs. Internal to the Coordination Group. Notes posted to Slack.

Monthly Brief

Editorial. One issue per month. Highlights work shipped, pattern of the month, ladder progressions. Goes to all of DS. Read in 5 minutes.

Annual Festival

1min.ai. May–August. The public moment. Highlights published externally. The biggest single comms event of the year.

Always-on Slack

The texture. Wins, losses, links, half-thoughts. Where the day-to-day life of the practice actually lives.

09 · Operating Rhythm

Cadences that interlock.

Different rhythms for different jobs. Daily Slack texture. Bi-weekly Coordination. Bi-monthly Lab sprints. Monthly Brief. Quarterly Governance review. Annual Festival. The wheel below shows how they line up across a year.

Annual operating rhythm — cadence wheel showing how each ritual lines up across the calendar year Year of Practice 2026
Always-on Slack
Daily texture · low friction
Daily
Coordination sync
Sprint reviews · blockers · hand-offs
Bi-weekly
Lab sprint
Pattern + playbook output
Bi-monthly
Monthly Brief
Editorial · all of DS
Monthly
Foundations
Programme onboard cohorts
Quarterly
Governance review
KPI · direction · capacity
Quarterly
1min.ai Festival
May → August · annual moment
Annual
10 · KPIs

What we count, what we don't.

Three KPIs at the top. Each tied to a layer. Each a leading indicator, not a vanity metric. We resist the urge to measure what's easy to measure when what's measurable isn't what matters.

+2
Lab adoption
Average ladder rungs gained per designer per year. Measures whether the Lab's patterns actually move people forward.
60%
Practice adoption
Share of active cases using at least one validated AI workflow. Measures whether the Practice translation is actually landing.
100%
All DS reach
Designers who have engaged with at least one Foundations module. Measures whether the floor is actually rising.
11 · WIP & Open Questions

What we're still figuring out.

No model lands fully formed. These are the seams we're watching and the questions we don't yet have clean answers to. Calling them out so the system can self-correct.

Watching Lab capacity vs Practice demand

Lab is small by design — but it gets pulled into Practice work that isn't really R&D. If 30% of Lab time is spent on Practice support, the Lab stops inventing. The fix is probably a clearer triage rubric, not more headcount.

Watching Ladder gaming

The fluency ladder is self-reported. It works because it's used in good faith. If it ever becomes a promotion gate, the rungs become a lie. Keep it descriptive. Keep it public. Keep it out of HR systems.

Open Festival sustainability

1min.ai is a four-month commitment. If only the same studios show up year after year, it ossifies. The brief, the categories, and the editorial board need to keep rotating.

Watching Tooling churn

The tooling stack needs to evolve, but every change costs the practice attention. Refresh quarterly, not monthly. One studio test before broad rollout.

A+ · Appendix

The receipts.

Source material, definitions, and the decisions behind the model. Click any section to expand.

A
Glossary
Lab · Practice · All DS · Pillar · Designation · Affiliate · Friend · Ladder · Foundation · Brief · Festival

Lab. The R&D node. ~20 designated. Owns invention, validation, and packaging of patterns and playbooks.

Practice. The regional translation layer. AMR, EMESA, AP+ME. Adapts what the Lab makes for local craft and case context.

All DS. Every designer in every studio. The user base of the system and its primary feedback engine.

Pillar. One of four operating buckets in the Practice — Explore, Learn, Apply, Share.

Designation. A formal role inside the Lab — Member, Affiliate, Friend.

Ladder. The seven-step Fluency Ladder for All DS — Curious → Trying → Working → Crafting → Building → Teaching → Leading.

Foundations. The onboarding programme. Quarterly cohorts. Self-paced after kick-off.

Brief. The monthly editorial. ~5 minutes to read. One issue, every studio, every month.

Festival. 1min.ai. Annual. May–August. Briefed, made, judged, shipped.

B
Decisions made & rejected
Why three layers, not two; why a ladder, not a certification; why a regional Practice, not a global one

Why three layers, not two. A two-layer model (Lab + everyone) collapses under the weight of regional variance. The Practice exists because Lab patterns need a translator before they can work in AP+ME the way they work in AMR.

Why a ladder, not a certification. Certifications create gatekeepers. Ladders create language. We want shared vocabulary, not a credentialing apparatus.

Why a regional Practice, not a global one. One global Practice would centralise faster than it could adapt. Regional means each region's Practice has authority to shape what reaches its studios.

C
Audit method
Studios audited · interviews · pilot workflows · what we asked · what we filtered out

Twelve studios audited across all three regions. Thirty-seven practitioner interviews — split roughly 60/40 between IC designers and design directors. Nine pilot workflows ran in parallel for at least one full case cycle each.

Filtering: we ignored claims of AI use that weren't backed by an artefact. We weighted answers from people who could show their work over those who couldn't.

D
Ladder rubric, in detail
What "Crafting" looks like on a Tuesday vs what it looks like to a director — concrete behaviours per rung

Each rung is described by behaviours, not capabilities. Curious: reads about it, can name three tools. Trying: has used a tool on real work; can show one artefact. Working: can describe their default pipeline, including AI steps; quality is consistent.

Crafting: their work has a recognisable hand inside an AI workflow. Building: has authored a workflow, custom assistant, or script others use. Teaching: regularly mentors or runs sessions. Leading: shapes direction at studio or practice level.

E
Festival mechanics
Brief structure · categories · judging panel · publishing standard · winner archive

Brief released first week of May. Open theme but always paired with a craft constraint (this year: 60-second piece, generative video).

Categories rotate annually but always include: Best Solo, Best Studio Collective, Best Use of New Tooling, Best Editorial. Judging panel is a mix of Lab Members, regional leads, and one external guest critic. Winners archived in the Showcase, with editorial write-up.

F
Tooling stack — current snapshot
Approved tools · regional variations · review schedule

Stack is reviewed quarterly by the Coordination Group, ratified by Governance. Regions can flag local additions but the canonical stack ships from the centre.

Currently includes: image generation, video generation, code-pair, copy-pair, internal pattern repo, custom GPTs for specific case workflows. Specifics rotate quarterly.

G
Risk register
IP · client confidentiality · brand consistency · vendor lock-in

Every approved tool has a documented IP and confidentiality posture. Client work uses only approved tools — no experimental tools touch client data without explicit sign-off.

Brand consistency: all client-facing AI output goes through the same review chain as any other deliverable. Vendor lock-in: deliberately maintain redundancy across at least two providers in each major capability category.

H
Change log — v9 → v10.5
What changed since the last public version of this doc

v10.0: introduced the three-layer model. Replaced the previous flat hub-and-spokes diagram. First articulation of the seven-step Fluency Ladder.

v10.5: codified Governance vs Coordination split (was previously merged). Added Friends as a third Lab designation. Added explicit risk register and IP posture (Appendix G).

v10.5 · live document

AI raised the floor.
DS raises the ceiling.

BCG Design Studios BCG Design Studios