AI Scaling Blockers
AI Scaling Blockers is a diagnostic canvas from Board of Innovation for teams whose AI pilots are not turning into scaled use. Participants map what is holding them back in four areas: strategy and organisation, technology and data, governance and operating model, and people and culture. They then cluster the blockers, prioritise the ones to remove in the next 6 to 12 months and define actions for each.
Facilitation script
- 1
Frame the session: which AI initiatives are in scope and what scaling would mean for them. Introduce the four areas of the canvas.
10 min - 2
Silent mapping: everyone writes blockers from their own experience, one per note, and places them in the four areas.
15 min - 3
Walk through each area, clarify notes and cluster related themes. Mark blockers named by more than one function.
20 min - 4
Prioritise the blockers most critical to remove in the next 6 to 12 months, for example by dot voting followed by a short discussion.
10 min - 5
Define a targeted action or enabler for each priority blocker, with an owner and a date.
20 min - 6
Agree the date for revisiting the canvas and how progress will be reported.
5 min
Tips
Ask for examples from specific pilots; 'no data strategy' is a label, while 'the churn model cannot go live because nobody owns the customer table' is a blocker someone can remove.
Have people write silently before any discussion so that the technology side does not set the frame.
Watch for blockers that are placed in one area but caused in another, for instance a people problem that is really an unclear ownership problem.
Limit the priorities to what the people in the room can influence.
Common pitfalls
Writing abstract labels such as 'culture' or 'data quality'. Nobody can act on them and the same canvas could be produced by any company
Inviting only the technology side. The canvas fills up in one area and the organisational and governance blockers stay invisible
Prioritising a dozen blockers. Effort is spread thinly and none of them is removed within the year
Defining actions that belong to people outside the room and have no owner inside it, so nothing happens after the session
Running the diagnostic once. New barriers appear as the first ones are removed, and the original canvas goes out of date
Variations
For a remote session, use a digital whiteboard with the four areas and private note mode during the mapping step. In large organisations, run the mapping separately per business unit and compare the canvases to find blockers that recur across units. Board of Innovation also publishes an AI Readiness Radar, which looks at enablers; the two can be used together. The source does not state a duration or group size, so the figures here are practical estimates.
Where it fits
When to use it
Several AI pilots have run and few or none have reached regular use
Functions blame each other for slow progress and you need a shared picture of the obstacles
A steering group has to decide where to invest enablement effort over the coming year
The same obstacles keep appearing in different projects
You are setting up an AI governance or transformation programme and need a starting diagnosis
When not to use it
The organisation has not run any AI pilots yet. There is nothing to diagnose, so start with the AI Opportunity Radar or an AI Innovation Day
One specific pilot failed for a technical reason. Use a root cause method such as an Ishikawa Diagram
The people who could remove the blockers are not in the room and will not see the result. The session then produces a complaint list
A single team wants to agree how it uses AI day to day. AI Working Agreements fit that purpose
You are weighing whether to go ahead with one particular change. Force Field Analysis compares driving and restraining forces for a single change
Related methods
Frequently asked questions
What is AI Scaling Blockers?▾
AI Scaling Blockers is a diagnostic canvas from Board of Innovation for organisations whose AI pilots do not scale. A cross-functional group maps blockers in four areas, clusters them, prioritises the most critical ones for the next 6 to 12 months and defines actions. It is revisited regularly to track progress.
What are the four categories on the canvas?▾
The four areas are strategy and organisation, technology and data, governance and operating model, and people and culture. Typical entries include scattered pilots without ownership, legacy systems and siloed data, unclear roles and weak guardrails, and low AI fluency or scepticism. The split helps a group see that stalled scaling is rarely only a technology problem.
Who should attend an AI Scaling Blockers session?▾
Bring people from each function that touches AI work: business owners of the pilots, IT and data, legal or risk, and HR or change management. Include at least one person with the authority to remove blockers. The source does not prescribe a group size; four to twelve people is workable.
How long does the diagnostic take?▾
Board of Innovation does not state a duration. One to two hours is realistic for a group that knows the pilots in question. Larger groups or a broad portfolio need the longer end, mostly for clustering and action planning.
How is it different from Force Field Analysis?▾
Force Field Analysis weighs driving and restraining forces for one proposed change. AI Scaling Blockers looks only at barriers, sorts them into four fixed organisational areas and is specific to scaling AI. Use Force Field Analysis for a single decision and this canvas for a portfolio-level diagnosis.
Plan your next workshop with AI
Workshop Weaver helps you combine methods like AI Scaling Blockers into a complete, timed agenda in minutes.
Try it freeMethod descriptions on Workshop Weaver are original content written by our team, based on established facilitation practices. This method was inspired by work from Board of Innovation.