Case 06

Strategy and Design Thinking Analysis

Situation

The Learning and Enablement team at Meridian Group wants to improve adoption of a new internal Copilot-style productivity tool. The tool is already available to all staff, but usage is uneven. Some teams use it daily for drafting, summarising, and meeting preparation. Others tried it once and stopped.

Leadership believes the tool can reduce repetitive work and improve the quality of communication. However, the early rollout focused mainly on access and basic training. There was limited follow-up on use cases, behaviour change, measurement, or manager involvement.

The team must decide how to improve adoption over the next quarter without creating a large programme. Budget is limited. The sponsor wants visible progress within 8 weeks.

Current Barriers

- Many users do not know what good prompting looks like in their specific role.

- Some users treat the tool as a chatbot, not a work-product assistant.

- There is confusion around safe use of internal information.

- Managers have not been given team-level adoption playbooks.

- No one is measuring whether usage leads to better or faster work.

- There are no internal champions or peer examples yet.

Prompt

Act as my strategy and design-thinking colleague.

Context:
[paste the situation, problem, idea, meeting notes, proposal, or project update]

Goal:
I want to think through this clearly and make a better decision.

Apply these frameworks:
1. SWOT analysis
2. Root cause analysis using 5 Whys
3. Stakeholder mapping
4. Value vs effort prioritisation
5. Weighted option scoring

Output:
Present the analysis in clear tables.
Highlight key insights, risks, assumptions, and recommended next steps.
End with a practical recommendation I can act on.

Example output from ChatGPT

Meridian Group – Copilot Adoption Strategy

Strategy and design-thinking review | 8-week intervention | 12-week planning horizon

Executive assessment

Meridian Group does not primarily have an access problem. It has a workflow adoption problem.

The organisation has deployed the technology but has not sufficiently helped employees understand where to use it, how to use it safely, or why it improves their daily work.

The next quarter should focus on shifting employees from trying a chatbot to repeatedly producing useful work with AI.

Time to demonstrate progress

8 weeks

Recommended approach

Focused pilot

Primary success measure

Repeat value

Core recommendation: Launch an eight-week, manager-led adoption sprint across three representative teams. Equip them with role-specific workflows, safe-use guidance, peer champions, and lightweight measurement of time saved and output quality.

Use weeks 9–12 to refine and extend the approach based on evidence.

1. SWOT analysis

DimensionFindingsStrategic implication
StrengthsTool available to all employees; some teams already use it daily; leadership supportBuild on existing users and real internal examples rather than starting another generic rollout
WeaknessesInconsistent prompting skills; no manager playbooks; weak measurement; limited follow-upPrioritise workflow enablement, management accountability and outcome tracking
OpportunitiesReplicate proven use cases; create peer champions; improve writing and meeting preparationStart with common, low-risk, high-frequency tasks
ThreatsSensitive-data misuse; inaccurate AI outputs; employee scepticism; competing prioritiesMake safe use visible, require review of outputs and keep the programme small

Key SWOT insight

The strongest opportunity is already inside Meridian: employees who are finding value.

Rather than teaching everyone what Copilot can do, Meridian should demonstrate what colleagues are already achieving and make that behaviour easy to replicate.

But two conditions matter: examples must be approved for safe use, and any claimed improvement must be measured rather than assumed.

2. Root cause analysis – 5 Whys

Problem: Employees try the tool once but do not continue using it.

1. Why?

They do not experience enough value to return.

2. Why?

They experiment with generic questions rather than recurring work tasks.

3. Why?

Training explains features and basic prompts rather than role-specific workflows.

4. Why?

The rollout was managed mainly as technology enablement, not behaviour change.

5. Why?

Hypothesised root cause

No clear owner was assigned to sustained adoption and measurable work outcomes.

This is a working causal hypothesis, not a proven diagnosis. The first week should test it using interviews with frequent users, one-time users, non-users and their managers.

Root-cause-to-intervention mapping

Root causeInterventionEvidence of progress
No clear personal valueRole-specific use-case demonstrationsMore users complete a real task successfully
Weak workflow integrationReusable prompts and end-to-end examplesRepeat use of selected workflows
Unclear safe-use boundariesOne-page approved-use guideImproved understanding in scenario checks
Limited manager involvementWeekly 10-minute team practiceIncreased participation across pilot teams
No feedback or accountabilitySimple adoption and value scorecardObservable changes in usage and output quality
No peer reinforcementChampions sharing examplesMore independently adopted workflows

The central design question becomes:

How might we help an employee complete one recurring task faster and better, with confidence that they are using AI safely?

That is a much more actionable question than "How do we get everyone using Copilot?"

3. Stakeholder mapping

Influence vs interest

HIGH INFLUENCE – HIGH INTEREST

Manage closely

Executive sponsor, Learning & Enablement lead, pilot managers

Weekly progress and decisions

HIGH INFLUENCE – LOWER INTEREST

Keep satisfied

IT leadership, Information Security, Compliance

Early control review and exception escalation

LOWER INFLUENCE – HIGH INTEREST

Engage actively

Daily users, prospective champions, interested employees

Co-design, testing and peer examples

LOWER INFLUENCE – LOWER INTEREST

Listen and support

Occasional users, sceptics, non-pilot employees

Short examples and optional assistance

Provisional mapping. Actual influence and interest should be validated with the stakeholders.

StakeholderWhat they needResponsibility
Executive sponsorEvidence of meaningful progressApprove scope and remove blockers
Learning & EnablementA repeatable programmeOwn delivery, content and adoption metrics
Pilot managersMinimal effort and useful workflowsEmbed use cases into team routines
EmployeesPractical value, safe-use confidenceTest workflows and give feedback
ChampionsRecognition and time to contributeDemonstrate practices and support peers
IT / Security / ComplianceAppropriate safeguardsApprove usage boundaries and measurement approach

Critical stakeholder insight: Managers are the bridge between learning and consistent behaviour. If they do not reinforce the use of Copilot in actual work, even good training may fail to change habits.

4. Value vs effort prioritisation

The first eight weeks should favour activities that create visible employee value without heavy technical work or a large training programme.

Intervention priority matrix

Quick wins

High value – Low effort

Role-specific prompt packs

Safe-use guidance

Peer demonstrations

Strategic bets

High value – Higher effort

Manager-led adoption sprint

Workflow outcome measurement

Supporting activities

Lower value – Low effort

Launch reminders

General awareness messages

Defer

Lower near-term value – High effort

Large training curriculum

Custom AI adoption portal

Qualitative prioritisation based on the scenario, not measured impact or effort.

Suggested delivery priority

InitiativeValue / 5Effort / 5Priority
Role-specific workflow and prompt packs52P1
Approved safe-use guide52P1 – prerequisite
Manager adoption playbook53P1
Pilot champions and peer examples42P1
Lightweight outcome measurement53P1
Company-wide awareness campaign22P3
Custom training platform25Defer

Scores are directional estimates. Effort is scored from 1 (lowest) to 5 (highest). The safe-use guide is a prerequisite even if other activities appear attractive.

5. Weighted option scoring

Four possible strategies are worth comparing.

OptionDescriptionMain trade-off
A. Broad training refreshOffer new courses to all staffWide reach, limited workflow reinforcement
B. Champions networkRecruit enthusiastic users to share examplesLow cost, but adoption may remain uneven
C. Focused adoption sprintCombine workflows, managers, champions, safeguards and measurementRequires coordination and manager commitment
D. Technical optimisationInvest in integrations, automation and advanced featuresPotential longer-term gains, high near-term effort

Interactive decision matrix

Adjust the weights or individual ratings to test how the recommendation changes. Ratings are illustrative judgments, not validated organisational data.

Scoring model

Step 1: Adjust decision criteria weights. They are automatically normalised to 100%.

Adoption impact

30%

Speed to value

25%

Budget fit

20%

Scalability

15%

Risk and governance fit

10%

Step 2: Change option ratings from 1 (weak) to 5 (strong). Higher is better for every criterion, including budget and safety.

CriterionOption AOption BOption COption D
Adoption impact3354
Speed to value4441
Budget fit3541
Scalability4354
Risk and governance fit3353

Weighted scores

1. Focused adoption sprint

4.55/5

2. Champions network

3.65/5

3. Broad training refresh

3.40/5

4. Technical optimisation

2.55/5

Highest-scoring option

Focused adoption sprint

A decision aid, not proof of programme effectiveness. Validate scores with stakeholders and actual constraints.

With the initial weights, the focused adoption sprint scores 4.55/5, followed by the champions network at 3.65/5, broad training at 3.40/5, and technical optimisation at 2.50/5.

The sprint combines the strongest elements of training and peer learning with manager involvement and outcome measurement. It is therefore the preferred option, provided Meridian can secure manager commitment.

6. Recommended 8-week action plan

The pilot should test whether a small amount of targeted support can produce sustained behaviour change.

Proposed milestones – targets and activities are recommendations, not existing Meridian commitments.

TimelineKey actionsDeliverableOwner
Week 1Select three teams, interview users and non-users, capture baselinePilot charter and baselineEnablement lead
Week 2Select 2–3 workflows per team; approve data-use boundariesWorkflow packs and safe-use guideEnablement + Security
Weeks 3–4Train managers and champions; run practical demonstrationsTeam playbooks and first completed tasksTeam managers
Weeks 5–6Reinforce through weekly team practice; collect feedback and examplesAdoption scorecard and improvementsChampions + Enablement
Week 7Compare outcomes against baseline; identify what workedEvidence-based findingsEnablement + Analytics
Week 8Present results and decide whether to expand, refine or stopSponsor decision packExecutive sponsor
Weeks 9–12Extend successful patterns selectivelyScaled adoption toolkitEnablement + Managers

What should employees actually learn?

Avoid generic prompt-engineering classes. Teach complete tasks with clear inputs, usable prompts, verification steps and finished outputs.

AudienceExample workflowWhat good looks like
OperationsConvert service notes into a structured handoverAccurate actions, owner and deadline
ManagersTurn meeting notes into a decision briefClear decisions, risks and next steps
Corporate staffImprove a complex email or presentationClearer, concise communication suited to the audience

Every workflow should explain when AI is appropriate, what data may be entered, what must be checked, and who remains accountable for the final output.

7. Measurement and success criteria

The sponsor needs evidence of value, not a chart showing how many times someone opened Copilot.

KPIHow to measureSuggested week-8 target
Repeat adoption% of pilot employees using an approved workflow in at least 2 separate weeks during weeks 5–860%
Workflow completion% of participating employees completing 3 real work tasks with Copilot70%
ProductivityMedian time difference for comparable tasks, including checking and rework15% faster
Output qualityBlind review using a consistent rubric for accuracy, completeness and clarityNo deterioration
Safe-use understandingShort scenario-based assessment90% correct
Manager reinforcement% of pilot teams completing scheduled weekly practice80%

These are proposed targets, not forecasts. Final thresholds should be agreed after baseline measurement. Report both participation and absolute numbers so the sponsor can understand the sample size.

Important: Faster work is not necessarily better work. If employees save five minutes drafting but spend ten minutes correcting errors, Meridian has not achieved its objective.

8. Risks, assumptions and evidence gaps

Priority risks

RiskLikelihoodImpactMitigationOwner
Managers do not participateMediumHighSecure explicit time commitments before launchSponsor
Sensitive information entered incorrectlyMediumHighApproved-use guidance, access controls and escalationSecurity / Compliance
AI outputs contain errorsHighMedium–HighHuman review and task-specific quality checksTeam managers
Adoption is measured without business valueHighHighBaseline, task comparison and output quality measuresEnablement / Analytics
Champions become overloadedMediumMediumLimit commitment and rotate responsibilityEnablement
Pilot succeeds only with enthusiastsMediumHighInclude sceptics and infrequent users in pilot samplingEnablement

Risk assessments are provisional.

Critical assumptions to validate

  • Meridian's Copilot tool supports the intended workflows and has appropriate enterprise data protections.
  • There are three teams willing to allocate limited time to the pilot.
  • Basic usage information and task outcomes can be measured in a privacy-appropriate way.
  • Security and Compliance can approve a practical safe-use guide quickly.
  • Existing daily users have examples that can be shared without exposing confidential information.

Evidence missing before launch

Missing evidenceWhy it mattersHow to obtain it
Baseline active and repeat useEstablishes whether adoption improvesExisting analytics, subject to privacy approval
Reasons for abandonmentTests the root-cause hypothesisShort user interviews
Most frequent repetitive tasksIdentifies high-value starting pointsManager and employee workshops
Current safe-use policyEstablishes permissible workflowsSecurity / Compliance review
Typical task duration and qualitySupports an outcome-based business caseBefore-and-after task sampling
Capacity of managers and championsDetermines feasibilityPilot participation commitments

9. Eight-week decision gates

Instead of automatically scaling after the pilot, use evidence to decide.

Sponsor readiness check

0/5 met

Safe-use boundaries approved and no unresolved serious control issues

At least 60% repeat adoption achieved in pilot

Median task time improved by at least 15%

Output quality maintained or improved

Pilot managers commit to continuing the approach

Indicative decision

Hold expansion until governance and safe-use requirements are satisfied.

This decision guide assumes each checkbox is backed by verified pilot evidence.

10. Final practical recommendation

Recommend that the sponsor approve a focused eight-week Copilot adoption sprint, rather than another company-wide training campaign.

The approval should cover three pilot teams, one accountable Enablement owner, a manager and champion for each team, 2–3 approved workflows per team, and lightweight measurement. Use existing tools and training channels wherever possible.

The next action is to organise a 45-minute planning meeting with the Enablement lead, pilot managers, IT/Security and the executive sponsor. Leave the meeting with a named owner, three selected teams, an agreed governance review, a baseline measurement plan and a start date.

The decision requested from leadership is simple:

Approve a small, measurable eight-week pilot to demonstrate whether Copilot can help employees complete recurring work faster, with equal or better quality and appropriate safeguards. Expand only where the evidence supports it.

Bottom line: Stop measuring success mainly by whether people have access or have attended training. Start measuring whether employees can repeatedly produce better work, safely, with less effort.

That is the behaviour Meridian needs to scale.