Business Strategy
Executive view
Technical view
Why this matters
AI Solution Engineers at manager level are asked to “find AI opportunities” but promoted when they protect capital—funding what advances strategy and killing vanity pilots. Executives do not lack technology options; they lack conviction that a programme changes competitive position within the planning horizon.
Weak strategy fluency produces familiar failure patterns:
-
The parallel universe pilot. A brilliant underwriting assistant runs while corporate strategy prioritises broker digitisation and direct channel growth. The pilot completes; the business unit never adopts because it solved someone’s hobby problem.
-
The core capability outsourced by accident. A “fast” SaaS copilot embeds proprietary workflow and customer data in a vendor tenancy with no exit path—contradicting platform strategy and data sovereignty commitments made to the board.
-
The cost-leader dressed as innovation. A bank under cost-to-serve pressure funds a marketing GenAI content factory. Expense rises; NPS unchanged; strategy scorecard misses.
Your role is to translate strategy documents into AI decision filters: which themes get capacity, which capabilities must be owned, what the operating model looks like after adoption, and how to explain trade-offs in language the strategy office uses—where to play, how to win, capabilities required.
Partners expect you to open alignment workshops with the client’s stated priorities (from annual report, investor day, BU plan)—not a generic AI maturity model.
Learn
Vision and mission
Definition. Mission states purpose—why the organisation exists. Vision describes the desired future state—what success looks like at horizon (often 5–10 years). Together they set boundaries for investment.
Why it matters. AI narratives that contradict mission (“automate all human advice” for a trust-based wealth brand) trigger brand and regulatory pushback. Vision sets scale ambition—regional efficiency player vs global digital leader implies different AI bets.
How to use on engagement. Quote exact phrasing from latest annual report or strategy deck when drafting alignment statements. Map use case to mission/vision explicitly: “Supports vision pillar X by …”
Pitfalls. Confusing vision with technology roadmap. Using mission statements so generic any project fits (“deliver shareholder value”).
Mini example. Insurer vision: “Most trusted digital insurer in UK SME commercial.” AI FNOL triage supports trust + digital; generic image generation for social media does not—unless linked to measurable broker acquisition KPI.
Corporate strategy vs business unit (BU) strategy
Definition. Corporate strategy spans the group: portfolio choices, capital allocation, group-wide capabilities, M&A, risk appetite. BU strategy covers how a division competes in its market: products, channels, segments, local operating model.
Why it matters. Group may prioritise platform consolidation while a BU wants best-of-breed copilot. Funding and architecture follow corporate constraints; benefits accrue to BU P&L. Misalignment causes double funding or orphan systems.
How to use on engagement. Identify decision rights: who approves—group CIO, BU COO, or both? Tag each use case corporate enabler (shared platform) vs BU differentiator (segment-specific).
Pitfalls. Solving BU pain that corporate architecture will retire. Ignoring inter-BU politics on data sharing.
Decision table.
| Signal | Likely owner | AI implication |
|---|---|---|
| Group “one platform” mandate | Corporate IT / CDO | Build shared retrieval + governance |
| BU P&L under pressure | BU COO | Prioritise cost-to-serve automation |
| Regulated entity firewall | Group risk / compliance | Separate models, data residency |
Differentiation vs cost leadership
Definition. Porter’s generic strategies (simplified): cost leadership (lowest cost producer), differentiation (premium for unique value), or focused variants on a segment. Stuck in the middle fails to choose.
Why it matters. Cost leaders fund automation, deflection, STP with tight ROI. Differentiators fund experience, personalisation, advisor augmentation where margin supports it. Same AI feature (copilot) has different success criteria.
How to use on engagement. Classify the client’s declared strategy from investor materials. Score use cases: cost, revenue, risk, speed—weight scores by strategy type.
Pitfalls. Applying cost logic to a differentiation brand (e.g. cutting human touch in private banking). Differentiation theatre—GenAI wrapper on commodity product.
Mini example. Low-cost carrier strategy: AI for crew rostering and disruption comms beats bespoke lounge concierge bot. Full-service carrier: loyalty personalisation and agent augmentation align.
Operating models
Definition. Operating model is how strategy is executed: organisation structure, processes, people, technology, governance, and performance management—often depicted as people, process, technology with clear RACI.
Why it matters. AI changes process (human-in-the-loop vs STP), people (roles, skills, capacity), and technology (platform boundaries). Strategy fails if operating model stays unchanged—“bolted-on bot.”
How to use on engagement. Produce current vs target operating model sketch for one value stream (e.g. FNOL). Show where AI sits, handoffs, escalation, metrics.
Pitfalls. Technology-only TOM with no role redesign. Ignoring unions and HR in cost-leader contexts.
Diagram: simplified servicing operating model shift
Value chains
Definition. Value chain maps activities from inbound logistics to service—primary activities (operations, outbound, marketing, service) and support (HR, tech, procurement). Each activity adds margin or cost.
Why it matters. AI should attach to high-leverage activities: large cost pools, bottleneck cycle times, or differentiation-critical steps—not marginal back-office unless strategy is pure cost takeout.
How to use on engagement. Draw industry value chain; mark £/FTE/time heat. Prioritise AI where heat meets strategic theme.
Pitfalls. Starting from available data not value chain. Optimising support functions while core operations starved.
Mini example. In insurance, claims handling and distribution (broker enablement) often dominate strategic AI for combined ratio and growth; HR policy FAQ is secondary unless EX strategy is core.
Build, buy, partner, and outsource
Definition. Build—internal development for control and IP. Buy—licence COTS/SaaS. Partner—co-develop or revenue share with specialist. Outsource—transfer operation to MSP/BPO, possibly with AI embedded.
Why it matters. Strategy defines core vs context (see frameworks). Wrong mode leaks IP, traps data, or burns time rebuilding commodity layers.
How to use on engagement. Score each capability: strategic differentiation, regulatory sensitivity, integration depth, time-to-market. Output build/buy/partner recommendation with exit criteria.
Pitfalls. Build for undifferentiated RAG plumbing. Buy for core advisory logic in regulated advice. Partner without data processing agreements and model audit rights.
Decision table (illustrative).
| Capability | Cost leader tendency | Differentiator tendency |
|---|---|---|
| Generic LLM API | Buy | Buy (+ fine-tune select flows) |
| Enterprise knowledge platform | Buy/partner; configure | Build/partner with strong ACL |
| Regulated decision support | Partner with audit | Build with HITL |
| Contact deflection | Buy omnichannel + integrate | Build experience layer |
Core vs non-core and platform strategy
Definition. Core capabilities create competitive advantage and are protected. Non-core can be commoditised or outsourced. Platform strategy consolidates shared data, integration, identity, and AI governance for reuse.
Why it matters. Corporate AI programmes fail when every BU buys its own copilot—strategy office pushes one platform, many products.
How to use on engagement. Map use cases to platform services (retrieval, prompt registry, eval, observability) vs product features. Align with cloud and data strategy already approved.
Pitfalls. Federation without standards—ten copilots, zero audit trail. Platform purism delaying BU value for 18 months.
Platform, data, and cloud strategy
Definition. Platform strategy commits the enterprise to shared capabilities (identity, integration, data mesh/fabric, AI governance, MLOps). Data strategy defines ownership, quality, access, and monetisation rules. Cloud strategy sets hosting posture: public, private, sovereign, multi-cloud, and exit constraints.
Why it matters. AI programmes fail when each BU buys isolated copilots on different clouds with incompatible logging—contradicting group platform and data strategy approved by the board. Model hosting, vector stores, and PII residency must align with sovereign or industry cloud commitments before you recommend architecture.
How to use on engagement. Request the approved platform/data/cloud strategy documents (not architect oral tradition). Map AI components to reuse mandates and data classification tiers. Flag exceptions early with strategy office—not at security review week.
Pitfalls. Shadow AI SaaS outside data strategy. Data lake fallacy—data exists but is not licensed for model training. Cloud strategy bypass via expensed credit cards.
Mini example. A bank commits to single sovereign cloud for customer data. Recommending a US-only SaaS copilot with no EU residency fails strategy regardless of demo quality—partner/private deployment or group platform becomes the only aligned path.
SWOT analysis
Definition. Strengths, Weaknesses, Opportunities, Threats—internal vs external factors affecting strategy.
Why it matters. AI opportunities often sit in O (automate rising contact volume) and T (fintech competitors with faster onboarding). SWOT grounds AI in external competition, not internal tech enthusiasm.
How to use on engagement. Run a facilitated SWOT in discovery; tag each AI idea with S/W/O/T linkage. Kill ideas that address none.
Pitfalls. Wishful strengths (“we have great data”) without data governance evidence. Static SWOT—refresh quarterly in fast markets.
PESTLE analysis
Definition. Political, Economic, Social, Technological, Legal, Environmental macro factors.
Why it matters. AI in regulated industries is Legal + Political heavy: AI Act, GDPR, FCA guidance, model bias liability. PESTLE explains why governance spend is strategic, not overhead.
How to use on engagement. Document 2–3 PESTLE drivers per use case in risk section of alignment paper. Link to responsible AI investments.
Pitfalls. Treating PESTLE as academic—must tie to go/no-go (e.g. automated credit decision legality).
Porter’s Five Forces
Definition. Industry attractiveness via: rivalry, supplier power, buyer power, substitutes, new entrants.
Why it matters. High buyer power (SME brokers, price comparison sites) pushes cost and speed AI. Substitute threat (insurtech MGAs) pushes digital experience AI.
How to use on engagement. One paragraph: “Five Forces imply priority on X; this use case strengthens …”
Pitfalls. Generic force lists without client-specific evidence. Ignoring supplier power of hyperscaler AI pricing.
Business Model Canvas (BMC)
Definition. Nine blocks: customer segments, value propositions, channels, relationships, revenue streams, key resources, key activities, key partners, cost structure.
Why it matters. BMC shows where AI touches the business model—not only cost. Example: AI enables new revenue stream (usage-based analytics product) or changes key activities (STP claims).
How to use on engagement. Overlay AI interventions on BMC; highlight blocks that change materially.
Pitfalls. Canvas filled with buzzwords; no quantified cost structure change.
Jobs to Be Done (JTBD)
Definition. Customers “hire” products to make progress on a job (functional, emotional, social)—not to interact with features.
Why it matters. “Job: renew commercial policy in 10 minutes without broker callbacks” beats “job: use chatbot.” AI features map to job steps and failure moments.
How to use on engagement. Interview snippets → job map → pain points → AI intervention per step. Measure job completion rate, not message count.
Pitfalls. Solution-first jobs (“use AI”)—circular. Ignoring emotional jobs in trust categories (health, wealth, claims).
Wardley mapping
Definition. Wardley maps plot value chain components on evolution axis (genesis → custom → product → commodity) vs value chain visibility.
Why it matters. Commodity components (generic LLM inference) should be bought; genesis/custom (proprietary underwriting rules + model orchestration) may be built. Maps communicate build/buy to technical and exec audiences.
How to use on engagement. Place components: data pipeline, vector store, model API, domain rules engine, UI, audit. Decide invest vs outsource by evolution stage.
Pitfalls. Over-precision in map placement—use for decision dialogue, not physics. Maps without movement over time (what commoditises in 24 months).
Three Horizons framework
Definition. H1—extend and defend core business. H2—emerging opportunities scaling up. H3—options for future businesses.
Why it matters. AI portfolio balance: H1 copilots and automation (fast ROI); H2 new AI-enabled products; H3 exploratory (agents, new data businesses). Prevents all H3 or all H1 portfolios.
How to use on engagement. Tag each use case H1/H2/H3; align funding gates—H3 needs learning metrics, not year-1 EBITDA.
Pitfalls. Labeling everything H1 to get funded. H3 starvation when competitors commoditise your core.
Playing to Win (Lafley & Martin)
Definition. Five choices: winning aspiration, where to play, how to win, capabilities required, management systems.
Why it matters. Maps cleanly to AI decision papers—executives know this language from strategy offsites.
How to use on engagement. One-page Playing to Win table for the AI programme nested under corporate choices.
Pitfalls. How to win = “use AI” (not a choice). Missing management systems (OKRs, governance, benefits tracking).
Mini template.
| Choice | Client (example) | AI programme alignment |
|---|---|---|
| Aspiration | Top-quartile combined ratio | LAE and fraud AI |
| Where to play | UK SME commercial | Broker-facing assistants first |
| How to win | Fast, accurate quotes | Submission ingestion AI |
| Capabilities | Data platform, MGA ops | Build retrieval + HITL underwriting assist |
| Management systems | OKR on quote TAT | Weekly leading metrics on STP rate |
AI alignment to strategy
Definition. Strategic alignment means a use case advances an explicit strategic theme with measurable linkage, fits operating model and build/buy principles, and has agreed stop rules if strategy shifts. Alignment is necessary but not sufficient—you still need economic credibility from topic 01 and feasible delivery from Stages 2–3.
Why it matters. Alignment is the gate between discovery and funding. Misaligned work consumes scarce data science and change capacity. In group structures, misalignment also triggers architecture rejection when corporate platform mandates conflict with BU shortcuts.
How to use on engagement. Produce alignment statement (5–8 lines): theme, mechanism, KPI, horizon, build/buy stance. Review with strategy office or sponsor exec. Maintain an alignment register across programmes—executives should see one page showing how AI portfolio maps to corporate choices, not a slide per project with inconsistent vocabulary.
Pitfalls. Retrofit alignment after picking the tool. Theme sprawl—one project claiming five strategic pillars weakly. Static alignment—strategy refreshed annually but AI portfolio not re-tagged. Proxy alignment—"supports innovation culture" without KPI.
Decision table: alignment strength
| Level | Evidence | Funding implication |
|---|---|---|
| Strong | Named theme + KPI + owner in BU plan | Eligible for multi-year funding |
| Medium | Theme match; KPI indirect | Pilot with strict leading metrics |
| Weak | Generic digital language | Defer or absorb into existing ops budget |
| None | Contradicts strategy or risk appetite | Stop; document rationale |
Worked alignment (bank cost-to-serve). Theme: digitise servicing. Mechanism: authenticated assistant + copilot on top 12 intents covering 41% of volume. KPI: cost-to-serve index −6% on targeted journeys; CSAT floor −2 pts max. Horizon: H1, 18-month rollout waves. Build/buy: group retrieval build; channel buy; models buy with private deployment. Stop rule: if month-9 deflection < 10% with CSAT breach, halt wave 3.
Frameworks and methods
| Framework | Best for | Time box | Pair with |
|---|---|---|---|
| Playing to Win | Exec decision papers | 2–4 hours workshop | Business case |
| Wardley map | Build/buy/platform | Half-day technical + strategy | Architecture topic 08 |
| JTBD interviews | Customer-facing AI | 6–10 interviews | UX topic 23 |
| Three Horizons | Portfolio balance | 1-hour leadership review | Opportunity discovery topic 05 |
| SWOT / PESTLE | Discovery kickoff | 90 minutes | Risk register |
| Porter Five Forces | Sector entry / competitive pressure | 60 minutes | Industry topic 03 |
| BMC overlay | Business model innovation AI | 2 hours | Business fundamentals topic 01 |
| Value chain heat map | Prioritisation | 2 hours | Process mining data |
Method: strategic alignment workshop (half-day)
- Read last annual report / investor deck—extract 3–5 explicit themes (30 min).
- Map active AI ideas to themes; drop orphans (45 min).
- Classify cost vs differentiation weighting (30 min).
- Wardley sketch for shared platform components (60 min).
- Draft alignment statements + H1/H2/H3 tags (45 min).
- Assign build/buy/partner per component (30 min).
When not to use heavy frameworks. Crisis response (operational outage), regulatory remediation with fixed scope, or single mandated compliance project—use light alignment paragraph only.
Strategy diagnostics checklist (AI portfolio review)
Use quarterly with sponsor and strategy office—30–45 minutes:
- Theme coverage: Are any corporate themes with no AI portfolio entry (gap) or many overlapping entries (sprawl)?
- Horizon balance: What % of spend is H1 vs H2 vs H3? If H3 > 25% without H1 delivery, rebalance.
- Operating model delta: For top three programmes, is TOM change funded in the same business case as technology?
- Build/buy coherence: Do Wardley components match actual contracts (any commodity built in-house)?
- Management systems: Do OKRs reference the same KPIs as alignment statements?
- Stop list hygiene: Are killed ideas documented to prevent zombie resurfacing?
- PESTLE refresh: Any new legal constraint (AI Act, conduct guidance) affecting in-flight work?
Document outcomes in the alignment register—one table executives can read without opening technical backlogs.
Architecture and operating model notes
Strategy drives target architecture principles:
STRATEGY CHOICE ARCHITECTURE PRINCIPLE AI EXAMPLE
---------------------------------------------------------------------------
Cost leadership Standardise, reuse, measure Shared copilot platform
Differentiation Best experience, controlled IP Custom orchestration + brand UX
Regulatory trust Audit, residency, HITL Private inference, full logging
Platform business API-first, partner ecosystem External developer API with quotas
Governance alignment. Management systems from Playing to Win must include: AI review board charter, benefits realisation cadence, kill criteria tied to strategic KPIs—not only technical SLAs.
Data and cloud strategy. Corporate choices on single cloud, sovereign cloud, or multi-cloud constrain model hosting. Strategy doc beats architect preference.
Real-world scenarios
Scenario A: UK retail bank — “digitise servicing, reduce cost-to-serve”
Declared strategy (illustrative). Annual report themes: (1) digitise core servicing, (2) reduce cost-to-serve 15% by 2028, (3) protect trust and conduct, (4) grow SME banking share.
Use case portfolio.
| Idea | Alignment | Verdict |
|---|---|---|
| Authenticated servicing assistant + agent copilot | Themes 1, 2; KPI: cost-to-serve, CSAT | Fund H1 |
| GenAI marketing copy for campaigns | Weak on 1–2; brand risk on 3 | Defer |
| SME onboarding doc extraction | Themes 1, 4; revenue + cost | Fund H2 |
| Experimental autonomous trading agent | Not in retail bank strategy | Kill |
Operating model. Move from queue-first to digital-first with copilot-augmented agents; new roles: knowledge curator, conversation designer, model risk reviewer.
Build/buy. Buy omni-channel layer; build/partner on bank-specific retrieval with ACLs; buy base models; build conduct guardrails.
Measurable outcomes (36 months). Cost-to-serve −8% on targeted journeys; digital completion rate +12 pts; conduct complaints flat or down; payback < 24 months on H1 bundle.
Strategic alignment statement (example). “The authenticated servicing programme advances digitise servicing and cost-to-serve by deflecting high-volume, low-risk intents and reducing handle time on assisted contacts, measured by bank-wide cost-to-serve and journey CSAT, implemented on the group knowledge platform with BU-specific policy corpora—H1 core.”
Scenario B: European P&C insurer — loss ratio, retention, and broker digitisation
Declared strategy. Reduce combined ratio toward 96%; improve retention in SME commercial; digitise broker submissions; strengthen claims excellence as differentiator in cat-prone regions.
Use case portfolio.
| Idea | Alignment | Verdict |
|---|---|---|
| FNOL document AI + triage | Claims excellence; LAE | Fund H1 |
| Underwriting RAG on submissions | Broker digitisation; speed | Fund H1 |
| Generic image GenAI for ads | Low strategic linkage | Kill |
| Predictive cat exposure analytics | Differentiation in regions | H2 |
Wardley view. Commodity: base LLM API, OCR. Product: workflow BPM. Custom: policy rules + HITL underwriting. Build custom orchestration; buy commodity layers.
Three Horizons. H1: FNOL + submission ingestion (ROI). H2: broker portal copilot with quote tracking. H3: parametric product pricing experiments.
Measurable outcomes. LAE/claim −10% on automated paths; quote turnaround −25% on SME commercial; retention +2 pts on targeted broker cohort; loss ratio stable or improved despite volume growth.
Build-vs-buy note (summary). Partner for document extraction only if audit trail and retraining rights retained; prefer group-built retrieval on sovereign cloud; buy telephony/chat channel—not differentiator.
Scenario C (stretch): UK grocery — margin defence and waste reduction
Strategy. Margin defence in inflation; fresh excellence as differentiation; ESG waste reduction commitments public.
AI alignment. Demand forecasting and markdown optimisation (H1) align strongly; in-aisle GenAI concierge (H3) weak unless tied to basket conversion in premium formats.
Lesson. Strategy filters which AI—here operations ML beats customer GenAI for board narrative.
Cross-industry comparison: where strategy points AI
| Industry | Common strategic theme (2025–26) | AI bets that typically align | AI bets that often misalign |
|---|---|---|---|
| Retail banking | Cost-to-serve, digital completion, conduct | Servicing assistant, copilot, onboarding doc AI | Public unauthenticated GenAI on regulated advice |
| P&C insurance | Combined ratio, broker digitisation, claims excellence | FNOL triage, submission ingestion, fraud signals | Generic marketing content factories |
| Grocery retail | Margin defence, waste, promo effectiveness | Forecasting, markdown optimisation | Novelty in-store avatars without conversion KPI |
| Wealth management | Trusted advice, AUM retention | Advisor augmentation, meeting summarisation with supervision | Fully automated portfolio recommendations without governance |
Use this table as a hypothesis starter, not a substitute for reading the client's actual strategy documents.
Practice exercises
Primary exercise: strategic alignment statement
Task. Choose a fictional or real financial institution with a stated strategy (from public materials). For one proposed AI use case (servicing assistant, FNOL triage, or broker submission intake), write:
- Five-line strategic alignment statement (themes, mechanism, KPI, horizon, H1/H2/H3).
- Playing to Win mini-table (five rows).
- Build/buy/partner recommendation for three components (model, retrieval, workflow).
- One paragraph “why not” for a misaligned use case (e.g. marketing GenAI).
Artefact criteria.
| Criterion | Pass |
|---|---|
| Quotes or paraphrases real strategic themes | Yes |
| Names measurable KPI linked to theme | Yes |
| Build/buy rationale references core vs commodity | Yes |
| Misaligned case rejected with strategy logic | Yes |
| No vendor name as strategy rationale | Yes |
Stretch exercise: Wardley map and portfolio tag
Task. Draw a Wardley map (ASCII or diagram) for the chosen use case with ≥8 components. Tag three use cases across Three Horizons. Write executive decision paper outline (max 800 words): problem, strategic fit, options, recommendation, risks, next 90 days.
Artefact criteria. Map includes movement annotation (12–24 months); portfolio has at least one stopped idea with reason; decision paper readable by non-technical COO.
Questions you should be able to answer
- What are the top three corporate strategic themes this year—and where are they documented?
- Which business unit owns the P&L for this problem—and does group strategy agree?
- Is the client pursuing cost leadership, differentiation, or a focused hybrid—and how does that weight AI ROI?
- Which Playing to Win choices does this use case support—or contradict?
- Where on the value chain does this use case sit, and is that chain segment strategically priority?
- What job is the customer or employee trying to get done—and does AI help complete it?
- Is this capability core or context—should we build, buy, or partner?
- What does the target operating model change (roles, handoffs, KPIs)?
- Which horizon (H1/H2/H3) is this—and what evidence gate applies?
- What competitive force (Porter) or PESTLE factor makes this urgent?
- How does this fit platform and data strategy—reuse or one-off?
- What happens to strategic metrics if we stop this programme?
- What management systems (OKRs, governance) must exist for scale?
- Which alternative investments compete for the same capital?
- What would misalignment look like in 12 months—symptoms and costs?
Negative cases
| Failure mode | Symptom | Prevention |
|---|---|---|
| Strategy retrofit | Alignment paragraph added after vendor selection | Themes first; ideas second |
| Theme hoarding | One project claims all pillars | One primary theme per programme |
| Build the commodity | Custom LLM platform for generic chat | Wardley: buy commodity inference |
| Buy the core | Proprietary underwriting logic in SaaS black box | Core logic owned; audit rights |
| H3 portfolio in crisis | Only experiments; core KPIs miss | Balance H1 delivery ≥70% capacity |
| Operating model denial | Bot live; agents unchanged; benefits zero | TOM workshop before build |
| BU vs group warfare | Duplicate copilots, no standards | Corporate platform mandate with BU config |
| Differentiation on cost | Premium brand automated away | Segment-specific service levels |
| Ignoring PESTLE | Automated decisions breach conduct rules | Legal sign-off in alignment gate |
| JTBD ignored | Low job completion; high traffic | Measure job success, not sessions |
Case study: the misaligned image generator. A insurer under combined ratio pressure funded a marketing GenAI image and copy workstream because the CMO saw competitor ads. After 9 months: £1.4M spend, no measurable quote volume or broker NPS movement; conduct queried ambiguous health claims imagery. Programme stopped; H1 FNOL triage remained unfunded. Root cause: no value chain heat map or Playing to Win filter—activity mistaken for strategy.
Case study: platform duplication across BUs. A banking group with one digital platform strategy allowed three BUs to procure separate copilot SaaS products. Benefits fragmented; audit could not trace advice across channels; run-rate tripled. Group architecture paused scale and mandated shared retrieval and logging. Recovery took 14 months—strategy was known; management systems failed to enforce it. Lesson: alignment includes enforceable standards, not only slide alignment.
Case study: differentiation overreach in cost programme. A telco in cost leadership mode launched a premium AI concierge for all customers. Handle time rose (customers experimented); NPS mixed; OPEX up for custom UX. Meanwhile network fault triage—clear cost play—starved. Rebalance to segmented where-to-play: concierge for high-ARPU only; automation for mass market. Strategy was cost leader; product design assumed differentiator.
Related playbook content
- Discovery — opportunity framing before alignment sign-off
- Consulting — problem structuring and executive narrative
- Business case and prioritisation — fund the aligned portfolio
- Architecture — translate platform strategy into target design
- Governance — management systems and decision rights
- Discovery frameworks — structured discovery patterns
- Business Fundamentals — KPI and payback language for alignment
- Industry and Domain Knowledge — sector value chains and regulation
- AI Opportunity Discovery — ideation filtered by strategic fit
- Enterprise Architecture — platform principles from strategy
- Business Learning overview — sector briefings and power words
- How to use this Learning Map — reference depth and artefact standards
- 8D Framework — Define stage requires strategic clarity
- VALUE gate — evidence bar before scale investment
Practice checklist
- I extracted strategic themes from primary sources (annual report, BU plan)—not memory
- I can explain cost vs differentiation weighting for this client
- I produced alignment statement for one use case with measurable KPI
- I stopped at least one misaligned idea with written rationale
- I completed build/buy/partner notes for three components
- I sketched operating model change beyond “add chatbot”
- I linked programme to management systems (OKR/governance), not only tech
- I filed artefacts in pattern library with strategy doc references
Discussion
Comments
Share feedback or questions about this page. No account required.
Loading comments…