AI and business value
AI strategy starts with business outcomes, not tools. This module sets out where AI creates value and the choices leaders must make.
Learning objectives
- Distinguish outcome-led from technology-led AI strategy
- Name four value levers with examples
- Identify strong and weak fits for today's AI
- Describe three levels of AI ambition
1.1 Strategy first, technology second
An AI strategy is not a list of tools. It's a set of choices about where AI will create value for your organisation, what you'll need to make that happen (data, people, platforms, governance), and what you'll deliberately not do. The most common failure is starting with the technology — "we need a chatbot" — instead of a business problem.
Technology-led
"Let's roll out an AI assistant to everyone and see what happens." Usage spikes, then fades; nobody can say what changed.
Outcome-led
"Customer service spends 40% of its time drafting routine replies. Can AI halve that while keeping satisfaction steady?" A clear goal, owner and measure.
1.2 Four ways AI creates value
| Value lever | Examples |
|---|---|
| Productivity & cost | Drafting, summarising, document processing, coding assistance, automating routine steps |
| Revenue & growth | Personalised recommendations, faster sales proposals, new AI-enabled products |
| Customer & employee experience | 24/7 answers, faster resolution, less tedious work |
| Risk & quality | Fraud detection, compliance checks, fewer errors, better forecasting |
1.3 What today's AI is good at — and not
Strong fit
- Language-heavy tasks: drafting, summarising, classifying, extracting
- Pattern recognition in large datasets
- Tasks where a human checks the output
- High-volume, repetitive judgement calls with clear examples
Weak fit (today)
- Decisions needing guaranteed accuracy with no human check
- Tasks with little data or no clear definition of "good"
- Problems that are really about process, incentives or politics
- Anything where an error is catastrophic and unrecoverable
The non-technical background is in AI Literacy for Work and How AI and LLMs Work.
1.4 Three levels of ambition
- Individual productivity — staff use general AI assistants for everyday tasks. Quick, cheap, broad; value is real but diffuse and hard to measure.
- Process transformation — redesign a specific workflow (claims, onboarding, support) around AI. Larger, measurable gains; needs integration and change management.
- New products and business models — AI becomes part of what you sell. Highest potential and highest risk.
Most organisations need all three, in different proportions. A good strategy is explicit about the mix.
1.5 Questions a leadership team should answer
- Which two or three business outcomes matter most over the next 18 months?
- Where does AI plausibly move those outcomes — and how would we know?
- What data, skills and platforms do we have, and what's missing?
- What risks are unacceptable, and who decides?
- How much will we invest, and how will we decide to stop or scale?
Pick your organisation (or one you know). For each of the four value levers, write one specific opportunity and the measure that would show it worked.
See a worked example
Example — online homeware retailer: Productivity: AI drafts replies to routine support emails (measure: handling time per ticket). Revenue: personalised product bundles (measure: average order value). Experience: instant order-status answers (measure: contact rate, CSAT). Risk: anomaly detection on refunds (measure: fraudulent refunds caught).
Key takeaways
- Start with outcomes, not tools.
- Value comes from productivity, growth, experience and risk reduction.
- AI fits language-heavy, high-volume, human-checked work best.
- Balance individual productivity, process redesign and new products.
- Leadership must set priorities, risk limits and stop/scale rules.