Capstone: a three-horizon technology roadmap

0:00 Capstone: three-horizon roadmap

This is where everything comes together. In this capstone, you'll build a three-horizon technology roadmap for a real business, a decision document leaders can act on, not a trend report. We'll cover the three horizons model, the components of a strong roadmap, a condensed worked example, and how to present it so it actually changes decisions.

0:24 The three horizons

The Three Horizons framework, popularised in the book The Alchemy of Growth and used widely by strategists, organises initiatives by time and certainty. Horizon one defends and improves the core, roughly now to twelve months, using proven technology. Horizon two builds emerging opportunities over one to three years, with pilots and early scaling. Horizon three creates options for the future, three years and beyond, through research, partnerships, small experiments and monitoring.

0:55 Why it matters

Why does this matter? Because most technology plans fail in one of two ways. Either they're a list of trendy tools with no link to the business, or they're so focused on this year that the organisation is blindsided three years later. A three-horizon roadmap forces you to connect technology to real problems today, build a pipeline of tested opportunities for tomorrow, and keep cheap options open for the future. It's also the document that turns everything you've learned in this course into decisions your leaders can act on.

1:34 Managing a farm

Here's an analogy. The three horizons are like managing a farm. Horizon one is this season's harvest: you tend it carefully and measure yield. Horizon two is the new orchard you planted two years ago: it needs investment and patience before it pays. Horizon three is the seed bank and the test plots: small, cheap experiments that might become the crops of the future. A farm that only harvests starves later. A farm that only experiments starves now.

2:08 Simple example: a design studio

A simple example for a five-person design studio. Horizon one: use AI tools to speed up mock-ups and client presentations, measured by hours saved per project. Horizon two: pilot a new service designing AR product try-ons for local retailers, with a gate of two paying clients in six months. Horizon three: watch AI-generated interactive 3D and spatial design tools, with a signpost of when major clients start asking for them. Three lines, three horizons, three different ways of measuring progress.

2:43 Manage each horizon differently

Time ranges differ by industry. What matters is balance, and managing each horizon differently. Horizon one runs on efficiency and KPIs. Horizon two runs on experiments and learning milestones. Horizon three runs on options and signposts. Judge a horizon three idea by horizon one metrics, and you'll kill every future option before it can teach you anything.

3:08 Nine components

A strong roadmap has nine components. A one-page business context. A summary of your horizon scan. Your scenarios with indicators. An initiative portfolio across all three horizons, each with a scorecard summary. A robustness check across scenarios. Enablers: data, skills, governance, security including post-quantum planning, partnerships and budget. A timeline with stage gates and evidence thresholds. Risks and ethics. And signposts with a review cadence.

3:36 Example: coffee chain, H1

Here's a condensed example: a specialty coffee chain with forty cafés in the UK and UAE, online subscriptions and a loyalty app. Horizon one: AI-assisted demand forecasting and scheduling, AI-drafted customer service in English and Arabic with human review, an agent-friendly website and feeds with AI search monitoring, and a post-quantum cryptography inventory.

3:59 Example: H2 and H3

Horizon two: a digital twin of roasting and cold-chain logistics for maintenance and energy, personalised subscription recommendations with on-device features for privacy, readiness for agentic commerce, and a synthetic media policy with Content Credentials. Horizon three: watch robotic drink preparation and smart-glasses training, take an option on a university partnership for sustainable cooling in the Gulf, and hedge against AI-dominated discovery by investing in verified content and community.

4:29 Stage gates

Every initiative needs a gate. For example: the digital twin moves to scale if its predictions beat the maintenance baseline for three months. Smart-glasses training moves to pilot when devices meet comfort and heat criteria and two peer operators report production use. Gates like these stop hype-driven scaling and give leaders confidence that progress is earned. The lesson text has a roadmap template, and a prompt that makes an AI play a sceptical board member and a pragmatic CFO to critique your work.

5:05 Presenting

When you present, lead with the decisions you need and the horizon one value case. Show the portfolio in one visual. Show how initiatives hold up across your scenarios. Be explicit about uncertainty and what would change your mind. And end with review dates. Your roadmap will be judged on evidence, balance, rigour, robustness, enablers and risk, and clear communication. It's a living document, so revisit it at least every six months.

5:36 Balance the portfolio

Watch the balance of your portfolio. A common failure is putting nearly everything in horizon one, because it's comfortable and measurable, leaving no pipeline for the future. The opposite failure is a roadmap full of exciting horizon three ideas with no near-term value to fund them. A healthy roadmap shows clear value now, a few credible pilots in horizon two, and small, cheap options in horizon three.

6:05 Stress-test before presenting

Before you present, stress-test your roadmap with the critique prompt in the lesson text. Ask the AI to play a sceptical board member, then a pragmatic CFO. Where is the portfolio unbalanced? Which initiatives lack owners or thresholds? Which claims need a source? Which initiative is weakest across scenarios? Fix what it finds, then have a trusted colleague do the same. Two rounds of critique usually make a roadmap dramatically stronger.

6:36 Three mistakes

Three common mistakes on roadmaps. First, everything in horizon one, which feels safe but leaves no pipeline for the future. Second, a roadmap full of exciting horizon three ideas with no near-term value to fund them. Third, no stage gates, so initiatives drift forward on enthusiasm or get quietly abandoned. Give each initiative an owner, a gate, and a date when you'll look at the evidence.

7:05 Try this now

Try this now. Choose a real business and write its one-page context: strategy, customers, markets and constraints. Then list three candidate initiatives for each horizon. For each, write a one-line problem it solves, a quick scorecard across the eight dimensions, an owner and a gate with an evidence threshold. Check each initiative against your four scenarios. Finally, paste the draft into an AI assistant with the critique prompt, asking it to play a sceptical board member and then a pragmatic finance director, and revise based on what holds up.

7:44 Course complete

Congratulations, you've reached the end of the course. You can now judge emerging technology with clear lenses and a scorecard, understand the frontier of AI, robotics, spatial computing, digital twins, quantum and post-quantum security, weigh energy and trust constraints, anticipate changes in search and work, and run horizon scanning and scenario planning. Your final step: build your three-horizon roadmap, critique it with the prompt, revise it, and then take the final assessment. Well done.