The Converge Methodology
From "AI is interesting" to delivered, governed value
The Converge Methodology is the method we use to take an organisation from curiosity about AI to a prioritised, deliverable roadmap, and then to working systems. It is published here in full, with the artefacts we use, because a method you can inspect is one you can trust.
Most AI programmes stall because they start with tools
The pattern is familiar. A vendor demo impresses the board, a pilot is funded, and six months later nobody can say what it changed. The tools were never the problem. What was missing was a shared understanding of where AI creates value in this business, which opportunities are ready and which are not, and a way of deciding between them that the business trusts.
The Converge Methodology fixes that in order. It builds a common vocabulary first, opens the field of ideas as wide as it will go, then narrows it with the business's own judgement into a small set of well-formed, assessable use cases, each placed on a realistic time-to-deploy horizon. Only then does anything get built.
We have run it across engineering, manufacturing, procurement, sales and marketing functions for a global industrial group, and for the leadership team of one of the UK's largest magazine publishers. The artefacts below are the ones we used.
How the Converge Methodology runs
A full programme takes eight to twelve weeks from first session to roadmap. A smaller business, or a single function, can run it in four.
Research and preparation
We tailor everything to your sector before anyone sits in a room: the competitor AI landscape, the vendor landscape for your functions, the regulatory map, and your own systems and data estate.
Awareness
Leadership and functional sessions that build a shared vocabulary. We introduce the AI Opportunity Pyramid, the four levels of AI, and the responsible-AI principles, worked on scenarios from your own business.
Divergent and convergent discovery
One divergent workshop per function to surface every idea without judgement, then one convergent workshop per function to cluster, score and force-rank them. We pre-score as a straw man; the room corrects it.
Consolidation and prioritisation
Ideas that share a capability are consolidated into single use cases ("one licence, many benefits"), written up in your own investment format, assigned an adoption pathway, and sequenced by readiness.
Pilot and scale
Delivery of the quick wins, the data-in-hand proofs of value and the foundations, with us as adviser, builder, or taking a bounded pilot on end to end.
What we use, and publish
Every consultancy names a framework. These are the ones we actually put in front of clients; each has its own page.
The AI Opportunity Pyramid
Four layers of time-to-deploy, from off-the-shelf copilots (days) to embedded and bespoke AI (quarters). Reading an opportunity's position alongside its priority is what separates "valuable" from "valuable and ready now".
The four levels of AI
Copilot, assistant, agent, agentic system: one scenario from your own business shown at each level, so leaders can see the shift from AI that assists people to AI that owns outcomes within governed boundaries.
The four adoption pathways
Research-led, product-led, system-embedded and data-foundation. Each use case follows the pathway its nature demands; treating them all as "switch it on" is the surest way to disappoint.
The convergence scorecard
Ten elements scored for every idea, from strategic fit and feasibility to cost, benefit, risk and data position, so any two opportunities in any function can be compared like for like.
Three findings that hold across every programme
Data, not technology, is the constraint. The tools for most opportunities already exist and are proven. What decides whether an opportunity can be delivered, and how soon, is whether the data it depends on exists in a usable form. In our last full programme, only around a third of the ideas raised had data ready today.
Appetite and readiness pull in opposite directions. The themes that attract the most ideas are frequently the least ready, because the data behind them is missing or scattered. So the work is sequenced by readiness, not enthusiasm, and the data foundations start early because so much depends on them.
A blanket block on AI tools is the costliest mistake. It does not stop AI use; it pushes it into unsanctioned personal accounts, which is the very data-leakage risk the block was meant to prevent, and it forfeits the cheapest, highest-value wins. The answer is governed, sanctioned access: enterprise licences with no-training terms, data classification, role-based phased rollout and a usage policy.
About the method
How long does a Converge Methodology programme take?
A full programme across five functions runs eight to twelve weeks from the first awareness session to the prioritised roadmap. A single function or a smaller business can complete it in four. Pilot and scale is scoped separately once the roadmap is agreed.
Who needs to take part?
The leadership team for the awareness session, then the people who do the work in each function for discovery, usually six to fifteen per workshop with product management and IT represented in each. Our last full programme involved close to seventy people across ten workshops.
What do we get at the end?
A findings report with every idea captured and traceable, a set of consolidated use cases written in your own assessment format with problem, value, cost, complexity, data position, risks and candidate vendors, a readiness-sequenced roadmap, and a recommended first quarter.
Does it only apply to generative AI?
No. Generative AI is one branch of the toolkit. The awareness sessions cover classical machine learning, computer vision, optimisation and digital twins as well, which in manufacturing and operations often offer faster returns than language models.
Can we run it ourselves?
The artefacts are published so you can. Most clients find the value is in the facilitation and the pre-scoring: an outside straw man that the room challenges produces a better-calibrated result than a blank page.
Ready to find out where AI creates value in your business?
Start with a leadership awareness session, or go straight to discovery.