Converge Methodology artefact
Adopting AI is not a single act: the four pathways
The opportunities a business identifies reach deployment along different routes, and treating them all as though they can simply be switched on is the surest way to disappoint. Every consolidated use case in a Converge Methodology programme is assigned one of four pathways: what actually has to happen before it can deliver, and in what order.
Four routes to deployment
Research-led
For opportunities that depend on data or capabilities the business does not yet know it can rely on. Starts with "do the required data and capability actually exist?", proceeds through a bounded prototype, and ends in a clear decision to commit or close. Funded as a research programme, not scoped as a business case with fixed returns until the evidence exists.
Product-led
For the large share of opportunities a market product already meets. Market scan, selection, validation on a bounded scope, adoption. The discipline is that the scan comes first: a shortlist of candidates is the start of selection, not the end, and the choice depends on fit with existing systems, data and licensing.
System-embedded
For capabilities arriving inside systems the business already runs, most obviously the ERP and CRM, as vendors add AI to their products. The pathway is validation, not acquisition: confirm the feature does what is claimed, check the data it draws on, weigh governance and change, then switch it on. The cost may be modest; the diligence is not optional.
Data-foundation
For opportunities where the data does not exist, cannot be reached, or is scattered across systems in inconsistent formats. No amount of tooling helps until the data is addressed, so these begin with a collection and pooling initiative that precedes research, product selection or system adoption alike. Master data and a trusted knowledge foundation are enablers, not ends.
The pathways are not a ranking
A use case follows the pathway its nature demands, and several of the most valuable opportunities begin with a research or data step precisely because that is where the real work lies. In our last full programme, the themes that attracted the most ideas were frequently the least ready, because the data behind them was missing or siloed; the themes the business could act on immediately were those where the data was already in hand.
Two consequences run through every roadmap we produce. First, the data-foundation work runs underneath the ambitious opportunities as a genuine precondition, which is why it is recommended as an early, cross-cutting investment rather than a by-product of any one use case. Second, the work is sequenced by readiness, not enthusiasm: data-in-hand opportunities proceed now, while the data-dependent ones enter their pathway first, so that effort is spent building the ground the ambitious use cases will later stand on.
The five data positions that decide sequencing
Every use case is assigned a data position in discovery, assessed from what participants report and confirmed by audit as the first scoping task: low dependency (runs on tooling, not on a body of data), in hand (exists, accessible, usable), siloed (exists but unjoined across systems or devices), partial (some strands have data, others do not) or absent (must be created or instrumented). The position decides sequencing, not feasibility, and it is the single most useful fact to know about any AI opportunity.
About the pathways
Can a use case follow two pathways?
In sequence, yes. A data-foundation step often precedes a product-led selection: build the master data, then choose the forecasting tool that will use it. The roadmap shows both steps and their order.
Does product-led mean you resell products?
No. We run the market scan and selection independently and hold no reseller margin. Where a build is the better answer we say so, and where a product is, we say that too.
How long does a research-led pathway take?
Typically one to two quarters to a commit-or-close decision, with engineering time ring-fenced and a concept demonstrator as the output. The point is to make the decision on evidence rather than to fund an open-ended programme.
Find out which pathway your ideas are on
Discovery assigns one to every use case, with the data position behind it.