The model below was written in 2016 for UK retail boards. It is republished as written; only the closing note is new.
Stage 1: Brilliant Basics. Platform, product data, mobile, social proof.
Stage 2: Cultural Change. Breaking the store and online silo. Shifting measurement.
Stage 3: Inspiration. Full-price model, organic-first, lifestyle content.
Stage 4: Channel Agnostic. Single customer view, cross-channel commerce.
Stage 5: Exploiting Technology. Clienteling, AI personalisation, AR.
A note from 2026
Ten years on, the destination stage has a budget line. 97% of retailers are increasing AI investment; 6% are generating enterprise-level value (McKinsey, July 2026). The gap is the model’s original claim: the stages are a sequence, not a menu.
AI personalisation was named as the destination in 2016 not because it was fashionable, it was barely conceivable, but because everything before it is what makes it work: product data (stage 1), measurement culture (stage 2), a proposition worth personalising (stage 3), and a single view of the customer (stage 4). A decade later the failure mode is unchanged: buying stage 5 while standing on stage 1.
Two things the model could not see. Agentic commerce has moved discovery upstream of the website entirely, which promotes stage 1’s least glamorous line, product data, from hygiene to strategy: it is now your brand’s voice in conversations you are not in the room for. And the boundary between stages 4 and 5 has collapsed, because the single customer view and the intelligence acting on it are becoming one capability.
The sequence still holds. The businesses generating value from AI in 2026 are the ones that cleared the first four stages, whether they used this map or not.
Back to Thinking