The context AI is missing is mostly not about the customer. Jochem van der Veer, Co-Founder and Chief Executive Officer of TheyDo, breaks customer journey context into five dimensions, and four of them describe the organization rather than the person it is serving. These are the rules it operates under, the outcome it wants from this interaction, the stage it considers the relationship to have reached, and the conditions of the moment.
Only one rule covers what a company knows about the individual, which is the dimension enterprises have spent a decade and a considerable budget building. “Most implementations get one or two of these,” van der Veer says. “All five is what makes the action appropriate rather than just automated.”
Fluency Disguises the Gap
Enterprises hold enormous volumes of customer information, distributed across systems, departments, and separate points in the journey. That volume allows a model to produce a plausible response without judging whether it fits the situation in front of it.
Failure looks like success, which is what makes it durable. A reply arrives: accurate, well-timed, articulate, and wrong, and nothing in the record indicates whether this person, at this moment, should be reassured or sold to. Van der Veer calls the outcome automation without judgment. An answer that reads well clears most internal reviews, leaving the customer as the only party positioned to notice it missed.
The Objective Was Never Written Down
The directional dimension catches organizations hardest, covering whether an interaction exists to reassure, resolve, retain, or cross-sell. “Without a clear direction, the right data still produces the wrong output,” van der Veer says.
No customer database has ever recorded what an interaction is for, and no amount of enrichment will surface it, since the objective was never the customer’s to supply. A system holding complete history and no stated purpose simply optimizes toward whatever it was last pointed at, usually revenue, and it does so in the middle of a service complaint.
Constraint behaves the same way, living in legal documents, consent registers, and capacity limits rather than in any profile, while temporal position sits in whatever lifecycle model the company maintains and rarely updates. The knowledge that would prevent the error was internal all along.
Recognizable Incidents Rather Than Abstract Risk
The consequences are familiar to every executive: retention offers sent while complaints remain unresolved, cross-sells that overlook eligibility or consent, identical treatment for new customers and long-standing customers at risk of churn, and teams optimizing individual interactions while the real friction lies elsewhere in the journey.
Each follows from automating the response before establishing whether it was the right response. Van der Veer notes that without customer perspective, the output comes back technically correct and contextually wrong.
One Shared Model Rather Than Five Partial Ones
Context has to be structured across the journey instead of reconstructed separately by each team, so product, service, marketing, operations, and data reason from a single model of the situation rather than assembling their own.
Journey management supplies that structure, connecting evidence, friction, business outcomes, ownership, and action in one frame, and van der Veer argues that AI should reason over the shared context rather than sitting as another interface above disconnected systems.
TheyDo positions itself as an Experience Context Platform on that basis. Traceability carries as much weight as accuracy here. A leader defending a recommendation to a regulator or a board needs to show which evidence, journey stage, objective, and constraints produced it, where each of those is an internal record rather than a customer attribute.
The readiness question most enterprises ask themselves points outward, toward what they hold about the customer. The useful version points inward. An organization that cannot state what a given interaction is for, under what constraints, at what stage, has not failed to gather enough data. It has failed to write down what it is doing, and AI executes that omission faithfully at scale. To learn more about journey context in AI-enabled customer operations, connect with Jochem van der Veer on LinkedIn or visit TheyDo.
