Opinion · Operating model · Organisation

Who owns the engagement layer — Marketing, Ops, or Product?

Once a campaign tool. Now a real-time decisioning layer.
The technology has moved on. The org chart has not.

Picture a single real-time decision. A high-value customer's behaviour shifts in a way that signals they are about to leave. The system reads the signal, scores the risk, and fires a retention offer before the customer has finished the thought. It is a good decision, made well, in the moment. Now ask a harder question than how it was made: who owned it? Marketing, because it is an offer? Operations, because it is a retention process? Product, because it happened inside the app? Risk, because it touches eligibility and price? In most enterprises nobody is quite sure. And that uncertainty, not the technology, is now the binding constraint.

How the engagement layer escaped Marketing

For most of its history this was simple. The engagement layer sent campaigns, and Marketing owned campaigns. The tools were marketing tools, the budget was a marketing budget, and the metrics were marketing metrics. That settlement has quietly collapsed. The same layer that once sent a newsletter now makes decisions that are operational (whether to intervene on a churn signal), financial (which offer, at what price, to whom), product (what the customer sees in the app at that instant) and regulatory (whether this customer may be contacted on this channel at all). The capability outgrew the function that owned it, and no one redrew the lines.

Gartner's research on the function shows how unsettled the question has become. Around nine in ten large organisations now have a Chief Experience Officer, Chief Customer Officer, or equivalent. Only one in ten of those officers reports to the CMO; most report to the CEO or COO. The role most often nominally responsible for the customer experience does not typically sit inside the function that executes most of it. The org chart has stopped being a settlement, and started being a contest.

The growth-versus-comms fault line

Even inside Marketing, the ownership is contested. The rise of growth as a discipline split the function in two. Growth roles think in experiments, funnels, real-time triggers and revenue, and reach for the engagement layer as an execution surface. Communications and brand roles think in narrative, consistency, and reputation, and reach for the same layer as a publishing surface. They are not the same job. They often report to different leaders. They frequently want the engagement layer to do incompatible things at the same customer in the same week.

The tool sits in the middle, and the org absorbs the conflict. A growth team triggers a clearance offer at lunchtime; a brand team is, that day, running an awareness campaign with a tone of voice that has nothing to do with discounts. Both fire at the same customer in the same session. From outside the building, neither is wrong. From inside, the only two parties whose interests were aligned were the customer and the tool.

What unclear ownership costs

When no one owns the engagement layer, the symptoms follow. Tools multiply, because each claimant buys its own rather than fight for control of a shared one, which is exactly how integration hell — the subject of the previous piece in this series — begins. Decisions collide at the customer: a retention offer and a risk hold and a product nudge all fire on the same person in the same session, because no single policy reconciles them. Accountability evaporates exactly as it does across vendors, except now it does so across departments. And the customer, who neither knows nor cares about the org chart, experiences the seams directly, as contradiction, repetition, or silence.

The technology stopped being the constraint. The org chart became one.

When the decision-maker is no longer human

And the question is about to get harder. Earlier pieces in this series have argued that the engagement layer is becoming a real-time decisioning layer, and that the next wave of decisions inside it will be made by autonomous agents rather than by humans clicking through a campaign tool. Agents do not respect the org chart. They will read a signal, score it against whatever policy is closest to hand, and act, on whatever channel is open, against whatever target someone configured weeks earlier. If platform, policy, and outcome are not separated cleanly before the agents arrive, the contradictions they produce will be faster, more frequent, and harder to unwind. The ownership question stops being a quarterly debate. It becomes structural.

Toward a workable model

The instinct to hand the whole thing to one function is a trap, because the claims are shared by design. A more durable model separates three things that are usually conflated: the platform, the policy, and the outcome. One core, common rules, distributed goals. The fight is only unwinnable while all three are treated as a single prize.

The question to put on the agenda

This is not a technology decision; it will not be solved by another procurement. It is an operating-model decision. And it belongs on the executive agenda as one explicit question, asked once and answered clearly: when a real-time decision is made about a customer, who owns the platform that made it, who owns the rules it followed, and who owns the result? An enterprise that can answer those three cleanly has resolved the binding constraint. One that cannot will keep buying capability it cannot govern, and the customer will keep feeling the seams the org chart drew.

What kind of platform deserves to sit at the centre of that ownership question is one this series will keep returning to.