Sky's VIP loyalty overhaul made Marketing Week this week. The headline framed it as an insights story: Customer research revealed people wanted valuable rewards, not generic perks. The implication: Sky listened, learned and responded.
That framing is generous.
The signals Sky eventually actioned were not hidden. They were sitting in behavioural data for years: open rates declining quarter on quarter, redemption rates flatlining, cohorts disengaging in predictable sequences. The insight was never missing but the capacity to act on it was.
This distinction matters because it changes where the fix belongs. If the problem is insight, you invest in research. If the problem is execution velocity, you rebuild how signals become actions. Sky's story is the second one dressed as the first.
Marketing teams rarely suffer from a shortage of data. They suffer from a surplus of data and a deficit of action. The signals that eventually drove Sky's overhaul were almost certainly visible in their CRM systems long before anyone commissioned a research project.
Cohort analysis would have shown which customer segments were disengaging. Behavioural triggers would have flagged when redemption patterns shifted. Churn propensity models would have identified at-risk subscribers before they cancelled. None of this requires advanced analytics - it requires someone looking at the data and having the operational capacity to respond.
The research that informed Sky's overhaul didn't uncover hidden truths. It gave internal stakeholders permission to act on truths that had been administratively inconvenient.
When a loyalty programme isn't working, the evidence accumulates in plain sight. What delays the response is the gap between seeing the signal and making a decision.
Most CRM teams review cohort performance monthly; some review it quarterly. By the time the insight reaches a brief, travels through creative development, passes compliance and enters the deployment queue, the customer it was meant to retain has already cancelled.
This is the execution velocity problem: the data exists, the segmentation exists, the triggers exist but the operational machinery between signal and action moves too slowly to matter. A customer who shows disengagement signals in week one receives a retention intervention in week six. By then, the decision is already made.
Sky's public admission that their loyalty programme was failing lands at the same time as many CMOs are being asked to prove personalisation ROI. The timing is instructive. Boards aren't questioning whether personalisation works but they are questioning whether marketing teams can execute it fast enough to justify the investment.
When loyalty programmes underperform, the instinct is to commission more research to understand the customer better, uncover hidden needs and develop new strategies. This instinct is expensive and often counterproductive.
Sky's research told them customers wanted rewards that felt valuable. This is not a revelation. It is a restatement of basic value exchange principles. The research didn't discover something new, it validated something obvious with enough rigour to justify organisational change.
The real constraint was never knowledge. It was the distance between knowing and doing. Sky's CRM systems must have contained years of behavioural evidence that generic perks were not working. The evidence was detailed, timestamped and segment-specific. It was also trapped in dashboards that described the past without triggering action in the present.
This is the gap where marketing value gets destroyed every week. Not in the insight. In the translation of insight to intervention. When that translation takes months, the insight becomes archaeology. Interesting, perhaps. Actionable, no longer.
The marketers winning in CRM and loyalty are not those with better insights. They are those who can collapse the time between signal and action. When a high-value customer shows early disengagement indicators, the intervention reaches them within days, not weeks. When a cohort's redemption behaviour shifts, the offer structure adapts before the next campaign cycle.
This is where AI agents create separation. They create separation by compressing the execution timeline rather than generating more insights. Behavioural signals that would previously wait for a monthly review now trigger immediate workflow adjustments. Cohort movements that would previously require manual analysis now surface automatically with recommended actions attached.
The shift isn't from human decision-making to automated decision-making. It's from batched decision-making to continuous decision-making. The marketer still sets the strategy, defines the guardrails and approves the framework. The system handles the velocity.
Sky's overhaul will likely improve their loyalty metrics. But the improvement will come from actioning signals faster, not from discovering signals that were previously invisible. The research provided organisational permission. The value will come from operational change.
Sky's story is useful precisely because it is honest about failure. A major brand admitted publicly that their loyalty programme wasn't delivering. That admission creates cover for every CRM leader sitting on similar evidence and struggling to get organisational traction.
The lesson isn't that research unlocks insight. The lesson is that insight without execution velocity is waste. The signals are already in your systems. The question is whether your operational model can process them fast enough to matter.