Warm audiences convert because the demand already existed. Harvesting it isn't growth, and the dashboard can't tell the difference.
Some budget lines survive on the strength of their reporting, and retargeting has the best reporting in the building. Audiences who visited, browsed, carted or bought, re-reached with precision, converting at multiples of any cold audience, at costs per acquisition the rest of the account cannot approach. Every dashboard frames it as the efficient core of the program, and every quarterly review protects it accordingly. The numbers are accurate. The interpretation is the problem, because the line is being read as acquisition, and the mechanics underneath it are loyalty mechanics, the maintenance of demand that already existed, priced and celebrated as the creation of demand that did not.
The distinction is not pedantic, it decides whether the spend grows the brand, and the brand science gives the question a firm frame. Growth, the Ehrenberg-Bass tradition has shown across categories and decades, comes overwhelmingly from penetration, recruiting more buyers, mostly light ones, rather than from extracting more from the engaged few. Double jeopardy makes the point structurally, smaller brands suffer twice, fewer buyers who are also slightly less loyal, and the escape is not deepening loyalty, it is widening the buyer base. Hold the retargeting line against that frame and the question writes itself. Who, exactly, is in those audiences?
By construction, the pools contain people who already moved toward the brand, visited the site, engaged the content, carted the product, bought before. They are the warmest fraction of the market, self-selected by demonstrated interest, and that self-selection is doing nearly all the work the dashboard attributes to the ads. These were, disproportionately, the people most likely to buy anyway, that is what the behaviours that put them in the pool revealed. The conversion rate is high because the demand pre-existed, the ad intercepted a journey already in motion, and attribution awarded the journey to the interception.
This is why the honest metric for the line is not conversion rate but incrementality, conversions that occurred because of the exposure, measured against the held-out counterfactual, and where rigorous holdout testing is run, the gap between attributed and incremental performance on warm audiences is usually the most sobering number in the account, with the size of the gap varying by brand, pool construction and category, which is exactly why it has to be measured rather than assumed in either direction. Attributed performance counts the harvest and implies the ad grew it, incrementality asks how much of the harvest would have arrived anyway, and for audiences selected on intent, the answer is often, most of it. The maths can be put plainly. A high conversion rate multiplied by low incrementality is a small number of genuinely created sales, purchased at auction prices, and reported as triumph.
Meanwhile the framing error compounds on the cost side. The brand is paying acquisition-grade media prices, real-time auction CPMs, to re-reach people whose acquisition already happened or was already underway. There is a name for marketing aimed at existing and near-existing customers, loyalty marketing, and it is a legitimate discipline with its own honest economics, CRM, owned channels, service, the inexpensive maintenance of a relationship that exists. Retargeting conducts that discipline through paid auctions, at cold-reach prices, against an audience that mostly did not need persuading, which is how a maintenance function ends up consuming an acquisition budget while reporting itself as the budget's best performer.
The cost is not the retargeting spend itself, it is the allocation it justifies. Every comparative review where warm-audience efficiency beats cold-audience cost sends budget downhill, toward the pools, away from the broad, light-buyer reach where penetration is actually built, and the drift is self-affirming, the more the account harvests, the better its short-term numbers look, the more harvesting earns at the next review. The supply side of the system, the demand creation that fills the pools in the first place, gets starved by the very arithmetic its output makes possible. Accounts can run this loop for a long time before the symptom surfaces where dashboards do not look, penetration flat, the pools refilling slower, the brand quietly living off demand banked in earlier years. Harvesting is not growth, and the confusion of the two is among the most expensive in the channel.
None of this argues for zero retargeting. Capture is a real job, some journeys genuinely benefit from a timely nudge, abandoned considerations, renewal moments, and a disciplined capture layer earns its modest place. The discipline is the point, the line sized to its incremental contribution, measured by holdout rather than attribution, priced against the loyalty alternatives, CRM and owned channels, that often do the same maintenance for a fraction of the auction cost. Run that audit and most retargeting lines shrink substantially, not to zero, to honesty.
What the recovered budget buys is the unglamorous engine the dashboard never flattered, broad reach into the light and non-buyers who constitute growth, carried by creative built for people with no existing intent, distinctive, native, attention-earning work attached to category entry points, the demand creation that makes every downstream number possible. Media creates the opportunity to see your ads. Creative determines whether they work. The market where they have the most working to do is precisely the one no retargeting pool contains, the buyers the brand has not met yet.
The reporting will look worse before it looks better, that should be said plainly. Cold reach converts slowly, its effects accrue in brand tracking and penetration data over quarters, and the first reviews after reallocation will miss the warm efficiency the account gave up. What arrives instead, with patience, is the only thing the warm pools could never produce, new buyers, the widening base that double jeopardy says is the entire game, and pools that refill because the brand is once again creating the demand it harvests. Loyalty marketing is worth doing, it is not worth doing at acquisition prices, and it is never worth mistaking for growth.