Highest Value
By the AdFlint research team · Last reviewed July 2026
Meta bid strategy that spends the budget seeking the greatest total purchase value rather than the greatest number of results.
Delivery skews toward people Meta predicts will spend more, so average order value tends to rise while result count falls. It is the value counterpart to Highest Volume and depends entirely on accurate value being sent with purchase events. The mistake is judging it on cost per purchase, because compared with a volume strategy on that single metric it will nearly always look worse.
Key takeaways
- Highest Value optimizes for total predicted purchase value across the ad set's delivery, skewing toward people likely to place larger orders, with no minimum efficiency floor the way ROAS Goal has.
- It depends on accurate purchase value through the pixel and Conversions API - without reliable value data it behaves much like Highest Volume, just noisier.
- Judge it on total revenue and average order value, never on cost per purchase, since compared on that single metric it will almost always look worse than a volume strategy by design.
- It fits stores with real spread in order value and a revenue-first goal; it is a poor fit for narrow retargeting audiences or accounts with uniform, low-variance order values.
In practice.
Highest Value is set at the ad set level as a bid strategy option and, unlike Highest Volume, it optimizes explicitly for the total dollar amount of predicted purchase value the budget can generate rather than the largest number of results. Mechanically, delivery skews toward people the system predicts will place larger orders or become higher-value purchasers, which means the auction is willing to pay more per individual result if the predicted value justifies it, and willing to pass on cheaper opportunities that look low-value. It carries no minimum ROAS threshold the way ROAS Goal does - it simply spends the full budget chasing the highest total value, with no floor on efficiency along the way.
It depends on the same input as ROAS Goal: accurate purchase value passed through the pixel and Conversions API, ideally with a connected product catalog. Without reliable value data, Highest Value has no meaningful signal to differentiate a $20 order from a $200 order, and it will not produce noticeably different delivery than Highest Volume. It also interacts with audience size more than Highest Volume does, because narrowing delivery toward predicted high spenders effectively shrinks the pool of people the ad set is willing to show to, which can be a real constraint on a small or already-narrow audience.
Highest Value matters for ecommerce accounts with a wide spread in order value where total revenue, not purchase count, is the actual business metric that matters - a store selling both a $15 accessory and a $300 core product benefits from a strategy that actively seeks the customers likely to buy the $300 item. It matters less, and is often a poor fit, for accounts with narrow or uniform order values, where there is little value variance for the algorithm to differentiate on, and matters least for lead-gen or non-ecommerce accounts with no purchase value signal at all.
The most common mistake is judging Highest Value on cost per purchase, the same single metric that fits Highest Volume, and concluding the strategy underperforms - compared head to head on that one number, Highest Value will nearly always look worse, since it is explicitly not optimizing for the cheapest or most numerous purchases. The fair comparison is total revenue and average order value over a comparable period. A second mistake is applying Highest Value to an account with thin or unreliable purchase value data, where the strategy cannot actually tell high-value opportunities apart from low-value ones and just produces noisier, less predictable delivery than Highest Volume would have. A third is running Highest Value on a narrow retargeting audience, where the already-limited pool gets squeezed further as the system chases only the highest-value subset within it, sometimes to the point of meaningful under-delivery.
In reporting, total purchase value and average order value are the metrics that show whether Highest Value is doing its job, not cost per purchase or purchase count, both of which are expected to look worse than a volume-oriented strategy on the same budget. Comparing average order value under Highest Value against the account's baseline average order value under Highest Volume is the cleanest single before-and-after check of whether the strategy is actually skewing toward higher spenders as intended.
Practically, Highest Value fits a specific situation: real spread in order value, reliable value tracking through the Conversions API, and a business goal of total revenue rather than purchase count or cost efficiency. Outside that situation, either Highest Volume for raw scale or ROAS Goal for a controlled efficiency floor is usually the better fit.
Comparing Highest Value against a Highest Volume baseline
Suppose your store sells items ranging from $25 to $250, and a Highest Volume ad set spending $2,000 a month produces 100 purchases averaging $60 each, for $6,000 in total value and a $20 cost per purchase.
You test Highest Value on an identical $2,000 budget. It produces only 70 purchases, a lower count, at a higher $28.57 cost per purchase - on that single metric it looks like a step backward. But average order value rises to $95, and total purchase value comes in at $6,650, about 11 percent higher than the Highest Volume ad set despite fewer transactions and a worse per-purchase cost. Judged on the metric the strategy is actually designed to move, total revenue, it outperformed; judged on cost per purchase alone, it would have been wrongly cut.
Highest Value compared with
The settings this gets confused with, and how to tell them apart.
Common questions.
Why does Highest Value show a worse cost per purchase than Highest Volume?
That is expected and by design - Highest Value is optimizing for total predicted purchase value, not the cheapest or most numerous purchases, so it will typically show fewer purchases at a higher individual cost while producing more total revenue, which is the metric that actually matters for this strategy.
Do I need a product catalog to use Highest Value?
It is strongly recommended, especially for stores with varied order sizes, because catalog data gives the system a much stronger signal for predicting which browsers are likely to become high-value purchasers than purchase events alone.
Is Highest Value the same as ROAS Goal?
No - Highest Value chases the largest total purchase value with no efficiency floor, while ROAS Goal bids toward a minimum blended return on ad spend and will hold back delivery if it cannot find purchasers likely to clear that ratio, making ROAS Goal the more efficiency-conscious of the two.
Why is my ad set under-delivering on a retargeting audience using Highest Value?
Highest Value narrows delivery to the subset of an audience predicted to place higher-value orders, and on an already-small retargeting audience that narrowing can shrink the eligible pool enough to cause meaningful under-delivery; Highest Volume or a more modest ROAS Goal is often a better fit for small audiences.
You should not need to know this to advertise.
AdFlint handles the settings for you, inside the Google and Meta accounts you already own.
Try AdFlint free