A $40 first purchase and a $140 first purchase both look like a conversion in a mediocre Meta ads account. That is the problem. Purchase value optimisation tells Meta to pursue the second customer more aggressively, because revenue matters more than a tidy cost-per-purchase report.
For Shopify founders already spending serious money on Meta, this is not a clever setting to flick on and forget. It is a shift in what your ad account is trained to value. Done properly, it can increase average order value, improve revenue efficiency and stop your budget being funnelled towards the cheapest possible buyers. Done badly, it can make delivery unstable, favour discount-heavy orders and hide an underlying offer problem.
What purchase value optimisation actually changes
Most ecommerce accounts optimise for purchases. Meta then looks for people likely to complete a purchase at the lowest achievable cost. That can be useful when an account has limited conversion data, a low average order value, or a simple objective: get more customers through the door.
But a purchase is not a fixed unit of value. One buyer might order a single $35 item using a welcome code. Another buys a $220 bundle at full price. If Meta is only asked to find purchases, it has little reason to distinguish between them.
Purchase value optimisation uses the value passed with your Purchase event to find people likely to generate more revenue. In practical terms, you select value as the conversion outcome rather than volume. Meta’s delivery system then weighs the probability of conversion against the likely revenue from that conversion.
That sounds obvious. Most agencies still optimise for the easy metric because it produces prettier screenshots. Cheap purchases, low cost per acquisition and a stable-looking dashboard can coexist with flat revenue and declining profit. Founders do not pay staff, suppliers or freight bills with a cost-per-purchase metric.
When value optimisation is the right move
Value optimisation works best when there is meaningful variation in order values and enough purchase volume for Meta to learn. A Shopify brand selling a $90 hero product with almost every customer buying one unit may see little difference. A brand with bundles, subscriptions, multi-item carts, upsells or a broad product catalogue has far more to gain.
It is especially relevant when your current campaigns are attracting bargain hunters. If sales spike whenever a discount is offered, but average order value falls and full-price products struggle, optimising for purchase volume can amplify the wrong behaviour. Meta will find more of what you reward.
There is a trade-off. Purchase volume optimisation often produces more orders at a lower cost. Value optimisation may produce fewer orders while generating more revenue. Neither outcome is automatically better. The right answer depends on contribution margin, repeat purchase behaviour, stock position and cash flow.
A founder selling high-margin consumables may accept a lower first-order return because a second purchase is likely. A founder selling bulky, low-margin products cannot. This is why generic ROAS targets are lazy. Your account needs a commercial target, not an agency benchmark copied from another business.
Do not confuse revenue with profit
Meta can optimise for purchase value. It cannot magically understand gross margin unless your data structure gives it a useful signal. If the highest-value orders come from products with thin margins, free-shipping thresholds or costly returns, revenue optimisation can send more budget towards less profitable sales.
Before changing optimisation, review which products and bundles drive the highest contribution after product cost, shipping, payment fees, discounting and returns. If a $180 bundle creates less profit than a $110 full-price product, do not celebrate higher average order value without checking the maths.
For most founder-led brands, the practical answer is not to build an over-engineered attribution model. It is to know the few commercial numbers that matter and use them to set guardrails: minimum blended MER, acceptable new-customer acquisition cost, average discount rate and contribution margin by product group.
The data has to be clean before Meta can learn
Value optimisation is only as credible as the Purchase event you send to Meta. If values are missing, duplicated, recorded in the wrong currency or disconnected from actual completed orders, you are training the algorithm on rubbish.
Check that Shopify purchase events pass the correct order value and AUD currency. Confirm browser and server events are properly deduplicated, so a single order is not counted twice. Refunds will not always be reflected in real time, which is another reason to judge performance over a sensible window rather than react to yesterday’s result.
Your catalogue also needs scrutiny. Broken variants, incorrect pricing, unavailable products and landing pages that do not match the ad can distort performance. Meta does not fix a messy store. It simply spends money according to the signals it receives.
Then look at volume. There is no magic purchase threshold that applies to every account, but a campaign with a trickle of weekly purchases is not a good candidate for aggressive value-based optimisation. The system needs enough conversion feedback to identify patterns. If spend is below roughly $3,000 per month, your bigger problem may be insufficient learning volume, not campaign settings.
How to test purchase value optimisation without torching performance
Do not rebuild a functioning account because someone on LinkedIn says value optimisation is superior. Run a controlled test that gives you a real answer.
Start by defining the business question. For example: can a value-optimised prospecting campaign increase new-customer revenue and average order value without pushing blended acquisition cost beyond our margin limit? That is a useful question. Can it beat last week’s ROAS is not.
Keep the offer, creative quality, landing pages and attribution view as consistent as possible. Test against your existing purchase-volume approach with a meaningful budget and enough time to move beyond early volatility. If your account is seasonal, compare against a comparable trading period rather than a random seven-day window.
Watch revenue, average order value, new-customer mix, blended MER and contribution margin. Cost per purchase still matters, but it is a diagnostic metric, not the prize. A higher cost per purchase is acceptable if each acquired customer is worth materially more and the economics hold.
Campaign architecture still matters
Value optimisation does not rescue a bloated account. If you have six campaigns competing for the same audience, dozens of underfunded ad sets and creative that has been running since the last financial year, changing the conversion setting will not create growth.
Keep prospecting structure simple enough to concentrate data and budget. Let creative do the heavy lifting on audience qualification. Use distinct campaign roles where they are justified, such as new customer acquisition, retargeting and catalogue-led activity, but do not split campaigns merely to make reporting look sophisticated.
The creative strategy must also match the commercial goal. Ads that lead with a permanent discount will predictably attract price-sensitive shoppers. To lift purchase value, test bundles, comparison angles, product education, use cases, social proof from higher-value customers and offers that increase cart size without training buyers to wait for a sale.
The mistakes that make the numbers lie
The first mistake is judging success entirely inside Ads Manager. Meta attribution is directional, not a substitute for your Shopify revenue, cash position or actual customer behaviour. If platform ROAS rises while total store revenue stalls, investigate before declaring victory.
The second is using an unrealistic ROAS floor. A strict minimum return target can restrict delivery so heavily that Meta cannot find enough buyers. It may protect a report while starving the business of scale. Set targets from your margin and growth requirements, then review whether they are helping or constraining delivery.
The third is changing five variables at once. New creative, a new offer, new landing pages, broad targeting and value optimisation launched on the same day gives you no usable learning. When results move, you will not know why.
Finally, do not use value optimisation to avoid fixing a weak proposition. If customers do not understand why your product costs more, Meta cannot manufacture demand. Better campaign settings cannot compensate for poor creative, a slow mobile experience, vague product pages or an offer with no reason to act now.
Make Meta accountable to the sale, not the spreadsheet
Purchase value optimisation is valuable because it forces a harder conversation: are your ads producing the kind of revenue your business actually wants? Not just more transactions. Not a lower cost figure. Profitable orders from customers who strengthen the business.
For established Shopify brands, that question should shape every decision in the account. Give Meta clean value signals, build creative that earns higher-value carts, and measure the outcome against real commercial constraints. If the numbers cannot survive that level of scrutiny, they were never performance marketing results. They were reporting theatre.