Google Ads Product Value Optimization is a beta that lets retailers adjust the conversion value of specific products, or whole groups of products, so automated bidding pushes harder on the inventory that actually matters to the business. Search Engine Land reported the test on September 17, after Google Ads freelancer Adriaan Dekker spotted it live in an account. Google's own product page describes it as a way to “align your ad spend with true business value.” That is marketing language for something we have wanted for years: bidding that understands profit.
What Google has actually confirmed
Keeping strictly to the two sources above, here is the whole of it. You can create value adjustments for products or product attributes such as brand or category. You do not have to restructure campaigns to do it. Automated bidding then uses those adjusted values when optimising Performance Max and Shopping campaigns, and Google says this happens in real time. Google names three use cases explicitly: profit, seasonal sell-through, and best-sellers.
That is the confirmed set. Everything past that point is unconfirmed, so we are going to say so rather than fill the gap. There is no published eligibility list, no announced general availability date, and no Help Center documentation we could find at the time of writing. Google's product page carries a publish date of May 13 and was last updated on September 16, which is consistent with a beta quietly widening rather than a launch. If you want the feature, the honest answer today is to ask your Google rep whether your account is in the test.
Why this matters more than it sounds
Target ROAS has one blind spot that no amount of bid tuning fixes. It optimises the revenue you report, and revenue is not the number you keep. A store with a 60 percent gross margin hero product and a 15 percent margin accessory looks identical to Smart Bidding if both sell for the same price. So the algorithm does the rational thing with the information it has, and buys whichever converts more cheaply. Very often that is the cheap, low-margin, high-return-rate item. Your ROAS holds up. Your bank balance does not. We wrote about that gap in detail in why your ROAS looks fine but you are not profitable.
Until now, the workarounds were all structural or plumbing. You split campaigns by margin band and lost the pooled conversion data that automation needs. Or you sent profit as the conversion value through a custom feed and accepted the development work and the reporting confusion that comes with it. Product Value Optimization is the first version of this that sits on top of your existing structure, which is why it is worth paying attention to even while it is in beta.
The catch, and it is a real one
The output quality is entirely a function of the rules you write. Search Engine Land makes this point plainly and we agree with it: prioritising a product because it is a bestseller is not the same as prioritising it because it generates the most incremental profit. Those two lists overlap far less than most founders expect. Bestsellers are often bestsellers because they are cheap, heavily discounted, or the thing people buy once and never again.
Three ways we expect accounts to get this wrong. Uplifting products by revenue rank, which just tells the algorithm to do harder what it was already doing. Using gross margin without subtracting returns, shipping and payment fees, which flatters anything bulky or frequently sent back. And uplifting seasonal inventory then forgetting to remove the rule in January, which leaves bids inflated on stock you no longer care about.
There is also a measurement consequence worth planning for. If you adjust the values Smart Bidding optimises against, the conversion value in your Google Ads reporting stops being your actual revenue. Your reported ROAS becomes an internal index, not a number you can hand to a finance team or compare against last quarter. That is a fine trade to make deliberately. It is a painful one to discover three weeks later.
What to do this week
Nothing here requires access to the beta, which is the point. The prep work is the hard part, and it is useful whether or not Google ever turns this on for you.
Build a true contribution margin per SKU. Not gross margin. Selling price minus cost of goods, minus payment processing, minus shipping and packaging, minus a return rate allowance for that specific product. Export it next to your top 50 SKUs by spend. Most stores we audit cannot produce this table on request, and until it exists, every profit-based bidding conversation is guesswork.
Rank by profit contribution, not revenue. Multiply that margin by units sold over the last 90 days. Then compare the top ten against your top ten by revenue. The difference between those two lists is what this feature is actually for, and looking at it usually changes someone's mind about where the budget goes.
Group into three or four bands, not fifty rules. Value adjustments apply to attributes such as brand and category, so decide now which product grouping in your feed maps to margin. If your custom labels are empty or hold something meaningless, fix them in the feed this week. That is the work that makes the rules possible later.
Snapshot your baseline before you touch anything. Last 60 days of revenue, orders, average order value and blended margin per campaign. Once adjusted values are running, your in-platform ROAS is no longer comparable to history, so you need the before picture recorded outside the platform.
Ask your rep, then wait. If you get access, change one thing. One band, one adjustment, one campaign, and leave it alone for a full purchase cycle plus the bidding relearn. A beta plus three simultaneous changes produces a result nobody can read.
The honest summary
This is a good feature pointed at the right problem, and it will reward the small number of ecommerce accounts that know their real margins by SKU. For everyone else it is a faster way to tell Google's automation the wrong thing with more confidence. The bottleneck was never Google's willingness to optimise for profit. It was that almost nobody can tell it what profit is.
Working out that number is step one of our free 15-point Profit Audit: contribution margin by product, what your reported ROAS is hiding, feed and custom label hygiene, and whether your conversion values survive contact with your bank account. Thirty minutes, and you keep the findings either way.