Google Ads Product Value Optimization: Bid for Margin, Not Just Revenue

Your ROAS can look great while your margin quietly bleeds

Here's a scene most D2C founders recognise. The dashboard says 4x ROAS, the team is happy, and yet the bank balance isn't growing the way the numbers promise. When you break it down, the culprit is usually the same: your best-selling product is a low-margin entry item, Smart Bidding has fallen in love with it because it converts cheaply, and it's soaking up budget that your genuinely profitable SKUs never see.

That's the exact problem Google is now trying to solve at the bidding layer. On Sep 14, 2026, Google announced Product Value Optimization, a beta feature that lets you tell automated bidding that not every conversion is worth the same — even inside the same campaign.

If you've ever wanted to say "sell more of this, less of that" without slicing your account into a dozen fragile campaigns, this is aimed squarely at you. It's early, it's a beta, and it rewards preparation, so let's be precise about what it does and how to use it without wrecking what's already working.

What Product Value Optimization actually does

Automated bidding has always treated a conversion as a conversion. A sale worth ₹800 and a sale worth ₹800 look identical to the algorithm, even if the first one nets you ₹90 and the second nets you ₹450. The system optimises toward the number it can see — revenue — because that's what you fed it.

Product Value Optimization adds a second layer. Instead of only reacting to sale price, it lets you assign a value multiplier to products based on attributes you choose — brand, category, or a custom label like "high-margin" or "core-SKU." You're effectively telling Google: treat a conversion on this product as worth 1.3x what its revenue implies, because it's worth more to my business.

Two things make this practical rather than theoretical. First, it works with Performance Max and Shopping campaigns without a restructure — you're not rebuilding anything. Second, it overlays your existing bidding strategy rather than replacing it. Your target ROAS or Maximize Conversion Value setup stays exactly as it is; the multipliers just re-weight where the algorithm leans.

Revenue bidding vs value bidding, side by side

The shift is easier to see in a table. Imagine two products in the same Shopping campaign:

Under plain revenue bidding, the entry serum wins the budget tug-of-war because it converts easily and cheaply. Under value bidding, you hand the moisturiser a multiplier that reflects its real contribution, and the algorithm starts fighting harder for the sale that actually pays your bills.

The point isn't to starve your cheap products. Entry items do real work bringing new customers in. The point is to stop pretending a ₹90-margin sale and a ₹540-margin sale deserve the same bid.

Setting it up without breaking what works

Because this is a beta and it touches bidding, restraint beats enthusiasm. A sensible sequence:

1. Get your margin data honest first

Before you touch Google Ads, pull real contribution margin by product — after COGS, shipping, payment fees, and your realistic return or COD-cancellation rate. This is the whole foundation. If your margins are a spreadsheet full of assumptions, fix that before you let an algorithm act on them.

2. Group products with custom labels in Merchant Center

Create clear custom labels — "high-margin," "core-sku," "clearance" — so you have clean segments to apply values against. Vague grouping produces vague bidding.

3. Request beta access, then start conservative

Enroll through your Google rep. When it's live, resist the urge to set dramatic multipliers on day one. Keep 1.0x as your baseline and start hero products around 1.2x. A modest gap lets you see whether the system responds sensibly before you widen it.

4. Change one thing at a time

Don't launch value optimization the same week you change budgets, creative, and targets. If everything moves at once and results shift, you won't know what caused it. Isolate the variable.

5. Give it a learning window

Let it run at least two weeks before judging. Smart Bidding recalibrates when you introduce new signals, and the first few days will look noisy. Read the trend, not the daily wobble.

Where Indian D2C brands will feel it most

This lands differently depending on your catalogue shape.

If you sell a wide range with mixed margins — say a beauty or supplements brand with cheap trial sizes and premium hero products — this is close to made for you. Your biggest leak is usually budget flowing to trial SKUs that acquire customers but barely profit. Value multipliers let you keep acquiring while nudging spend toward the products that fund the business.

If you run COD-heavy categories, bake your cancellation reality into the margin number you feed the system. A product with a headline 40% margin and a 25% RTO rate is not a 40% margin product. Value optimization is only as smart as the contribution figure behind it.

If you're a single-hero-product brand, the feature does less for you — there's no meaningful value gap to express. Your gains still live in creative, offers, and landing pages, not in re-weighting a catalogue of one.

Getting the margin model and label structure right before you switch this on is fiddly, ongoing work, and it's the difference between a feature that prints profit and one that quietly optimises for the wrong thing. If you'd rather have that built and monitored properly, it's the core of what our growth marketing team does for D2C accounts.

A quick worked example

Say you run a skincare brand with three tiers. A ₹500 cleanser at 12% margin, a ₹900 serum at 30%, and a ₹1,500 regimen kit at 48%. Under plain Maximize Conversion Value, Google chases the cleanser hardest — it converts cheapest and the algorithm sees three equal-looking "value" events. Your ROAS report looks healthy at 4.2x, but your contribution margin is thin because the cheap cleanser dominates your sales mix.

Now you apply value multipliers that reflect reality: cleanser at 1.0x, serum at 1.3x, kit at 1.6x. You haven't touched your target ROAS or your budget. Over the next two weeks, the algorithm starts bidding harder in auctions where the serum and kit can win, and eases off slightly on cleanser-only queries. Your headline ROAS might even dip a touch — say to 3.9x — while your actual profit per rupee spent climbs, because the sales mix shifted toward products that pay.

That last point is the one to internalise: with value bidding, a slightly lower ROAS number can mean a healthier business. You're no longer optimising the metric on the screen; you're optimising the money in the bank. Judge the change on contribution margin, not the ROAS headline.

The mistakes that will burn your margin

A few traps are worth naming out loud.

Setting multipliers off gut feel instead of margin data teaches the algorithm your biases, not your economics. Cranking a hero product to 2x on day one can pull so hard that your cheaper acquisition SKUs starve and new-customer volume drops. Forgetting to update values when margins change — a supplier price hike, a discount season — leaves the system optimising toward last quarter's reality. And treating this as "set and forget" misses the point: it's a steering input you revisit, not a switch you flip once.

What to watch after two weeks

Once it's had time to settle, check three things. Is budget actually shifting toward your high-value products, visible in the impression and spend distribution across SKUs? Is your blended contribution margin improving, not just your ROAS number? And is new-customer acquisition holding, or did you over-weight profit so hard that the top of your funnel dried up?

If all three look right, widen the gap between your baseline and hero multipliers a little and watch again. If acquisition dipped, ease the multipliers back. The goal is a portfolio that grows and pays — revenue that means something once the costs are out.

Bidding tools have spent years getting better at chasing conversions. This is the first time the machinery is being pointed at the question that actually matters to a founder: not how many sales, but which ones were worth having.