How do I optimize Google Ads to revenue, profit, and ROAS?

Optimize To Revenue First, Then Profit — In That Order

Google Ads only ever optimizes toward the number you give it, so the work of optimizing to revenue, profit, and ROAS is mostly the work of feeding it better numbers. There's a clear hierarchy, and most accounts are stuck on the bottom rung.

The bottom rung is counting conversions. Every purchase reports as "1", so Smart Bidding treats a $20 order and a $2,000 order as identical and pushes for volume — more orders, regardless of size. This is fine for lead gen but actively wrong for e-commerce, because it trains the algorithm to chase your cheapest, lowest-value products.

The next rung is revenue. You pass the actual order value with every purchase, so a $2,000 sale counts a hundred times more than a $20 one. Now bidding favours searches and products that produce more revenue, and you can switch from Maximize Conversions to a value-based strategy. This single change — dynamic conversion values instead of a fixed number — is the highest-leverage fix in most accounts.

The top rung is profit. Revenue still lies to you, because a $2,000 order on a 10% margin makes less money than a $400 order on 60%. The fix is to send profit — revenue minus cost of goods — as the conversion value, so the algorithm bids toward margin, not turnover. ROAS isn't a fourth thing to optimize; it's the lever you pull once values are flowing. You set a target ROAS and Google spends to hit it. Get the hierarchy right — counts, then revenue, then profit — and ROAS becomes a dial you can trust instead of a vanity number.

Step One: Send Real Order Values, Not Just Conversions

Before any bidding strategy can optimize to revenue, Google has to receive the revenue. The first job is dynamic conversion values: every purchase must report its actual dollar amount, not a flat placeholder. On Shopify, WooCommerce, or a custom Next.js store, this means the value is read from the order total and passed through the data layer to both the Google Ads purchase tag and GA4 — never hardcoded.

Verify it before trusting it. Open Google Ads, look at your Purchase conversion action, and confirm "Conversion value" is set to "Use different values for each conversion". Then check the numbers: total reported conversion value over the last 30 days should roughly match the revenue in your store's own reporting. If Google shows $40,000 and Shopify shows $90,000, your tracking is leaking and every bid decision is being made on bad data — fixing that comes before any optimization.

Decide what counts as the value. The cleanest default is order subtotal excluding tax and shipping, because tax isn't yours and shipping is usually a pass-through cost. Sending the full total including tax inflates your apparent ROAS and quietly biases bidding. Be consistent across every channel so the numbers reconcile.

The payoff is immediate even before you touch bid strategy: with real values flowing, your Google Ads and GA4 reports finally show revenue per campaign, per product, and per search term. You can see that one campaign drives 60% of clicks but 20% of revenue — the kind of insight that's invisible when every conversion is just a "1". Accurate values are the foundation; everything else in this article assumes they're in place and reconciling.

Step Two: Switch To Target ROAS And Set It Against Your Margin

Once real order values are flowing, the bid strategy that optimizes to ROAS is Target ROAS (tROAS): you tell Google the return you want from each ad dollar, and it raises bids on searches likely to produce high-value orders and lowers them on the rest. But tROAS only works with enough signal — value-based bidding needs a meaningful volume of conversions with values attached (a common rule of thumb is roughly 15+ purchases in the prior 30 days) before it bids reliably. Below that, start on Maximize Conversion Value with no target and add the target once data accumulates.

Setting the target is where most stores go wrong: they pick a round number like 400% because it sounds healthy. The correct anchor is your break-even ROAS, which equals 1 divided by your profit margin. At a 40% gross margin, break-even is 2.5x — below that you lose money on every sale before you've paid for anything else. At a 25% margin, break-even is 4x. Your tROAS target should sit comfortably above break-even, with room for shipping, returns, and overhead.

Don't over-constrain it. Set the target too aggressively and Google throttles spend to protect the ratio — your ROAS looks great while volume and total profit collapse. A 10x ROAS on $500 of spend makes less money than a 4x ROAS on $5,000. Optimize for total profit dollars, not the highest possible ratio.

Finally, segment. A single account-wide target forces high-margin and low-margin products to share one rule. Split campaigns by margin tier or product category so each carries its own realistic target, and the account stops subsidizing your worst SKUs with budget that should go to your best.

Step Three: Optimize To Profit By Feeding Google Margin, Not Revenue

To make Google bid on profit instead of revenue, send profit as the conversion value. This is the step that separates accounts that look profitable from accounts that are. Revenue-based ROAS treats a $1,000 order identically whether it carries 70% margin or 8% — so the algorithm happily pours budget into high-revenue, low-margin products that drain you.

The direct method is to pass profit (revenue minus cost of goods sold) as the conversion value in the data layer, rather than the order total. If your store knows each product's cost — and Shopify, WooCommerce, and most platforms can store it — you can compute order profit at checkout and send that number. Now a 70%-margin order reports far more value than an 8%-margin order of the same price, and tROAS automatically shifts spend toward what actually makes money. Your reported ROAS numbers shrink, which alarms people who don't understand the change, but they're finally honest: a 2.5x "profit ROAS" means you're genuinely doubling your money.

If full profit values are too complex to wire up, two approximations help. First, set per-product-category conversion values or use custom labels in your feed to group SKUs by margin tier, then run separate campaigns with margin-appropriate targets. Second, build a profit view in GA4 or a spreadsheet using exported COGS and judge campaigns there, even if bidding still runs on revenue.

Don't forget the costs Google never sees: returns, shipping subsidies, payment-processing fees, and discount codes all eat margin after the sale. A category with a 30% return rate needs its conversion values discounted accordingly, or you'll keep scaling a campaign that books revenue and refunds it a week later. Profit optimization means the whole P&L reaches the algorithm — not just the part that flatters it.

What Breaks Profit Optimization — And What To Watch

The fastest way to ruin revenue-and-profit bidding is broken or partial value tracking, so audit that first and continuously. If half your purchases report a value and half report "1", Google's model is corrupted and tROAS will make confidently wrong decisions. The same happens when a theme update, a checkout change, or a consent-banner setting silently drops the purchase tag — reconcile Google's reported conversion value against your store's real revenue every month, not once.

Watch for new-customer blindness. Optimizing purely to immediate ROAS or profit makes the algorithm avoid first-time buyers, who convert at lower rates than returning ones, and chase cheap repeat sales you'd have won anyway. If acquiring customers matters, layer in a new-customer value rule (Google supports a new-customer acquisition goal) so a first order is worth more to the bidder than its single-purchase profit suggests. Otherwise short-term ROAS looks fine while your customer base quietly stops growing.

Mind attribution and the learning phase. A data-driven attribution model spreads credit differently than last-click, so the same sale can show a different ROAS depending on the model — pick one and judge trends, not daily noise. After any major change (new target, profit values, restructured campaigns), Smart Bidding re-enters a learning phase of one to two weeks where results are volatile; don't react to it.

Finally, resist the single-metric trap. ROAS, revenue, and profit can disagree, and the honest scorecard reads all three together: total profit dollars first, profit ROAS to judge efficiency, revenue to confirm you're not shrinking the business to hit a ratio. An account optimized to one number in isolation almost always sacrifices another. If you want this wired up correctly — values reconciling, margins flowing, targets set against your real break-even — that's exactly the kind of build SearchPod does, on accounts you own and can take with you.

Related questions