A marketing strategist is interested in target ROAS bidding: the honest 2026 brief

Why a marketing strategist is interested in target ROAS bidding#

A marketing strategist is interested in target ROAS bidding for one plain reason: control over revenue rather than raw volume. Target ROAS sits at the value end of Google's Smart Bidding family, and Google states plainly that Smart Bidding sets bids with machine learning at the moment of each auction. Rather than treating every conversion as equal quarry, the algorithm predicts the value hiding in each auction and pounces where the scent of revenue runs strong, holding back where it runs thin.

The ROAS target calculation is plain arithmetic: conversion value divided by ad cost, multiplied by 100. Google's own worked example uses a shoe shop that wants $5 of sales for every $1 of spend. A 500% target means $5 of revenue per $1 of spend, because $5 divided by $1, times 100, is 500%. Set that target and the system keeps total conversion value close to five times total cost: some sales higher, some sales sagging, the average held to the number a marketing strategist actually named.

Since June 2026 the strategy stands on its own paws again. Google is relabelling Maximise conversion value with a Target ROAS as simply Target ROAS, restored as a standalone strategy in June 2026 alongside Target CPA (checked 11 August 2026); the bidding behaviour is unchanged, only the label moved. Whenever a marketing strategist is interested in target ROAS bidding this early, the label change is worth a moment of relief: nothing about how the algorithm bids has actually shifted. If you are mapping the whole toolkit, start at the folkfox Google Ads hub and follow the trail outward.

Target ROAS requirements by campaign type#

Target ROAS requirements differ sharply by campaign type, and they are eligibility gates, not gentle guidance. For Search campaigns and Shopping, Google asks for at least 15 conversions in the past 30 days at the conversion tracking level. Display carries no strict history requirement: new Display campaigns no longer need a conversion track record to adopt the strategy (checked 11 August 2026).

Target ROAS app campaigns run to a stricter drumbeat: Google asks for about 10 conversions every day, or 300 conversions in 30 days, and recommends running a Target CPA campaign first to establish a baseline before value bidding begins. If installs and in-app revenue are your quarry, our app marketing work lives on exactly this maths, and it is the campaign type where a marketing strategist is interested in target ROAS bidding earliest, because App gates it hardest.

Why does the practitioner floor sit so far above the printed gate? Because conversion values fluctuate far more than conversion counts. A fortnight of fat baskets followed by a week of small sales swings the measured return wildly, so seasoned buyers treat 50 or more conversions in the last 30 days as the working floor for value bidding, not 15. Below 50, the algorithm is guessing from noise; at 50 or more conversions a month it has a steady signal to stalk.

The value data the target feeds on#

The target is only as honest as the values beneath it. Before the strategy can be applied, every tracked conversion needs a value, and Google's value-based bidding guidance is blunt: you should be reporting two or more unique values across your conversion goals. One flat figure is not value data; it is a conversion count wearing a costume, and it starves the engine of the very signal a marketing strategist is interested in target ROAS bidding to read in the first place.

The classic corruption is the static placeholder. Passing a static arbitrary value such as $100 per lead teaches the algorithm that every lead is identical, so it cheerfully buys the cheapest leads it can find; agency audits of value-fed accounts report CPA inflated by up to 40% once volume starts wearing the costume of value. The remedy is dynamic or calculated values: automated data transfer matters most when values are dynamic rather than static, and transaction-specific revenue beats any estimate.

Conversion value rules feed the same value stream the target is measured against. Google states that conversion value rules are used in real time by Smart Bidding to steer Target ROAS, and that value rules can lift values for higher-value customers, devices or locations. Adjust a rule and you have adjusted the very ruler the target reads, so document every change and expect a fresh learning wobble. That evidence-first discipline is the same one behind our SEO and GEO services: measure honestly, then set targets, never the other way round.

Device adjustments at minus 100 only#

Under target ROAS bidding, most manual levers are politely ignored. Location, ad schedule, audience and demographic bid adjustments are simply not used by the algorithm; Google's bid adjustment documentation shows device as the single exception, honoured at minus 100% only. You can remove a device from the hunt entirely, but you cannot nudge it, whisper to it, or tempt it back with a smaller trim.

The Target ROAS help page says the quiet part plainly: existing adjustments are not used, there is one exception, and you do not need to delete the old settings because they will not be read. Treat that as a tidy-up prompt anyway. An account littered with legacy adjustments cannot mislead the machine, but it will mislead the next marketer who prowls through the den.

From 17 August 2026, budget-limited campaigns obey the target#

From 17 August 2026, budget-limited campaigns using target ROAS or Target CPA are pulled toward the stated target rather than being left to overachieve it (checked 11 August 2026). Google's change notice is explicit that such campaigns will more consistently perform toward your bid target. If your account has been banking a 700% actual return against a 500% stated target, the free lunch ends and delivery drifts to the number you wrote down.

The proportional-adjustment formula (ROAS-Target ROAS)/10 is a concrete practitioner tactic for the 17 August 2026 change.

The defensive move is to make every stated target match trailing reality before the deadline. The Bid Target Adjustment Tool is now live inside Google Ads (checked 11 August 2026), and Google will not adjust targets automatically, so review each budget-limited campaign, track its recent actuals, and rewrite conservative targets by hand. If that sweep sounds like a chore, it is precisely the housekeeping our PPC management team performs weekly for every account where a marketing strategist is interested in target ROAS bidding well enough to keep it honest.

How a marketing strategist actually runs the workflow#

Reading the rules is one thing; running the workflow is another. In practice, a marketing strategist is interested in target ROAS bidding not as a single switch but as a small, repeatable routine: count conversions honestly, audit the value feed, calculate the real trailing return, set the target near that reality, then leave the account alone while the learning period settles. Skip a step and the target becomes a guess dressed up as a strategy.

The static value trap from earlier in this piece and the August 2026 target change are not separate problems; they compound. A padded target fed by flat values used to be forgiven because budget-limited campaigns quietly overachieved it. From 17 August 2026 that forgiveness is gone, so a marketing strategist is interested in target ROAS bidding audits that fix both weaknesses at once, not one at a time. Fixing the value feed without resetting a stale target still leaves cost drifting upward; resetting the target without fixing the value feed just teaches the algorithm to chase a cleaner-looking fiction.

A marketing strategist is interested in target ROAS bidding partly because it scales: one honest target, applied to a well-fed algorithm, replaces dozens of manual bid tweaks. But scale cuts both ways. A stale target or a static value does not fail quietly in one ad group; it fails across the whole campaign at once, which is why the routine above is worth running on a calendar, not just when a client asks why costs crept up. Treat the review as seasonal housekeeping: a den swept clean before winter serves better than one swept mid-storm.

Reporting is where the whole routine either earns its keep or quietly falls apart. A marketing strategist is interested in target ROAS bidding numbers that a client can actually trust, and trust starts with naming the basis of every figure on the page: which window, which currency, which conversion actions are primary rather than secondary, and whether the return quoted is the platform's own attribution or a blended, cross-channel view. Two honest reports built on different bases will disagree, and that disagreement is not a bug in the maths, it is a mismatch in the definitions feeding it. Write the basis down once, in plain words, and reuse the same sentence every time the report goes out.

The last habit worth building is patience with the calendar itself. Google's own guidance keeps returning to the same two to four week window for the learning period to settle, and every stacked change (a new target, a new value rule, a switched conversion action) resets that clock. A marketing strategist is interested in target ROAS bidding results that arrive fast, but the honest version of speed here is discipline: change one variable, wait out the window, read the result, then decide the next move. Anything faster is not optimisation, it is noise dressed up as a decision, and the algorithm pays for that noise in wasted spend long after the marketer has moved on to the next tab.

Frequently asked questions#