If your B2B SaaS campaigns run on Target CPA, Target ROAS, or Maximize Conversions, almost every manual bid adjustment in your settings is doing nothing. Per Google’s “About bid adjustments” documentation, manual bid adjustments are not supported on Smart Bidding strategies — with a single exception, device. Location, ad-schedule, audience, and demographic modifiers you set on a Smart Bidding campaign are ignored, because the algorithm already prices those signals into every auction. The modifiers are a holdover from the manual-CPC era, and leaving them in place is account clutter that creates false confidence.
This post separates the one adjustment that still functions from the ones that are placebo, explains what a device adjustment actually does under Target CPA (it shifts the target, not the bid), and lays out the controls you should use instead to get real influence over delivery. The goal is a settings surface that reflects what is genuinely steering your spend — which is also the first thing any competent account audit looks for.
What still works: device, and only device
Device is the lone bid adjustment Google still honours under the main Smart Bidding strategies, and even then it behaves differently than it did under manual CPC. Under Target CPA, a device bid adjustment modifies the value of your CPA target for that device rather than multiplying a bid. Setting mobile to -20% tells Smart Bidding to chase a 20% lower CPA on mobile, which deprioritises it; setting it to -100% removes the device entirely. That -100% exclusion is the one case where a device adjustment still works as a true hard stop.
For B2B SaaS, this is more than a technicality. Mobile traffic on high-intent software queries converts to demos and trials far less reliably than desktop, yet the raw click and form-fill volume can look healthy. A deliberate negative device target on mobile — backed by segment data showing worse SQL rates, not a hunch — is a legitimate lever. The mistake is treating every device number as a bid multiplier it no longer is, or layering device preferences on top of a bidding strategy that is already optimising per auction. Set device intentionally, from evidence, and leave the rest of the adjustment grid alone.
What is placebo: location, schedule, audience, demographics
Location, ad-schedule, audience, and demographic bid adjustments are ignored on Target CPA and Maximize Conversions campaigns. Google’s support table lists them as unsupported, and the reason is structural: Smart Bidding already sets a unique bid for every auction using the searcher’s physical location and location intent, the time of day and day of week, and the audience and demographic signals available. A manual overlay would double-count the very signals the algorithm is built to weigh. Google states it can optimise bids down to the city a person is located in, even when your own location targeting is coarser.
The practical harm is false confidence. A manager sees a -30% ad-schedule adjustment on weekends and believes weekend spend is throttled; under tCPA it is not, and spend flows wherever the model predicts conversions. The same goes for a +25% adjustment on a “high-value geo” that quietly does nothing. These modifiers tend to be sediment from a manual-CPC setup that was migrated to Smart Bidding without cleaning the settings. They are not a lever you are pulling — they are a number the report still displays while the algorithm ignores it.
What to use instead for real control
If you want to influence delivery on a Smart Bidding account, use the controls the algorithm actually reads. For short, known demand changes — a product launch, a conference week, a pricing announcement — use seasonality adjustments and data exclusions: seasonality adjustments tell bidding to expect a short conversion-rate spike, and data exclusions tell it to ignore a window where tracking broke. Both are temporary and purpose-built, unlike a standing bid modifier.
For where and when ads show, use targeting and exclusions rather than adjustments: location targeting and exclusions govern eligibility (they still work because they control who is eligible, not what you bid), and campaign ad scheduling sets active hours. For structural device, geographic, or segment preferences, split them into separate campaigns with their own targets so each gets a clean Smart Bidding model. And to steer bidding toward quality rather than volume, feed it better conversion data — values and offline imports — not a modifier. That is the difference between shaping the inputs Smart Bidding optimises on and decorating the settings page.
The manual-bidding exception, and why it rarely applies
Bid adjustments still function fully on the legacy manual and enhanced-CPC strategies, which is why the adjustment grid has not disappeared from the interface. If a campaign runs on manual CPC, a -40% mobile modifier really does cut your mobile bids by 40%. This is the context the old playbooks were written in, and it is why so many managers still reach for adjustments reflexively.
For a B2B SaaS account spending $10k+ per month, manual bidding is rarely the right home, so the exception rarely applies. The volume and the long sales cycle reward letting Smart Bidding price auctions against good conversion data far more than hand-tuning CPC bids ever could — the same reasoning behind choosing Maximize Conversions over manual CPC once you clear the conversion-volume floor. If you are deliberately on manual bidding for a thin-data campaign, your adjustments work; the moment you switch that campaign to tCPA or Max Conversions, every adjustment except device goes inert, and the settings should be cleaned to match.
Audit your adjustments: a 10-minute cleanup
Run this pass across every Smart Bidding campaign. First, open the device column and keep only the device adjustments you can justify with segment-level conversion-quality data; reset the rest to 0%. Second, clear every location, ad-schedule, and demographic bid adjustment on tCPA, tROAS, and Max Conversions campaigns — they are inert, and removing them stops them from misleading whoever reads the account next. Third, move any intent behind those deleted adjustments to the correct lever: a geo you wanted to downweight becomes a location exclusion or its own campaign; an hour range becomes an ad schedule; a device becomes a device target.
The payoff is not a performance jump — inert settings were never affecting performance. The payoff is an account whose settings mean what they say, so the next optimisation decision is made against reality instead of a decorative number. Stale bid adjustments are one of the most common findings in a B2B SaaS account audit precisely because they survive agency handoffs and strategy migrations untouched. Clearing them is cheap insurance against managing to a lever that is not connected to anything.
Where this fits in a clean account
The rule to remember: on Smart Bidding, device is the only bid adjustment that does anything, and under Target CPA it shifts your target rather than your bid. Everything else — location, schedule, audience, demographics — is controlled through targeting, exclusions, seasonality adjustments, separate campaigns, and better conversion data, not through the adjustment grid. If your settings still carry modifiers from a manual-CPC era, they are clutter that invites bad decisions.
If you want a second set of eyes on what in your account is actually steering spend, a Google Ads audit checks the settings surface against what Smart Bidding honours, and our Google Ads management service runs the account on the levers that work rather than the ones that merely display. To spot inert settings yourself in a few minutes, start with the free 10-point audit checklist.
