What is value-based bidding?

Value-based bidding is a subset of Google Ads Smart Bidding strategies. Instead of optimizing for the highest possible number of conversions, the way Maximize Conversions or Target CPA do, it optimizes for the highest possible value of those conversions, within a budget or a target return on ad spend (ROAS).

The advertiser assigns a value to each conversion, either dynamic (the actual amount of each sale, for example) or static and different by conversion type (a booked demo is worth more than a downloaded PDF, a qualified lead more than a generic contact form). Google needs at least two distinct conversion values for the system to have anything to distinguish; if every conversion is worth the same, value-based bidding has no new information to optimize on and behaves like a volume-based strategy.

There are two ways to apply it: Maximize Conversion Value, with no efficiency limit, and Target ROAS, which also requires maintaining a minimum return per euro spent. Both are available for Search, Shopping, Video, and Demand Gen campaigns, though the minimum conversion volume requirements vary by campaign type. The dynamic value is calculated from the actual amount of each conversion, while the static value is a fixed number the advertiser sets once per conversion type and that stays the same without a manual update.

The bidding system needs reliable conversion values before it can optimize anything. That means setting up conversion tracking to record not just that a conversion happened, but how much it is worth: the order amount, a fixed value per action type, or a combination of both, depending on what fits the business.

With those values, the bidding model adjusts the bid in each auction not just by conversion probability alone, but by conversion probability multiplied by the expected value of that conversion. A search with lower conversion probability but higher expected value can receive a higher bid than one with higher probability but lower value, a distinction a purely volume-based strategy cannot make.

Google Ads recommends having at least fifteen conversions in the last thirty days before setting a numeric Target ROAS, and keeping the delay between the click and the recorded conversion short, ideally under seven days, because the model learns worse when value data arrives weeks after the triggering click. It also recommends feeding conversion data to Google as soon as it is available, rather than batching uploads.

Before turning the strategy on, Google Ads asks for conversion tracking that has already been running for about three weeks or one to two full conversion cycles, so the system has a history of values to start optimizing on instead of starting from nothing. The adjustment happens in each individual auction, not once a day, so a change in the declared values shows up in bids almost immediately.

In a B2B account, not every lead is worth the same. A demo booked by a hundred-person company is not equivalent to a form filled out by a student researching a school project, even though both count equally as a conversion if the bidding system only looks at the count. A volume-based strategy optimizes to get more of both equally; a value-based strategy can learn to prioritize the first one if the advertiser tells it, through a numeric value, that it is worth more.

That is the difference between counting leads and weighting them: moving from a quantity metric to a quality one requires the business to first decide what makes one lead worth more than another, and translate that into a number the bidding system can use. Without that groundwork, defining values by conversion type or connecting the CRM to pass the real sale value once the deal closes, value-based bidding has no advantage over optimizing by volume: it needs quality first-party data about the conversion, not just the fact that it happened, and that data does not exist unless someone defined it.

Without that prioritization, budget tends to spread evenly across leads of different worth, which in practice penalizes campaigns aimed at larger accounts compared to ones that attract high volume at low value.

• Define the value of each conversion type before turning the strategy on, not after: without at least two distinct values, the system has nothing to distinguish.
• Connect the real sale value from the CRM when the sales cycle is long, instead of leaving a fixed value assigned at the moment of the lead.
• Check that the delay between click and conversion stays short; if the sales cycle runs for months, consider valuing an intermediate conversion, like a qualified demo, on its own instead of waiting for the close.
• Accumulate at least fifteen conversions in thirty days before setting a numeric Target ROAS, as Google Ads recommends.
• Feed value data to Google as soon as it is available, not in weekly or monthly batches.
• Keep conversion tracking stable for about three weeks before turning the strategy on, so the system starts with history instead of from zero.

• Assigning the same value to every conversion and expecting value-based bidding to behave differently from volume-based bidding.
• Setting a Target ROAS before reaching the recommended minimum number of conversions, which leaves the system optimizing on too little information.
• Letting weeks pass between the click and the recorded conversion value, in a business whose sales cycle is already long.
• Turning on the strategy without having connected any first-party value data first: the system inherits the underlying problem instead of solving it.
• Changing the value criteria in the middle of the system's learning period, which resets the optimization process.