LTV-adjusted ROAS: bidding on predicted value

What is LTV-adjusted ROAS?

Standard ROAS divides the revenue a campaign is credited with by what it cost. In B2B that revenue is often a first invoice, a trial start or a fixed value per lead. It treats a small account that churns after one term the same as a large one that renews and expands.

LTV-adjusted ROAS replaces that first value with the predicted lifetime value of the customers the campaign brought in. The question changes from “what did this spend return this month?” to “what will the customers it won be worth?”

Why does first-sale ROAS mislead?

Three patterns show up again and again:

• Cheap wins look good. A keyword or audience that brings small, price-sensitive accounts can show a strong short-term ROAS and still lose money over a year.

• Slow buyers look bad. Larger accounts take longer to close, so short conversion windows credit them late or not at all.

• Flat lead values hide the difference. If every form fill is worth the same number, bidding has no reason to prefer the lead from a good-fit company.

A forward-looking value fixes the third problem directly, and helps with the first two because it rewards fit rather than speed.

How do ad platforms use values?

They bid on them. Google describes target ROAS as a strategy that predicts the value of a potential conversion and sets bids to maximize conversion value at the return you target.^[1] Those predictions come from the conversion values you report, and you must set values on the conversions you track before you can use it.^[1]

Meta works the same way. When the optimization goal is value, Ads Manager shows the bid strategy as highest value, and a ROAS goal sets the minimum return you want from spend.^[6] Meta warns that setting that floor too high can lead to under-delivery.^[6]

So the value you attach to each conversion is the instruction. A flat value tells the platform all leads are equal. A predicted value tells it which ones to pay more for.

What goes into a predicted value?

Start simple. A rules-based score that most teams can explain beats a model nobody trusts. Useful inputs for B2B:

• Firmographics: company size, industry, region and whether the account matches your ideal customer profile.

• Deal signals: the plan or product of interest, the number of seats or users requested and the source of the lead.

• Early behavior: for product-led businesses, activation and usage in the first days of a trial.

• History: average contract value, renewal rate and expansion for similar past customers.

Turn the score into a value per conversion, or into a few tiers with a value each. Agree the assumptions with finance, and use margin rather than revenue if margins differ by product.

How do you send values back?

From your CRM, because that is where the real outcome lives. Our guide to closing the loop from CRM to ad spend covers the plumbing for each platform.

• Google Ads. Enhanced conversions for leads is an upgraded version of offline conversion import: your site sends hashed lead details when a form is submitted, and you later import the same hashed data with the conversion and any click IDs.^[4] Google’s notice says that from June 15, 2026, these uploads move to the Data Manager API and are blocked in the Google Ads API, so new builds should use the Data Manager API.^[4] Google describes it as a unified API for sending audience and conversion data to its ad products.^[5]

• Meta. The Conversions API connects your marketing data, including offline events from a CRM, to Meta for targeting and measurement.^[7]

Send a value with each later stage, such as a qualified opportunity, not only with the first form fill.

When do conversion value rules help?

When you know a segment is worth more but cannot change the value at the source. Google’s conversion value rules adjust both reported values and the values Smart Bidding uses, and Google names lifetime value and margin differences as reasons to use them.^[2] Rules can use location, device and audience conditions, including first-party lists.^[2] Through the API, a multiply action must sit between 0.5 and 10.^[3]

They apply only to some campaign types, and Smart Bidding uses them with target ROAS and maximize conversion value.^[2] Check your campaign types against the current list before you plan around them.

How much data do you need?

More than many B2B accounts have at the deal stage. Google lists at least 15 conversions in the past 30 days for target ROAS on Search and Shopping, and higher thresholds for some other campaign types.^[1]

If closed deals are too rare, value an earlier stage that happens more often, such as a sales-accepted lead, and set its value from the predicted worth of the account. Keep the closed-won data for checking the model rather than for bidding.

How do you know the model works?

Compare predicted with actual value by campaign and by tier, on a schedule that fits your sales cycle. If high-tier leads do not close at higher values, fix the scores before you raise bids.

Track a few numbers beyond blended ROAS:

1. Predicted value per acquired lead, by campaign.

2. Pipeline and closed revenue per unit of spend, by value tier.

3. Retention and expansion of acquired customers once enough time has passed.

Test before you switch. Run value bidding against your current strategy in a controlled experiment, and judge it on pipeline, not on the reported ROAS that the new values themselves inflate. For waste that has nothing to do with value, start with how to find wasted ad spend, and see B2B PPC campaign structure for how to split campaigns so value signals stay clean.

Our PPC service sets up this kind of value feed as part of our B2B demand generation work.