What is value-based bidding and how it can drive smarter marketing success

Most paid media campaigns are built to win conversions, not revenue. Your bidding algorithm treats a $50 lead the same as a $5,000 one, and it has no way to know the difference unless you tell it. The result is a growing gap between the numbers in your Google Ads account and the numbers that actually matter to your business.

Value-based bidding closes that gap. This article explains how VBB works, what you need to set it up properly, how to assign conversion values when you do not have perfect data, and what separates the brands that get real results from the ones that see the strategy fail. You will leave with a clear framework you can apply to your own campaigns.

What is value-based bidding and how does it work

Value-based bidding is an automated bidding strategy that optimizes ad spend based on the expected value of each conversion. You define what each customer action is worth to your business. The platform's machine learning then bids more aggressively in auctions likely to produce higher-value outcomes, and holds back in auctions likely to produce low-value ones.

Traditional bidding strategies like target CPA or maximize conversions treat all conversions as equal. A newsletter signup and a $10,000 enterprise deal submission both count as one conversion. The algorithm cannot differentiate between them unless you supply the value data that makes that distinction possible.

VBB changes this by passing a monetary or proxy value alongside each conversion event. Google Ads evaluates thousands of real-time signals during every auction, including device type, location, time of day, search query, user history, and audience segment. It uses your value data to decide exactly how much to bid for each specific impression.

The two main VBB strategies in Google Ads

Google Ads offers two Smart Bidding strategies built for value-based optimization.

Google recommends starting any new VBB setup with a tROAS target set 20% below your historical ROAS average. This gives the algorithm room to learn before you tighten the constraints.

How the machine learning feedback loop works

1. A user clicks your ad and lands on your site.
2. Google assigns them a Google Click ID (GCLID), an anonymous identifier tied to that click.
3. When a conversion happens, you pass the conversion value back to Google along with the GCLID.
4. Google's AI maps the value to the signals present at auction time and updates its model.
5. Future bids adjust based on which signal combinations tend to produce higher-value outcomes.

The more accurate and frequent your value signals, the better the algorithm performs. Passing conversion data daily is best practice. Stale or batched uploads slow down the learning cycle, particularly in fast-moving industries.

Why conversion volume alone is no longer enough

Optimizing for conversion volume without value data means your algorithm may be filling your pipeline with the wrong customers. It cannot distinguish between a one-time buyer and a repeat enterprise client unless you give it the data to do so. As Jess Weber, Director of Account Performance at HawkSEM, puts it: "Conversions can come in, but conversion volume means nothing if they aren't going to become customers."

This is not a theoretical problem. In 2025, a benchmark study analyzing $996 million in Google ad spend across 100 consumer brands found that maximize conversions and target CPA strategies together accounted for roughly 43% of all spend, while value-based approaches including target ROAS and maximize conversion value accounted for about 48% combined. The brands optimizing for value are already the majority, but many are doing it without clean data, which undermines the results.

"Most advertisers come to us optimizing for lead volume. The first thing we do is connect what a lead is actually worth to the business, because that single change transforms how the algorithm behaves."

Tanner Medina, Co-Founder and Chief Growth Officer, Launchcodex

The profit gap that volume metrics create

Consider a B2B SaaS company running target CPA campaigns. The algorithm learns to produce form submissions at $120 each. What it does not know is that some of those submissions come from enterprise prospects worth $40,000 in annual contract value, while others come from small businesses that churn after 90 days.

By treating these leads equally, the campaign spends the same to acquire both. The sales team wastes time on low-quality leads. The business absorbs inefficiency at every stage.

Optmyzr's value-based bidding guide illustrates the exact cost of this gap. For a B2B company with an average order value of $3,000 and a 45% profit margin, the standard conversion value without lifetime value modeling is $270 per lead, based on a 20% lead-to-close rate. Add lifetime value modeling, where customers spend an average of $5,000 more over their relationship, and that same conversion is worth $720 to the algorithm. Nearly three times higher. That gap directly determines how aggressively the algorithm bids for similar prospects.

The first-party data urgency

Research from Madgicx found that 71% of marketers are actively expanding their first-party datasets, yet 66% expect reduced personalization ability due to tightening privacy restrictions. First-party data is the foundation of effective VBB. Without it, the algorithm cannot differentiate high-value prospects from low-value ones. Building that data infrastructure is the prerequisite, not an optional upgrade.

How to assign conversion values when you do not have perfect data

You do not need a clean CRM or complete revenue data to start value-based bidding. Most advertisers should begin with proxy values, which are reasonable estimates based on historical conversion rates and average deal sizes. Navah Hopkins, Optmyzr Brand Evangelist and one of the top 25 most influential voices in PPC, advises: "If you don't have perfect data, start with proxy values based on your lowest customer value and refine over time."

Waiting for perfect data is the most common reason advertisers delay VBB adoption and miss performance gains. Imperfect values that are consistently applied and regularly refined outperform no values at all.

A practical framework for calculating conversion values

There are three main approaches to assigning values, depending on your business model.

Revenue-based values apply directly to ecommerce. Each transaction passes its actual cart value as the conversion value. This is the most accurate setup available and what Weber describes as best practice for ecommerce: "Having dynamic values for each product and allowing Google to optimize performance via actual dollar values is best practice."

Profit-based values adjust for margin. If your product generates $3,000 in revenue but costs $1,650 to produce and fulfill, passing $3,000 as the conversion value tells the algorithm to optimize for gross revenue, not profit. Passing $1,350 (the margin) produces a more accurate signal.

Proxy values for lead generation require a calculation. Use this formula:

Conversion value = Average deal size x Profit margin x Lead-to-close rate

Conversion value = Average deal size x Profit margin x Lead-to-close rate

Conversion value = Average deal size x Profit margin x Lead-to-close rate

If your average deal is $20,000, your margin is 40%, and 15% of your leads become customers:

$20,000 x 0.40 x 0.15 = $1,200 per lead

$20,000 x 0.40 x 0.15 = $1,200 per lead

That is the value you assign to each qualified lead conversion event. As your CRM data matures, you can segment this further by lead source, industry, or company size to pass differentiated values.

Using conversion value rules for dynamic adjustments

Google Ads Conversion Value Rules let you adjust the relative value of a conversion based on conditions the algorithm cannot observe directly, such as:

• Device type (if desktop leads close at 2x the rate of mobile leads, assign a 2x multiplier)
• Geographic location (if enterprise clients cluster in specific metros, weight those conversions higher)
• Audience segment (CRM-matched lists of past high-value customers can carry a premium multiplier)

These rules do not change your base conversion values. They layer an adjustment on top so the algorithm can act on signals your tracking setup cannot capture on its own.

How offline conversions and CRM integration complete the picture

For most B2B advertisers, the most valuable conversion events happen offline. A lead submits a form online, but the sale closes in a call or in a CRM two weeks later. Without feeding that outcome back to Google, the algorithm learns only that a form was submitted. It has no idea whether that click produced $0 or $50,000 in revenue. Offline conversion tracking closes that loop.

As Navah Hopkins puts it: "If you're seeing a lot of junk leads, that's a sign you need to integrate offline conversion tracking and send quality signals back to Google."

How offline conversion tracking works in practice

1. A prospect clicks your ad. Google assigns them a GCLID.
2. Your form captures the GCLID alongside the lead details and stores it in your CRM (Salesforce, HubSpot, or similar).
3. When the lead progresses to a meeting, opportunity, or closed deal, your CRM records that outcome.
4. You upload the conversion event back to Google Ads using the GCLID, along with the associated value.
5. Google maps the outcome to the original click and updates its bidding model.

You can automate this upload through direct CRM integrations with tools like Zapier, native Salesforce or HubSpot connectors, or the Google Ads Offline Conversions API.

The 90-day rule you cannot ignore

Google Ads only applies offline conversion data to its bidding model if it is uploaded within 90 days of the original click. Any upload outside that window counts for reporting only and does not affect bidding optimization. For B2B companies with sales cycles longer than three months, you need to optimize to an earlier, faster-converting stage in your funnel, such as qualified lead, demo booked, or opportunity created, rather than waiting for closed revenue.

Enhanced conversions for leads

For advertisers who want to improve attribution accuracy without a full offline conversion workflow, Google's Enhanced Conversions for Leads feature hashes and matches first-party user data from form submissions (email addresses, phone numbers) to Google accounts. Research from GrowLeads citing Google internal data shows this can improve conversion tracking accuracy by 11% and increase measurable conversion value by 14%.

Server-side tracking as the infrastructure foundation

Browser-based pixels suffer from ad blocker interference, cookie deletion, and cross-device tracking gaps. A 2025 B2B PPC report from The Digital Bloom found that server-side conversion tracking via the Conversions API achieves 15 to 30% more complete conversion attribution than browser pixels alone. For VBB to work well, the data feeding the algorithm needs to be as complete as possible. Server-side tracking is the most reliable way to achieve that.

What results to expect from value-based bidding

The performance case for VBB is well-documented. Google data shows that advertisers switching from target CPA to target ROAS see a median 14% increase in conversion value at a similar return on ad spend. Some brands report even larger gains: a 30% lift in cost efficiency and 20% higher revenue after adopting VBB, according to figures cited by Google. Results depend on data quality and implementation.

These are not guaranteed outcomes. They reflect what happens when the strategy is properly implemented with clean conversion values, consistent data uploads, and enough conversion volume for the algorithm to learn.

Realistic expectations by business type

The role of AI and predictive LTV modeling

Standard VBB uses historical transaction values or proxy estimates. Advanced implementations use AI-based lifetime value prediction to assign more accurate prospective values at the point of conversion. According to Voyantis, AI-based LTV prediction models can increase ROAS by 20 to 40% over static value models by calculating each new user's predicted long-term revenue rather than relying on a fixed average.

"The real advantage comes when you stop feeding the algorithm a flat number and start feeding it a model. That is when the system starts optimizing toward customers who actually grow with you, not just ones who convert once."

Derick Do, Co-Founder and Chief Product Officer, Launchcodex

This approach requires dedicated data science tooling, but it represents where the industry is heading. Brands that build this infrastructure now build a compounding advantage over those who stay on static values.

VBB beyond Google: value-based optimization on Meta Ads

Value-based optimization is available on Meta Ads too. Meta runs a parallel system called Value Optimization within its Advantage+ campaigns. The principle is the same: feed the platform historical purchase value data, and it bids more aggressively for users whose profiles predict higher-value transactions. Brands using Meta's Value Optimization often see 20 to 30% higher average order values and more stable ROAS at scale.

Most published content on VBB focuses on Google. A full-funnel paid media program applies value-based logic across every paid channel, and the same data discipline that makes Google VBB work transfers directly to Meta.

How Meta's Value Optimization differs from Google VBB

Meta uses historical pixel data to build a user-level purchase value estimate. The system then bids more for users whose profile resembles past high-value buyers, adjusted for the probability and predicted size of a future purchase. Unlike Google's auction-time signals, Meta's model leans more on audience pattern matching.

For this to work well on Meta, you need:

• A clean Meta Pixel or Conversions API setup with purchase values passed for every transaction
• At least several months of purchase history for the algorithm to model from
• Consistent creative and offer structure so the algorithm can isolate audience signal from creative performance

Cross-platform value attribution

Running VBB across both Google and Meta creates an attribution challenge. Each platform claims credit for conversions, and each will optimize toward the signal it receives. Using a third-party attribution tool like Triple Whale, Wicked Reports, or a direct CRM-based attribution model alongside your platform data gives you a more accurate read on where value is actually being created. This prevents over-investing in one platform based on inflated last-click attribution.

Common mistakes that cause value-based bidding to fail

VBB fails when the data quality going into the system does not match the decisions expected out of it. The algorithm is only as good as the signals you supply. The most common failures follow predictable patterns, and all of them are preventable with the right setup.

A Dun and Bradstreet survey found that 34% of B2B marketers and sales leaders cite inaccurate customer data as a major obstacle to successful data-driven marketing. In a VBB context, bad data does not produce neutral results. It actively misdirects the algorithm.

Five mistakes to avoid

1. Assigning the same value to every conversion. If all your conversion actions share a single value, the algorithm has no differentiation to work with. This is equivalent to running maximize conversions. You must supply at least two distinct values for VBB to function as intended.
2. Setting tROAS too high at launch. A tROAS target above your historical average restricts auction eligibility while the algorithm is still learning. Set it 20% below historical ROAS and tighten it after the learning period stabilizes.
3. Optimizing to multiple funnel stages simultaneously. Google's documentation warns that bidding to more than one stage within a single strategy, such as website visit and qualified lead, leads to duplicative signals and weaker performance. Choose one stage with meaningful volume and value.
4. Skipping offline conversion uploads for B2B. Without closed deal data feeding back into the system, your campaigns optimize for form volume, not revenue. The algorithm learns to generate the kind of leads that are easy to get, not the kind that close.
5. Failing to hit the minimum conversion threshold. Google's minimum is 15 conversions in 30 days. In practice, Jyll Saskin Gales, a former Google Ads strategist and current Google Ads coach, recommends targeting 30 or more per month for the algorithm to make reliable decisions. Accounts below this threshold should run maximize conversions first to build volume before switching to a value-based strategy.

How to build a VBB-ready data infrastructure

Getting value-based bidding to perform is a data infrastructure problem, not a campaign settings problem. The companies that see the biggest gains from VBB do the foundational work first: clean conversion tracking, a functioning CRM integration, a defined value framework, and a consistent data upload cadence.

Here is the setup sequence that produces reliable outcomes.

Phase 1: Conversion tracking audit

Before changing any bidding strategy, audit what you are currently tracking. A conversion rate optimization audit is a practical starting point for identifying which actions carry real business value and which are cluttering your primary conversion column.

• Confirm that all key conversion actions are firing correctly and not double-counting.
• Remove or demote low-intent actions (page views, time on site) from your primary conversion column.
• Identify which conversions carry different business value and map each to a realistic monetary amount.

Phase 2: Value assignment framework

Use the proxy value formula introduced earlier and document it in a shared reference your team and platform partners can apply consistently.

• Build a table with each conversion event, its assigned value, the data source for that value, and a review date.
• Review and update values quarterly as CRM data matures.

Phase 3: CRM and data pipeline setup

• Connect your CRM (Salesforce, HubSpot, or similar) to Google Ads using the native integration or a tool like Zapier.
• Configure your forms to capture and store the GCLID at lead submission.
• Set up an automated daily upload of qualified conversion events with associated values.
• Implement Enhanced Conversions for Leads to fill attribution gaps that pixel-based tracking misses.

Phase 4: Conversion value rules

Once base tracking is clean and values are assigned, layer in Conversion Value Rules for any dimensions where you have evidence of value variation, such as geographic lift, device performance differences, or audience segment quality.

Phase 5: Bidding strategy transition

Do not switch directly from manual bidding or target CPA to target ROAS if your account lacks sufficient conversion volume. Use this transition path:

1. Run maximize conversions until you hit 30 or more conversions per month consistently.
2. Switch to maximize conversion value (no tROAS) and verify that value signals are being received correctly.
3. Add a tROAS target set 20% below your historical ROAS average.
4. Allow 14 days without bid or creative changes for the learning period to stabilize.
5. Tighten tROAS incrementally, in increments no larger than 10 to 15%, once performance is stable.

Value-based bidding is a system, not a setting

Value-based bidding is a more precise version of something every good marketing automation program already tries to do: spend more to acquire customers who are worth more to the business. The shift is that VBB makes this systematic, measurable, and automated rather than ad hoc.

The logic applies to any platform or channel that accepts conversion value data, whether that is Meta Ads, Microsoft Advertising, or a programmatic DSP. The discipline is the same: define what value means to your business, build the data infrastructure to capture it, and feed those signals to the systems making decisions on your behalf.

The advertisers seeing the strongest results are not the ones with the biggest budgets. They are the ones with the cleanest data, the clearest definition of what a good customer looks like, and the discipline to feed that information back to their campaigns consistently.

If your current campaigns are generating conversions but not revenue, the problem is almost certainly the signal you are giving the algorithm. Fix the signal, and the algorithm will follow.

FAQ

What is the minimum conversion volume needed for value-based bidding?

Google's official minimum is 15 conversions in the past 30 days, but experienced practitioners recommend 30 or more per month for the algorithm to learn reliably. Accounts below this threshold should build volume with maximize conversions first, then migrate to a value-based strategy once that floor is established.

What if I do not know the exact value of each conversion?

Use proxy values calculated from your historical lead-to-close rate, average deal size, and profit margin. A reasonable estimate applied consistently outperforms leaving value data blank. Refine the values quarterly as your CRM data improves.

Can value-based bidding work for lead generation, not just ecommerce?

Yes. Lead gen advertisers assign values based on estimated downstream revenue or lead quality scores. B2B companies use CRM-based lead scoring to pass a numeric value with each form submission, telling the algorithm which types of leads are worth bidding more to acquire.

How long does the VBB learning period take?

Google recommends allowing at least 14 days after switching to a value-based strategy before evaluating performance or making bid or creative changes. Accounts with lower conversion volume may need longer.

Is VBB only available on Google Ads?

No. Meta Ads uses a parallel approach called Value Optimization within Advantage+ campaigns. Microsoft Advertising also supports target ROAS bidding. Cross-platform value-based optimization is increasingly common for brands running multi-channel paid programs.

What happens if I set my tROAS target too high?

A tROAS target above your historical average restricts the number of auctions the algorithm enters, reducing impression volume while the system struggles to meet an unreachable target. Start 20% below your historical average and adjust upward only after performance stabilizes.