Value-based bidding for lead generation means giving Google Ads information about which leads matter more, then using Maximize Conversion Value or Target ROAS to optimize toward that value.
It solves a common measurement failure: a spam form, an existing-customer call, a valid prospect, and a booked appointment all appear as one conversion.
The bidding strategy is the last step. First you need a value model, an attribution path, and a reliable feedback loop. This guide builds those pieces for businesses where sales happen after a form or phone call, including a practical CallRail workflow.
What is value-based bidding for lead generation?
Google Ads Smart Bidding can optimize for:
- Conversion volume: Maximize Conversions or Target CPA
- Conversion value: Maximize Conversion Value or Target ROAS
The value strategies use the numbers reported with each conversion. For ecommerce, that number is often transaction revenue. For lead generation, it can represent:
- Actual closed revenue
- Gross profit or contribution value
- Expected revenue based on lead type or stage
- A validated relative score when financial value is not yet available
Google’s conversion-value guide explicitly supports transaction-specific values when individual sales or leads have different value.
Why raw leads are a weak bidding signal
Suppose a campaign generates 100 phone calls at a $100 cost per call. A conversion-volume report says the campaign produced 100 conversions for $10,000.
Call review shows:
| Call outcome | Calls | Share |
|---|---|---|
| Qualified prospects | 30 | 30% |
| Existing customers | 25 | 25% |
| Unsupported need or location | 20 | 20% |
| Vendor, spam, or unrelated | 15 | 15% |
| Missed or unresolved | 10 | 10% |
The real cost per qualified call is:
$10,000 / 30 = $333.33
If a second campaign produces only 80 calls for the same spend but 40 qualify, its platform cost per raw call looks worse while its cost per qualified call is better:
$10,000 / 40 = $250
When all calls share the same conversion action and value, bidding cannot see that reversal. Value-based measurement makes it visible.
Choose the deepest reliable outcome, not the deepest possible outcome
The best bidding signal balances business importance, coverage, and speed.
| Outcome | Business relevance | Typical speed | Common limitation |
|---|---|---|---|
| Raw call or form | Low | Immediate | Includes weak and invalid demand |
| Qualified lead | Medium to high | Minutes to days | Requires consistent qualification |
| Appointment or opportunity | High | Hours to weeks | Sales process can influence the result |
| Closed customer | Very high | Days to months | Low volume and long reporting delay |
| Collected profit | Highest economic fidelity | Weeks to months | Often too delayed or incomplete for primary bidding |
Do not automatically bid only to the deepest CRM stage. If closed revenue arrives 90 days later and only half of records are updated, a qualified-lead outcome available the same day may be the more useful bidding signal.
You can still keep later stages as secondary conversions for validation.
Google currently recommends choosing qualified lead or converted lead as the goal when setting up enhanced conversions for leads. Its offline conversion FAQ defines a qualified lead as one further qualified in a CRM or internal lead-generation system.
Four ways to calculate lead conversion values
Method 1: actual revenue
Send the revenue associated with the closed customer.
Use this when:
- Revenue is connected to the original lead.
- Refunds, cancellations, and duplicate deals are handled.
- The delay still produces adequate signal.
- Revenue is a reasonable proxy for what the business wants to maximize.
If margins vary significantly, revenue can favor large but unprofitable deals. Profit or contribution value may align better.
Method 2: expected value by lead type
Calculate expected value from observed close rate and average economic value.
Expected lead value = close rate × average value per customer
Example:
| Lead type | Close rate | Average collected revenue | Expected value |
|---|---|---|---|
| Booked replacement estimate | 30% | $12,000 | $3,600 |
| Qualified repair inquiry | 45% | $900 | $405 |
| General valid inquiry | 10% | $1,200 | $120 |
This is stronger than assigning values by intuition because the hierarchy is tied to observed outcomes. Recalculate on a fixed cadence and use enough history to reduce noise.
Method 3: expected gross profit
Use contribution economics when services have different margins.
Expected lead value = close rate × average collected revenue × contribution margin
If a lead type closes at 30%, produces $12,000 in average collected revenue, and has a 35% contribution margin:
30% × $12,000 × 35% = $1,260 expected contribution value
Define the margin consistently. Do not mix gross margin for one service with revenue for another.
Method 4: relative proxy values
When financial data is not ready, use a relative value system that consistently gives stronger leads more weight. LeadUp uses this method for scored phone leads:
Lead value = caller-type base value + lead-status base value + (star rating × per-star value)
You can set base values for caller type and lead status independently. For example:
| Setting | Example relative value |
|---|---|
| First-time caller base | 100 |
| Repeat caller base | 25 |
| Good-lead base | 50 |
| Bad-lead base | 0 |
| Value per star | 20 |
A four-star, first-time good lead would receive a value of 230:
100 + 50 + (4 × 20) = 230
A four-star, repeat good lead would receive a value of 155:
25 + 50 + (4 × 20) = 155
These figures are examples, not benchmarks. The spacing should reflect how you want Google Ads to prioritize caller types and lead quality. Proxy values are not revenue, so label the resulting metric as value/cost or a quality-value index rather than financial ROAS.
A practical phone-lead value model
For call-driven businesses, define outcomes from information available in the conversation:
- Caller status: first-time prospect, repeat prospect, or existing customer
- Need: supported service or case type
- Fit: location, eligibility, budget, jurisdiction, or other business-specific constraints
- Intent and urgency: research, active problem, or immediate need
- Next step: no action, follow-up requested, estimate booked, or appointment scheduled
LeadUp determines whether the call is a good lead and assigns a one-to-five-star score. The Google Ads push can then be configured to:
- Send only good leads, or send both good and bad leads.
- Keep first-time and repeat caller conversions separate.
- Use different caller-type base values for first-time and repeat callers.
- Use different lead-status base values for good and bad leads.
- Add a configurable amount for each star in the lead rating.
- Allow a user to increase or decrease the conversion value manually after it has been pushed.
This preserves more information than a single generic Phone Call conversion while keeping the value model understandable.
What CallRail sends to LeadUp after each call
CallRail is the post-call data source in this workflow. After the call, a webhook sends LeadUp the available call data, including items such as the recording and tracking number.
CallRail does not provide the transcript, summary, qualification, or LeadUp score used by this workflow. LeadUp creates and applies those layers after receiving the post-call data.
The LeadUp + CallRail value feedback loop
LeadUp’s value-based conversion tracking automates the processing and feedback layer after CallRail delivers the post-call webhook:
- Receive: CallRail sends the recording, tracking number, and other available call data to LeadUp.
- Transcribe and summarize: LeadUp creates its own transcript and call summary.
- Qualify and score: LeadUp determines whether the call is a good lead and assigns a one-to-five-star rating.
- Assign relative value: LeadUp combines the selected caller-type base value, good- or bad-lead base value, and configured per-star value.
- Select conversion events: You choose whether to push only good leads or both good and bad leads, with separate events for first-time and repeat callers.
- Push and refine: LeadUp sends the event and relative value to Google Ads. A user can manually increase or decrease the value afterward when the initial proxy needs correction.
This creates a useful intermediate signal without requiring a mature CRM. A CRM remains valuable for later stages such as opportunities, closed sales, collected revenue, and lifetime value.
Do not run parallel systems that upload the same event under the same action without a deduplication plan. Decide which system owns each stage and conversion action.
Configure Google Ads so bidding uses the right actions
1. Audit primary and secondary conversions
Primary actions appear in the Conversions column and are used for bidding when their standard goal is active. Secondary actions normally remain observational in All conversions. Custom goals are an important exception because actions included in a custom goal are used for bidding regardless of their primary/secondary setting. Google’s primary and secondary action documentation explains the behavior.
A practical structure might be:
| Conversion action | Role during validation |
|---|---|
| Raw phone call | Secondary |
| Qualified phone lead with value | Primary or initially Secondary during observation |
| Booked appointment | Secondary validation action, or Primary when volume and delay support it |
| Closed customer/revenue | Secondary validation action until coverage is reliable |
The correct design depends on volume and funnel overlap. If Qualified Lead and Booked Appointment can both fire for one person, including both in bidding may double-count progress unless the values and goal structure deliberately account for it.
2. Confirm the post-call data path
- Verify that the CallRail post-call webhook reaches LeadUp.
- Confirm the recording and required tracking details are present.
- Check that LeadUp creates the transcript, summary, qualification, and star rating.
- Confirm first-time and repeat callers use the intended conversion events and caller-type base values.
- Confirm good and bad leads use the intended lead-status base values.
- Verify the good-leads-only or good-plus-bad push setting matches the campaign goal.
3. Validate the relative value formula
Review examples across first-time and repeat callers at every star rating. Confirm that the resulting value differences are large enough to express your priorities without pretending to represent revenue.
After a value is pushed, compare it with the call and known outcome. Use the manual increase or decrease control when the initial proxy does not reflect the lead’s real value.
4. Monitor diagnostics and value coverage
Track at least:
- Eligible calls received
- Calls attributed to Google Ads
- Calls classified
- Calls sent with a non-null value
- Import errors
- First-time and repeat caller value distributions
- Distribution of values by campaign and week
A sudden rise in zero-value calls can reflect worse traffic or a failed classification process. Separate business outcomes from technical states.
When to switch to value-based bidding
Do not switch as soon as the first values appear.
Google recommends that Search and Shopping campaigns that recently began reporting conversion values, or changed how values are reported, wait four weeks or three conversion cycles after including the values in the Conversions column before adopting Maximize Conversion Value. Review the current guidance in Google’s Maximize Conversion Value article.
Before switching, confirm:
- At least two differentiated values occur with meaningful frequency.
- Coverage is stable across campaigns, locations, devices, and schedules.
- Values correlate with downstream qualified or closed outcomes.
- Recent reporting is evaluated after the full conversion delay.
- Budgets can tolerate spend changes.
- Stakeholders agree on what the value number represents.
Then choose:
- Maximize Conversion Value to pursue the most value within budget.
- Target ROAS to pursue value with an average return constraint.
Use our Maximize Conversions vs Maximize Conversion Value guide for the volume-versus-value decision, and the Target CPA vs Target ROAS comparison when the account needs an efficiency target.
Frequently asked questions
Do conversion values for leads have to equal revenue?
No. You can use expected values or relative values when revenue is unavailable. Make the meaning explicit and validate that higher values correlate with better business outcomes.
What does CallRail send to LeadUp?
CallRail sends available post-call data through a webhook, including information such as the recording and tracking number. LeadUp creates the transcript and summary, qualifies and scores the call, assigns the relative value, and handles the configured Google Ads conversion push.
Do I need a CRM for value-based bidding?
Not necessarily. A CRM is the strongest source for later sales stages, but LeadUp can create its own transcript and summary from the CallRail recording, qualify and score the lead, and push a relative proxy value without a complex CRM integration.
Should an unqualified lead have a value of zero?
Usually, if it has no acquisition value. Keep “unqualified” separate from “not yet reviewed” or “failed to process.” A technical unknown should not silently become a business judgment.
How often should lead values be updated?
Review them on a fixed cadence and whenever pricing, close rates, margin, qualification rules, or the sales process changes materially. Avoid frequent reactive changes based on small samples.
Bottom line
Value-based bidding for lead generation works when the value signal is more truthful than raw lead count.
Start with the deepest outcome you can report quickly and consistently. Calculate values from real economics where possible, validate the ordering, keep Google Ads goals clean, and observe the data before switching bids.
If phone calls are the missing feedback loop, see how LeadUp turns CallRail call data into qualified-lead conversion values.