Calculate cost per qualified phone lead by adding every cost attributable to the AI calling workflow for one defined measurement window, then dividing that total by the number of phone leads that met a written qualification rule in the same window.
Cost per qualified lead = total attributable workflow cost ÷ qualified phone leads
Do not use transfers as a substitute for qualified leads unless your written definition explicitly makes a successful transfer the qualification event. Keep qualification rate and transfer rate as separate diagnostics. If there are no qualified leads, report the result as “not calculable” rather than zero.
Start with the Q-COST framework
A useful calculation depends more on boundary discipline than spreadsheet complexity. The Q-COST framework makes those boundaries explicit:
- Q — Qualification contract: Write the exact conditions a lead must meet.
- C — Cohort and clock: Fix the population and measurement window.
- O — Outcome ledger: Count outcomes from records, not impressions.
- S — Scoped spend: Include costs attributable to that same cohort and window.
- T — Two-rate diagnosis: Report qualification and transfer rates separately.
This sequence prevents a common measurement error: changing the definition, denominator or cost scope after seeing the result.
Q — Write a qualification contract
A “qualified” lead is not self-defining. Before collecting numbers, write a rule that another reviewer could apply to the call record and reach the same decision. A useful contract contains:
- the required fit conditions, such as service area, use case or account type;
- the required intent signal, such as agreeing to a next step;
- any disqualifiers;
- the evidence field that proves each condition;
- the terminal outcome label stored in the workflow records;
- the treatment of repeat callers, incomplete calls and later reversals.
For example, “interested” is too subjective on its own. A better internal rule might require a matching service area, a stated need and consent to a defined next action. That is an operational definition, not a claim that the lead will buy.
Freeze the contract for the measurement window. If the rule changes, begin a new version or recalculate the full cohort under the new rule.
C — Fix the cohort and clock
Choose one start time, one end time and one inclusion rule. The numerator, denominator and spend must all refer to this same boundary. State whether “total calls” means attempts, connected calls or completed conversations. For the basic worksheet below, it means all call records included under your declared rule.
Also decide how to handle:
- calls started before the window but completed inside it;
- retries to the same contact;
- inbound and outbound calls;
- test, spam, wrong-number and duplicate records;
- leads whose qualification is confirmed after the call;
- messages sent before or after the call.
There is no universal choice for every workflow. Consistency and disclosure matter more than selecting a particular convention.
O — Build an auditable outcome ledger
Use one row per included call record, with a stable call identifier. At minimum, retain the call date, cohort label, outcome, qualification status, transfer status and reason for any manual correction. Where possible, keep the evidence fields used by the qualification contract alongside the outcome.
CallTurbo’s public site states that call history, outcomes, contact notes and follow-up tasks remain available in its CRM workspace. It also describes visibility into calls, transfers and outcomes. Those records can supply the workflow side of the calculation, but your team still has to define which recorded outcome counts as qualified.
Reconcile the final counts before calculating. A call should not be silently dropped because it has an awkward outcome. Put unresolved records in an explicit “unknown” category, then decide whether your denominator includes them.
S — Scope attributable spend
Create one cost row for each resource consumed by the included workflow. The core calculation is:
Total attributable cost = calling cost
+ AI voice cost
+ messaging cost
+ other attributable cost
Cost per qualified phone lead = total attributable cost
÷ qualified lead count
CallTurbo publicly describes credits used across eligible calling, AI voice and messaging usage. Use current, user-entered or product-supplied rates rather than copying a static rate into the worksheet. The CallTurbo pricing page is the appropriate product path for configured plan and usage inputs, although the public page may require JavaScript to display them.
“Other attributable cost” should be transparent rather than a catch-all. Depending on the decision you are making, it could include workflow-specific carrier charges, number rental, setup work, review labour or connected services. Include an item only if you can state its source and allocation method. Do not mix a monthly cost into a weekly window without allocating it, and do not include company-wide overhead selectively in only one workflow.
If credits cover several operations or campaigns, convert only the usage attributable to this cohort into cost. Record the rate version and billing period used. If a reliable allocation is impossible, show the disputed amount separately and calculate a range rather than pretending to have false precision.
T — Keep two operational rates separate
Calculate these companion measures with the same call denominator:
Qualification rate = qualified phone leads ÷ total included calls
Transfer rate = transferred calls ÷ total included calls
A transfer is a routing event; qualification is a status under your contract. One can occur without the other. A qualified lead might request a later appointment instead of a live transfer. A call might be transferred for human review before qualification is established. Report both counts even if your current workflow usually makes them coincide.
If you also calculate cost per transfer, label it clearly. It answers a different question from cost per qualified lead.
Reusable calculation worksheet
Complete this worksheet for one versioned measurement window:
| Input | Value to enter | Source or rule to record |
|---|---|---|
| Window start and end | — | Time zone and inclusion convention |
| Qualification definition/version | — | Written contract and outcome label |
| Total included calls | — | Export or deduplicated call ledger |
| Calling cost | — | Current usage record and rate version |
| AI voice cost | — | Current usage record and rate version |
| Messaging cost | — | Current usage record and rate version |
| Other attributable cost | — | Itemised amount and allocation method |
| Qualified lead count | — | Audited outcome ledger |
| Transferred call count | — | Audited transfer records |
Then produce four outputs:
| Output | Formula | Zero-denominator handling |
|---|---|---|
| Total attributable cost | Calling + AI voice + messaging + other | Can be zero if supported by the inputs |
| Cost per qualified lead | Total attributable cost ÷ qualified leads | “Not calculable: no qualified leads” when qualified leads = 0 |
| Qualification rate | Qualified leads ÷ total calls | “Not calculable: no included calls” when total calls = 0 |
| Transfer rate | Transfers ÷ total calls | “Not calculable: no included calls” when total calls = 0 |
Store the raw inputs with the output. A standalone cost-per-lead figure is difficult to interpret later if its window, currency, tax treatment, qualification version and exclusions have disappeared.
Example: one-week workflow in euros (illustrative)
The following values are hypothetical and demonstrate the method only. They are not CallTurbo prices, observed results or a benchmark.
A team measures one week of calls using a fixed qualification contract. Its worksheet contains:
| Input | Illustrative value |
|---|---|
| Total included calls | 1,000 |
| Calling cost | €180 |
| AI voice cost | €120 |
| Messaging cost | €20 |
| Other attributable cost | €80 |
| Qualified phone leads | 80 |
| Transferred calls | 50 |
First, add the attributable costs:
€180 + €120 + €20 + €80 = €400
Then calculate the principal measure:
Cost per qualified phone lead = €400 ÷ 80 = €5
Calculate the two diagnostics without treating them as interchangeable:
Qualification rate = 80 ÷ 1,000 = 8%
Transfer rate = 50 ÷ 1,000 = 5%
The correct report for this illustrative window is therefore €5 per qualified phone lead, an 8% qualification rate and a 5% transfer rate, subject to the stated cost scope and qualification contract. It would be wrong to divide €400 by 50 and label the resulting €8 “cost per qualified lead”; that is cost per transfer, not the requested measure.
Decision table: what should enter the calculation?
| Situation | Treatment | Reason |
|---|---|---|
| Usage belongs entirely to the measured workflow and window | Include it | It is directly attributable |
| A monthly fixed cost supports several workflows | Allocate it using a declared, repeatable rule | The full amount would distort one cohort |
| A cost cannot be allocated reliably | Exclude and disclose it, or calculate a range | False precision hides uncertainty |
| A call was transferred but fails the qualification contract | Count as transferred, not qualified | Routing and qualification are different events |
| A lead qualifies but chooses a later follow-up | Count as qualified; do not count as transferred | Qualification does not require a live handoff unless the contract says so |
| A retry reaches the same person | Apply the predeclared deduplication rule | Otherwise dialling strategy changes the denominator |
| Qualified lead count is zero | Report cost per qualified lead as not calculable | Division by zero is not a zero cost |
| Total call count is zero | Do not calculate either rate | Both rate denominators are zero |
| Qualification definition changes mid-window | Split the window or reclassify the whole cohort | Mixed definitions make comparison unreliable |
Implementation steps
- Name the decision. Decide whether the calculation will compare workflow versions, monitor one campaign or support budgeting. This determines which costs are relevant.
- Version the qualification contract. Specify fit, intent, exclusions, evidence and final label. Give the rule a version identifier.
- Declare the window and denominator. Record dates, time zone, call inclusion rule and duplicate treatment before exporting data.
- Export the outcome ledger. Retrieve call identifiers, outcomes, notes, qualification labels and transfer events. Preserve unknown outcomes instead of deleting them.
- Retrieve current cost inputs. Use billing or usage records for calling, AI voice and messaging. Record currency, tax treatment, rate version and any credit conversion used.
- Itemise other costs. Give each amount a label, source and allocation method. Keep excluded costs in a visible note.
- Reconcile counts and scope. Confirm that usage, calls, qualified outcomes and transfers all belong to the same cohort and period.
- Apply zero checks, then calculate. Test the qualified count and total-call count before division. Produce the four worksheet outputs.
- Review exceptions. Sample unknowns and manual corrections against the written contract. Log any reclassification.
- Publish the assumptions with the result. Keep the worksheet, source labels and qualification version together so another person can reproduce the calculation.
For repeated reporting, lock formula cells and make only the inputs editable. Add validation that rejects negative costs and counts, warns when transfers exceed included calls, and blocks rates when total calls are zero. A qualified count greater than total calls is also a signal to check whether one call can produce multiple leads; if it can, rename the denominator and formula rather than forcing the data into a call-based model.
Limitations
Cost per qualified phone lead is a bounded operating measure, not proof of revenue, profit, lead quality after handoff or causal performance. It is only as reliable as the qualification rule, recorded outcomes and cost allocation.
The basic model has several limitations:
- Outcome quality: Automated labels and manual reviews may be incomplete or inconsistent.
- Timing: A lead can qualify or be reversed after the measurement window closes.
- Allocation: Shared plans, labour and infrastructure may not map cleanly to one workflow.
- Denominator choice: Attempts, connections and completed conversations produce different rates.
- Currency and tax: Comparisons across regions or periods require consistent treatment.
- Workflow differences: Two campaigns may use different audiences, qualification rules and transfer policies, so their figures may not be comparable.
- Downstream value: The model does not estimate sales, lifetime value or return on investment.
Do not compare two cost-per-qualified-lead figures until their qualification contracts, windows, denominators and cost scopes are aligned. When they are not aligned, show the differences rather than ranking the workflows.
FAQ
What if there are no qualified phone leads?
Report cost per qualified lead as “not calculable: no qualified leads”. Keep total attributable cost and the zero qualified count visible. Do not report zero, because the workflow still incurred the recorded cost.
Should transfers count as qualified leads?
Only if the written qualification contract explicitly defines a successful transfer as the qualification event. Otherwise, maintain separate qualified and transferred fields and report both rates.
Should setup and staff time be included?
Include them when they are relevant to the decision and can be allocated consistently. Itemise the amount and method. For a short-run operating view, you may choose to exclude one-time setup; for a full-cost view, you may allocate it across a declared period or expected volume. Label the view so readers know which calculation they are seeing.
Which calls belong in total calls?
Use the population declared before measurement: for example, all valid attempts or all connected calls. Exclude tests, duplicates or spam only under a documented rule. The same rule must be used for every compared window.
Can I compare this figure across campaigns?
Yes, but only after checking that qualification definitions, cost categories, currency treatment, periods and denominator rules are equivalent. If they differ, the comparison mixes operating performance with accounting and policy differences.
How often should the calculation be updated?
Update it whenever the reporting window closes and whenever rates, workflow scope or the qualification contract changes. Keep each version so historical results retain the assumptions that produced them.