Visitor identification ROI measures financial return from the identification workflow against the full cost of operating it. A list of matched people is not pipeline, and pipeline is not realized revenue.
I am George, founder of Leadpipe. The framework below is designed to let a finance or RevOps owner check every input. Use incremental gross profit, less program costs, divided by program costs for realized ROI. Report sourced pipeline as a separate leading indicator.
This distinction matters when a trial produces useful records before the sales cycle has had time to close. You can validate data quality quickly without pretending the financial result is already known.
Define the unit and the observation window
Choose a cohort, such as accounts first accepted into a visitor-driven workflow during one month. Track that cohort over an agreed period. The observation window must be long enough for your sales process; it does not have to match the subscription billing month.
Keep people and accounts separate. Several people at one company may contribute to one opportunity. Repeated sessions from the same person are not additional people. A contact whose account was already in an active deal is not a new sourced account.
Our person-level versus company-level guide explains why the denominator matters. Retain the original cohort assignment when a record revisits later.
Use these metrics for different decisions
| Metric | Formula | What it tells you |
|---|---|---|
| Person identification rate | Unique matched people / eligible unique visitors | Coverage for the stated cohort |
| Cost per accepted person | Full program cost / unique people accepted for the workflow | Cost of usable output |
| Cost per accepted account | Full program cost / unique accounts accepted for the workflow | Account-level economics |
| Sourced pipeline | Value of distinct qualified opportunities attributed to the workflow | Potential future business |
| Realized ROI | (Incremental gross profit − program cost) / program cost | Financial return over the stated period |
| Payback date | First date cumulative incremental gross profit covers cumulative program cost | Time to recover the investment |
A zero denominator produces an undefined metric, not zero cost. Label it “not enough data” and keep the numerator visible. Pipeline divided by cost can be a useful pipeline multiple, but label it as such.
Build the funnel from observed counts
Start with eligible unique visitors in the selected region and page group. Record the eligibility rules, including bot, employee and other approved exclusions. Then measure actual unique identities returned rather than multiplying traffic by a vendor’s advertised match rate.
For each stage, keep the rejected count and reason:
- Matched: the provider returned a company or person record.
- Accepted: the record has the fit and fields your workflow requires.
- Eligible for action: it survives suppression and duplicate checks.
- Worked: an owner took the defined action.
- Qualified opportunity: sales accepted a real opportunity under your CRM rules.
- Closed revenue: the deal closed and revenue was recorded using the chosen convention.
Do not infer step six from step one. If many accepted records never reach step four, the immediate problem is operating capacity. Increasing identification volume may make that problem worse.
Use the trial scorecard for the first stages. For Orbit audiences, build a separate cohort starting with the audience delivered, because those people do not necessarily originate from your own website traffic.
Count the full cost of the program
Include the subscription, additional data, integration costs, and the labor needed to review and work the records. Allocate setup work across an explicitly chosen period rather than quietly excluding it.
Use incremental costs consistently. If an existing CRM subscription would be paid regardless of this program, show the allocation policy rather than counting the full company bill against a small experiment. Conversely, include a new paid integration required solely to run it.
The visitor identification pricing guide separates plan capacity from observed cost. A low advertised price per available credit can coexist with an expensive cost per accepted account if most records cannot be used.
Worked example: hypothetical, not a Leadpipe result
The following figures demonstrate the arithmetic. They are invented planning inputs, not customer results, forecasts or benchmarks.
| Input over a 90-day cohort window | Illustrative amount |
|---|---|
| Subscription and infrastructure | $900 |
| Operations and follow-up labor | $600 |
| Total program cost | $1,500 |
| Accepted unique accounts | 100 |
| Qualified opportunities attributed to the workflow | 10 |
| Pipeline value at creation | $50,000 |
| Closed revenue attributed under the selected convention | $10,000 |
| Assumed gross margin | 80% |
From these inputs:
- Cost per accepted account is $1,500 / 100 = $15.
- Pipeline multiple is $50,000 / $1,500 = 33.3x.
- Gross profit associated with the attributed revenue is $10,000 × 80% = $8,000.
- If all that gross profit were incremental, ROI would be ($8,000 − $1,500) / $1,500 = 433.3%.
The last line is conditional. Attribution alone does not establish incrementality. If only half the associated gross profit were incremental, the same calculation would produce ($4,000 − $1,500) / $1,500 = 166.7%. These are alternative assumptions, not two measurements.
You also cannot calculate a payback date from these totals. You need the dates on which costs and gross profit accumulated. A cost-to-profit ratio is not a number of months.
Distinguish attribution from causation
A sourced rule might require that the account had no active opportunity before identification and that the workflow produced the first qualified sales interaction. That makes reporting more consistent. It still does not prove the opportunity would not have happened otherwise.
For stronger evidence, design a holdout before launching outreach. Randomize at the account level where possible so contacts from the same company do not enter both groups. Use comparable eligibility and observation windows, and record any cross-group exposure.
A holdout that compares visitor-triggered follow-up with business as usual measures the incremental workflow effect. It may not isolate the identity provider from changes in rep behavior, response time or messaging. State what was actually tested.
Small samples and long sales cycles can leave the result inconclusive. Report the group sizes, elapsed time and observed outcomes rather than declaring a winner because one extra deal closed.
Keep sourced and influenced opportunities separate
An identified visitor from an existing opportunity may help the account owner prioritize a conversation. Record that as influence under a defined rule. Do not also count the entire opportunity as newly sourced.
Deduplicate opportunity IDs across people, campaigns and tools. Keep first identification time, first accepted action, opportunity creation and close date in the reporting record. Preserve the original campaign information rather than overwriting it with “visitor identification.”
For a monthly review, show the funnel counts, the rejection reasons, the amount spent and the maturity of each cohort. Financial results can follow later as deals close.
FAQ
Can I use pipeline to calculate ROI?
You can report a pipeline-to-cost multiple, but it is not realized financial ROI. Pipeline may never close, and closed revenue still needs a margin and incrementality treatment.
Is one trial enough to prove revenue lift?
Usually it is better suited to checking coverage, record quality and routing. The time needed to evaluate revenue depends on your traffic, sample size and sales cycle.
Should I include sales labor?
Include the labor attributable to running the workflow, using a stated allocation method. Otherwise two vendors with different manual review requirements can look artificially similar.
What if I cannot run a holdout?
Report attributed outcomes and the attribution rule. Use sensitivity scenarios for incrementality, label them as assumptions, and avoid presenting a causal revenue claim.
Put the model beside the workflow
Review Leadpipe Identification and the current plans, then start a trial with a named owner and a cohort definition. The trial is 7 days or 500 profiles, whichever comes first, no card. Measure usable records first and let financial outcomes mature.
