Campaign measurement
Campaign reporting terms glossary
Campaign reports often use familiar words as if everyone agrees on what they prove. This glossary keeps delivery, exposure quality, traffic, lead quality, matchbacks, attribution, and incrementality in separate lanes.
Use it before a campaign readout becomes a renewal recommendation. The point is not to make every report longer. The point is to keep each term attached to the evidence it can actually support.
Start by naming the lane
| Report lane | What it can usually support | Question before renewal |
|---|---|---|
| Delivery | Whether the campaign ran in the planned contexts, dates, placements, and device classes. | Did the delivered campaign match the package that was sold and approved? |
| Exposure quality | Whether impressions had a measurable, viewable, and valid opportunity to be noticed. | Which impressions were excluded, unmeasured, filtered, or concentrated? |
| Traffic quality | Whether clicks or visits showed useful on-site behavior after the ad interaction. | Did the destination and qualification rule match the campaign objective? |
| Lead quality | Whether form fills, registrations, or inquiries became useful buyer-side records. | Which leads were qualified, disqualified, missing status, or still in follow-up? |
| Matched outcomes | Whether exposed or tracked records could be joined to later outcomes. | What happened to unmatched people, duplicate outcomes, and prior intent? |
| Incrementality | Whether a designed comparison estimates what changed because of the campaign. | Was the counterfactual set before results were visible? |
Delivery and exposure terms
Package ID
The stable name for the sold context, placement set, and buying lane.
Do not use it as: proof that every impression reached the intended reader job.
Ask: Which URLs, placements, devices, and exclusions were included?
Placement ID
The specific ad location used for trafficking and reporting.
Do not use it as: a substitute for page context, creative size, or device mix.
Ask: Did this placement carry the same creative and destination as the others?
Served impression
An ad delivery event recorded by the ad system.
Do not use it as: proof that a person saw the ad.
Ask: How many served impressions were measurable and viewable?
Measurable impression
An impression where the system could evaluate viewability or exposure conditions.
Do not use it as: the whole delivery universe if unmeasured inventory is material.
Ask: Are viewability rates calculated on served or measurable impressions?
Viewable impression
An impression that met the report's stated viewability definition.
Do not use it as: evidence that the ad caused a response.
Ask: Which definition, device class, and measurement vendor were used?
Invalid traffic filtration
The process used to exclude traffic judged invalid under the report's rules.
Do not use it as: a guarantee that the remaining traffic is equally valuable.
Ask: Were filtered counts and denominator changes shown?
Frequency distribution
How often measured users, households, devices, or IDs were reached.
Do not use it as: a single average that hides high-frequency tails.
Ask: What share of delivery sat above the planned frequency range?
Context group
The editorial or page-intent grouping used to interpret delivery and response.
Do not use it as: a broad audience claim detached from the actual pages.
Ask: Which page set produced the result, and was it mixed with unlike contexts?
Traffic and destination terms
Clicks and visits are most useful when the report shows the destination, qualification rule, and reader action after arrival.
| Term | Plain meaning | Weak readout move | Better readout move |
|---|---|---|---|
| Click | A recorded ad interaction that sent the user toward a destination. | Equating click volume with qualified interest. | Report clicks beside destination type, device class, and qualified visit rate. |
| CTR | Clicks divided by reported impressions under the chosen denominator. | Ranking contexts by CTR alone. | Show denominator, creative theme, placement, and visit quality before ranking. |
| Qualified visit | A visit that met pre-set engagement, geography, bot-filter, content-depth, or landing-page rules. | Choosing the qualification rule after results are known. | Define the rule before launch and report excluded visits separately. |
| Engaged session | A session that crossed a time, depth, event, or return-visit threshold. | Treating engagement as a business outcome. | Use it as a traffic-quality diagnostic that may justify a deeper test. |
| Destination friction | Drop-off caused by page speed, form length, unclear offer, device fit, or tracking failure. | Calling a placement weak when the landing page changed the response. | Separate media delivery from destination performance before judging either. |
Outcome and attribution terms
Form fill
A submitted lead, registration, demo request, or contact form.
Do not use it as: proof of new demand.
Ask: Was the person already known, already in market, or already in follow-up?
Qualified lead
A lead that met the buyer's stated fit, intent, or follow-up criteria.
Do not use it as: causal proof that the campaign created the opportunity.
Ask: Which qualification rule was applied, and how many records lacked status?
Disqualified lead
A submitted record that failed fit, intent, duplicate, geography, account, or quality rules.
Do not use it as: a hidden adjustment removed from the readout.
Ask: Were disqualification reasons reported by context and creative theme?
Matched conversion
An outcome joined to an ad exposure, click, household, device, account, or customer record.
Do not use it as: the same thing as incremental conversion.
Ask: What was unmatched, duplicated, excluded, or already likely to convert?
Attributed conversion
A conversion credited to a campaign under a reporting rule or model.
Do not use it as: proof that the credited touch caused the conversion.
Ask: Is the report assigning credit or estimating lift?
Incremental conversion
An estimated conversion that would not have occurred without the campaign.
Do not use it as: a label for all conversions after exposure.
Ask: What holdout, geo comparison, matched baseline, or model identifies the counterfactual?
Comparison and decision terms
| Term | Use when | Report language it can support | Limit to show |
|---|---|---|---|
| No comparison | The report only shows delivery, response, and outcome counts. | Observed response in the measured campaign. | No causal or directional claim about what changed. |
| Prior flight baseline | The new flight is compared with an earlier campaign or period. | Directional change against that prior setup. | Seasonality, creative, destination, pricing, and audience changes. |
| Matched context | Similar pages, markets, accounts, or audiences were selected for comparison. | Directional difference under stated matching rules. | Unmatched differences and whether the match was chosen before results. |
| Holdout | Eligible users, geographies, accounts, or markets were protected from exposure. | Estimated lift for the tested population and window. | Leakage, power, compliance, contamination, and generalization. |
| Decision limit | The report states what action the evidence is strong enough to support. | Renew, revise creative, fix tracking, run a lift test, or stop the flight. | The decision should not exceed the comparison quality. |
Careful wording for reports
| Metric appears strong | Careful sentence | Avoid writing |
|---|---|---|
| High viewable rate | The campaign delivered a strong opportunity for measurable exposure in these placements. | The campaign worked. |
| High CTR | The creative and context produced more observed clicks under this denominator. | This context created the most demand. |
| High qualified visit rate | Traffic quality was stronger for this destination and qualification rule. | The audience was more valuable in every sense. |
| High form-fill count | The campaign produced more submitted records before lead-quality review. | The campaign generated incremental pipeline. |
| High matched conversion count | More outcomes matched to tracked records under this join rule. | The campaign caused those conversions. |
| Positive lift estimate | The test estimates lift for this design, eligible population, window, and uncertainty range. | The same lift will apply to all future spend. |
QA pass before a renewal
- Circle every term that sounds causal: drove, generated, incremental, impact, lift, attributed, produced.
- Rewrite the result as delivery, exposure quality, traffic quality, lead quality, matched outcome, or incremental effect.
- Check whether the denominator is served impressions, measurable impressions, viewable impressions, visits, leads, accounts, or eligible customers.
- Separate placement, creative, destination, device, and context before naming a winner.
- Use causal language only when the comparison was designed before results were visible.
Pair with
Use this glossary with the campaign reporting sample for report structure, the private marketplace readout export sample for row-level package, placement, device, creative, traffic-quality, and outcome-status fields, the campaign readout QA checklist after results arrive, the campaign KPI dictionary when naming metrics, the campaign data-layer spec before launch, the identity matchback checklist for joined outcomes, and the private marketplace campaign measurement checklist when planning contextual packages.