Use an audience waterfall to turn an intent score into a responsible advertising plan. Count the scored records, apply the client's campaign eligibility rules, and then track which approved records the destination accepts and actually reaches. A small deliverable group can reflect a strict use policy, a platform requirement, limited inventory, or a technical mismatch. Each reason calls for a different response.
Where the decision gets complicated
Imagine an agency working with a Pennsylvania business that installs dust collection equipment. The agency has a high-scoring segment of manufacturers and a separate first-party customer file collected directly through the business. Some manufacturers belong to companies the business already serves. Some have no approved identifier for the proposed destination. One score threshold can group accounts by relevance while an entirely different rule governs whether a specific activation path is allowed.
Cortex Signal Flow carries forward existing intent scores and premade segments. That can save audience design work. Confirm the destination, source terms, identifier type, use permission, and account eligibility before activation. Build a waterfall with a distinct count and timestamp at every step. If you can't document why a set of records moved from one stage to another, pause before committing budget based on the final total.
What to record at handoff
In a planning sheet, count account candidates, then count eligible identifiers separately. Next add a disposition for account approval, destination acceptance and delivery. Write the relevant data purpose beside each transition. A lead scoring feed can guide a message strategy without supplying approved identifiers for audience matching.
Keep a sample with an active customer, an out-of-territory company, a suppressed identifier and an account whose relationship is unresolved. Exercise the checks using authorized test data. If a prohibited record passes, pause the proposed route even if the eventual audience would meet the client's budget target.
Selection rule
Specify the equipment problem, site geography, score snapshot, and exclusion dates. Store a compact reason for why accounts are included. Put the next step in the operating notes. Keep a separate view for high-priority accounts that lack an eligible route.
Use the selection rule check to decide which owner needs the case next. An approved account-level signal can help a seller prepare, while a technician may need a verified site condition before accepting work. Write down what remains missing so the next owner does not mistake the signal for a confirmed project. A seller can ask a neutral question that helps the prospect describe the actual situation without asserting a browsing history or a level of readiness the evidence cannot establish. Ask whether the resulting next step answers the question the buyer may reasonably have.
Choose the route before asking for an upload file
An advertising destination can support contextual targeting, first-party audience matching or other audience types under different rules. Do not assume one approval applies across them. Put the proposed route, identifier lineage, collection purpose and source terms into the request before building a file. If account identifiers will be converted to another grain, describe that conversion and its limits. The agency needs to know whose data it is handling and what the platform expects. An approved research segment can still be useful in creative planning even when the activation route is unavailable.
Identifier grain
Use feedback about the identifier grain check to refine which evidence earns the next step. Translate the audience's unit into the actual identifier requested by the destination. If an account has several possible contacts, a count of contacts will not reconcile directly to the account total.
The review record for the identifier grain check should retain this disposition. Keep a per-grain count through the handoff and review ambiguous relationships. Record the reasons from the identifier grain check for cases the team does not advance. When a case is relevant but cannot be served, that explanation can support offer planning. When the research meant something else, it may point to a topic definition that needs correction. Save the unresolved question beside the decision so the next reviewer can reproduce the result.
Turn the identifier grain check into a short review record with the observed evidence, the proposed action and its owner. The next operator should be able to continue the work without being asked to infer missing details. If a check is unresolved, note which subsequent action must wait and what evidence can release it.
Follow two fictional records
A scored manufacturer has two approved email subscribers and one purchased contact whose use is limited to research. The agency may be able to use the account score to plan creative while excluding the purchased contact from the proposed upload. Apply the destination's actual policy before doing anything with the subscribers.
The resulting approved identifiers might represent fewer accounts than the planning segment. Keep the planning total and approval count in the brief. A rejection in a later stage should be assigned to its actual cause rather than counted as lower intent.
Source separation
A third-party score used to inform planning and a customer email uploaded for matching involve different purposes and checks.
Put the source separation check in the intake brief as a business question someone can answer. In discovery, the source separation check can suggest a helpful question and a clear reason for asking it. The buyer can supply a missing operational detail without being told what a source inferred about them. Record the confirmed answer separately from the original research cue. For this case, the expected next step is clear. Record lineage so the agency can explain the source of each activation identifier.
Try an adjacent case that differs on one important fact.
Agree on the point at which the source separation check becomes a decision someone can act on. Earlier evidence can stay visible in the approved working record. The final disposition should say which requirement was satisfied. A later change can then update the decision without turning earlier uncertainty into a historical fact.
Destination acceptance
Count records accepted by the platform separately from people or accounts it can match. Confirm whether the platform even returns the denominators you need. To close the review, write down what the decision owner is expected to do. Label undocumented stages as unavailable and avoid treating an acceptance response as delivered reach.
Use the destination acceptance check to decide which owner needs the case next. An approved account-level signal can help a seller prepare, while a technician may need a verified site condition before accepting work. Write down what remains missing so the next owner does not mistake the signal for a confirmed project. That work can be done on a small sample before asking a provider for more records. A documented exception can remain useful evidence while it waits for an approved decision.
Estimate delivery from the last stage you can substantiate
If a platform reports accepted rows but offers no stable match denominator, report accepted rows as accepted rows. If it reports matched audience size with a minimum threshold, explain how withholding affects your estimate. Reach is a later measure that depends on pacing and inventory as well as audience size. For planning, give the client a range based on approved inputs rather than assigning every scored account an equal probability of delivery. Verify where exclusions apply after the campaign begins, especially for current customers and withdrawn permissions.
Delivery reality
Agree on who can settle the delivery reality check. Measure reach or exposures using a known definition and retain cost separately. Tightening a score threshold may narrow eligible delivery beyond the budget model's assumptions.
Apply the rule while the source evidence is still available. Adjust the proposed budget or test a properly approved alternative audience before changing the data rules. Record the reasons from the delivery reality check for cases the team does not advance. When a case is relevant but cannot be served, that explanation can support offer planning. When the research meant something else, it may point to a topic definition that needs correction. Where a previous result remains relevant, keep it as a separate snapshot rather than restating it under the new rule.
Before treating the delivery reality check as ready for routine use, agree how to handle evidence that arrives after the initial decision. Identify which current state can change and which original snapshot remains frozen. This preserves the difference between the latest information and the evidence available when the earlier action was approved.
The outside evidence and its limits
Google's Customer Match policy requires qualifying first-party collection for that specific use and also sets account eligibility conditions. Confirm source and destination requirements before making an upload decision. This policy does not document a Cortex integration. See the Google Customer Match policy.
What the activation readiness denominator includes
The following example is hypothetical and uses scored accounts in each fictional audience as the denominator. It counts cases that passed documented eligibility review for the selected route. Save the underlying counts together with the label used to report activation readiness. Totals and rates will answer different questions when group sizes differ.
| Illustrative group | Count meeting rule | Eligible count | Rate |
|---|---|---|---|
| Stricter scoring segment under one approval rule | 84 | 210 | 40.00% |
| Broader first-party-led planning segment under the same approval rule | 99 | 180 | 55.00% |
The first result is 84 out of 210, or 40.00 percent. The second is 99 out of 180, or 55.00 percent. On unrounded rates the first is 15.00 percentage points below the second. If you present a relative change, keep the comparison denominator explicit. Using 55.00 percent as the reference gives a difference of about 27.3 percent in relative magnitude.
The broader planning segment yields more approved records in this example despite the lower apparent selectivity. These two groups may differ in provenance and are not a test of score quality.
Record the snapshot date beside the activation readiness result. For this hypothetical comparison, use the same written review standard for both groups. If a reviewer changes the comparison rule, keep the original activation readiness calculation and clearly identify the revised version.
Use activation readiness to decide which design deserves a buyer-facing test. Keep the reviewed examples beside the comparison and consider where they came from. A narrow sample can highlight a classification problem even if it cannot describe a wider market. Plan a repeat review with a different eligible sample before making a large source or campaign commitment.
Reconcile contacts and accounts without mixing the denominators
One authorized account may contribute several contacts. Several records may also resolve to the same destination unit. Count each level separately and state which is relevant to the campaign. If you start with accounts and end with people, a percentage between those counts can be misleading. Add an approved transition report rather than forcing them into one funnel. Keeping separate counts also helps a buyer recognize where an apparent scale problem is actually a scope or definition problem.
The activation readiness counts leave 126 cases that did not meet the rule in the first fictional group and 81 in the second. In actual data, inspect those remaining cases before calling them all failures. For example, some may be awaiting a decision, while others may be complete and outside the rule. Keep a reconciliation showing the final and unresolved dispositions. That distinction matters more than rounding the percentages to another decimal place.
Give the owner of activation readiness one concrete buyer action to review alongside the source evidence. A topic map should help the team decide how to educate, qualify or route a prospect. It can remain useful at account level even if no approved person-level action exists. Verify the offer fit and use boundary separately.
Keep an approval date for source and destination eligibility because either can change. If a client withdraws a record from a permitted use, document which audience refresh or exclusion process applies. Verify its behavior with approved test data before reopening the route.
For more on the signals behind this decision, read the Cortex guide to buyer intent data.
Take the strategy into a buyer conversation
Authorize audience activation only on an eligible route and budget from the resulting delivery opportunity. Draw the waterfall with the evidence you actually have at each stage. An approved identifier file, an accepted destination response and a reach report support different decisions. Bring the weakest link in that chain to the review before asking for a larger media budget.
Explore how Cortex Marketing AI uses buyer research and feedback to refine priorities for a specific offer.
Bring one buyer problem, a sample of approved source evidence, and the sales decision you want to improve. For help designing a test around that decision, call Cortex Marketing AI to book an intent consultation.
