Campaign planning
Contextual campaign brief template
Organization byline: Measurement Press editorial desk. This page does not list an individual author or reviewer. Read the editorial method and corrections policy.
About 12 min read
A contextual media buy gets easier to measure when the brief names the reader job, eligible inventory, creative promise, primary outcome, and comparison rule before the first impression runs.
Use this template for private marketplace, sponsorship, newsletter, native, or display campaigns where the value of the buy depends on professional context. It is meant to keep media planning, creative handoff, reporting, and claim language aligned.
Picture the kickoff call: the seller has a strong audience story, the buyer has a renewal threshold, creative has two messages, and analytics needs source fields before trafficking starts. The brief is the shared record that keeps those decisions from being rebuilt from memory during the final readout.
The one-page brief
The brief should be short enough to read before a kickoff call and specific enough to survive trafficking, analytics exports, and the final report.
| Brief field | Write down before launch | Why it matters later |
|---|---|---|
| Decision owner | The person or team that will use the result, plus the budget, renewal, creative, or testing decision at stake. | Stops the report from optimizing for a metric nobody planned to use. |
| Reader job | The task the reader is performing: choosing a method, evaluating a vendor, checking a source, learning a failure mode, or planning a campaign. | Keeps contextual fit tied to reader intent rather than a broad audience label. |
| Eligible inventory | Topic desk, page set, URL pattern, placement IDs, formats, device classes, and exclusions. | Prevents delivery from drifting into unclear run-of-publication inventory. |
| Creative promise | The offer, message, asset type, destination, and the reader problem the creative claims to help solve. | Separates weak context from weak creative if response is lower than expected. |
| Primary outcome | One main signal selected before launch: qualified delivery, engaged visit, qualified lead, brand survey movement, or designed lift. | Reduces post-flight metric shopping. |
| Comparison rule | Holdout, matched context, prior flight, matched market, business-as-usual baseline, or explicit no-comparison status. | Sets whether the readout can use descriptive, directional, or causal language. |
| Decision threshold | The result that would trigger renewal, creative change, budget shift, deeper test, or stop decision. | Makes the recommendation auditable instead of mood-based. |
Reader context
Contextual buying is strongest when the campaign names the reader's job, not just the page category. A reader evaluating a measurement method is in a different mode from a reader scanning a case study or printing a worksheet.
| Reader job | Good-fit campaign role | Weak readout habit |
|---|---|---|
| Choose a measurement method | Introduce a guide, calculator, research asset, analyst service, or product page that helps compare methods. | Pooling method-selector readers with broad awareness traffic. |
| Evaluate a vendor claim | Offer a review worksheet, benchmark, demo, governance resource, or evidence-quality service. | Treating all source-evaluation traffic as the same intent. |
| Plan a campaign | Promote a practical template, webinar, planning guide, or operational workflow. | Judging the buy only by last-click form fills. |
| Study a failure mode | Pair a case study with a measured next step, such as a checklist or advisory consult. | Assuming deep reading means immediate purchase intent. |
| Check a public or media claim | Support source-quality, research, data, or professional learning offers with careful claim language. | Reading engagement as proof that beliefs changed. |
Inventory and creative handoff
A campaign brief should make trafficking boring. Anyone setting up the flight should be able to map the package, placements, creative, destination, and reporting keys without interpreting a sales paragraph.
1. Name the package and placement IDs.Use the package ID, deal key, placement IDs, accepted sizes, device rules, and page-set exclusions from the media kit or inventory matrix.
2. Version the creative.Record creative ID, size, message angle, offer, brand name, destination URL, active dates, and backup creative for mobile or lower-size placements.
3. Keep destination quality visible.Preserve landing page, form path, page version, UTM values, destination action, and any follow-up rules that affect lead or outcome quality.
4. Log material changes.Creative swaps, page edits, audience changes, paused delivery, tracking fixes, or makegood inventory should appear in the final readout notes.
Measurement plan
The measurement plan should match the decision. A sponsorship flight can be valuable with a descriptive readout, but it should not borrow causal wording from a test it did not run.
| Decision | Minimum evidence | Stronger design | Bounded readout language |
|---|---|---|---|
| Renew a contextual package | Qualified delivery, viewable impressions, frequency, engaged sessions, and destination quality by page set. | Matched prior flight or matched page group with similar creative and destination. | The package delivered the intended context and produced observed response under the stated definitions. |
| Choose a creative route | Delivery and response by creative ID, format, destination, and message angle. | Balanced creative rotation with enough delivery per cell. | The observed response favors this creative in this context; it does not isolate creative effect if rotation was uneven. |
| Judge lead quality | Lead source trail, qualification status, disqualification reason, follow-up status, and duplicate rules. | Matched baseline or holdout for downstream quality under the same sales process. | The campaign generated reported leads with the stated quality mix. |
| Estimate brand movement | Survey sample source, exposed/control balance, question wording, field dates, and uncertainty. | Pre-registered brand lift design with respondent-quality and leakage checks. | The survey suggests movement in the measured sample and window. |
| Claim incremental impact | A protected comparison, assignment or matching method, outcome window, leakage check, and uncertainty statement. | Randomized holdout, clean geo test, or matched-market design selected before results were visible. | The design estimates lift for this population, window, and outcome, with the stated limits. |
Primary source trail and a simulated brief gate
Platform field boundary: Google Ads contextual-targeting guidance names Topics, Placements, Display or Video keywords, and Exclusions as scope controls. Google Ad Manager key-value guidance describes page or page-part labels passed with ad tags, while its targeting-key reporting guidance says only keys configured as reportable can be included in reports. The MRC viewable-impression guidance keeps served, viewable, and invalid states distinct. These sources support field design; they do not establish audience identity, delivery quality, performance, or causal lift.
The handoff fields here follow primary platform and standards guidance. Google Ads placement guidance groups topics, placements, keywords, and exclusions as content-targeting controls, while Google Ads Display guidance distinguishes narrowing reach with “Targeting” from observing a segment without changing reach. The IAB Tech Lab Content Taxonomy guide provides a common language for page topic context and attributes such as format, language, and source. Google Analytics campaign URL guidance recommends consistent source, medium, and campaign fields and uses content values to distinguish creative variants. These sources support a cleaner brief and data trail; they do not establish campaign performance or causal lift.
| Source boundary | Supports in this brief | Still requires campaign evidence |
|---|---|---|
| Contextual targeting guidance | Write the eligible topics, placements, keywords, and exclusions as explicit scope controls. | Whether the selected page set reaches the intended professional reader or creates demand. |
| Content-taxonomy guidance | Use a shared vocabulary for page topic, format, language, and source attributes. | That a label is complete, accurate, brand-safe, or representative of every delivered page. |
| Campaign URL guidance | Keep source, medium, campaign, and content fields consistent enough to distinguish planned variants. | That the fields survive redirects, exports, consent states, and downstream joins without missingness. |
| Measurement boundary | Pre-register an outcome, window, comparison rule, and claim ceiling before launch. | Observed delivery or response proves performance, incrementality, or causal lift. |
Simulated worked example — not a real campaign, result, or forecast: Assume a hypothetical contextual package is reviewed against five launch records. The first four are complete, but the comparison rule is blank.
| Launch record | Simulated state | Why it changes the readout |
|---|---|---|
| Eligible context and exclusions | Pass | The page set, topic boundary, and excluded routes are written before trafficking. |
| Placement and device map | Pass | Placement IDs, accepted sizes, and device classes can be joined to delivery. |
| Creative and destination record | Pass | Creative ID, message, landing page, active dates, and backup path are stable. |
| Source and outcome fields | Pass | UTM fields, package ID, placement ID, primary outcome, and outcome window are named. |
| Comparison and decision rule | Missing | No holdout, matched baseline, prior-flight rule, or explicit no-comparison status is attached. |
| Total | 4 / 5 = 80% | Revise: packet completeness is not permission to use causal or lift language. |
The 80% is a reproducible handoff check, not a quality probability, performance estimate, or policy approval. The strongest supported readout in this simulation is descriptive delivery and observed response under the stated fields. The decision changes if the missing comparison is designed before launch, if the page set or destination changes, or if source fields fail to survive redirects and exports.
Check whether the brief is actually joinable
A filled brief is not yet a usable measurement record. Before launch, test whether each field group has a named value and a key that can survive trafficking and the final export. This separates a persuasive planning document from one that can support a defensible readout.
| Field group | Simulated state | Required join or decision key |
|---|---|---|
| Decision owner and threshold | Pass | Owner, renewal or stop threshold, and action if the threshold is missed. |
| Reader job and eligible context | Pass | Reader task, page-set rule, topic boundary, and exclusions. |
| Placement and device map | Pass | Package ID, placement ID, format, device class, and active dates. |
| Creative and destination version | Pass | Creative ID, message angle, destination URL, and page version. |
| Outcome and source trail | Pass | Campaign, package, placement, creative, source/medium, and outcome-window fields. |
| Comparison and claim ceiling | Missing | Protected comparison, matched baseline, prior-flight rule, or explicit no-comparison label. |
| Total | 5 / 6 = 83.3% | Hold causal language: the record is joinable, but the missing comparison limits the claim to descriptive or bounded directional language. |
Reproduce the score as passing field groups / 6. If the source trail cannot join one of the five passing groups, the sensitivity case is 4 / 6 = 66.7%; that is a data-handoff failure, not evidence that the campaign performed worse. In either case, do not rank contexts, creatives, or destinations as causal winners until the comparison rule is locked before delivery. The test also assumes stable naming across trafficking, analytics, and sales or survey exports; it does not prove that the fields are complete, accurate, or sufficient for statistical inference.
Pre-launch QA
| Question | Pass condition | If not |
|---|---|---|
| Can the campaign be reported by context? | Package, page set, placement ID, and device class are captured. | Report at package level and avoid ranking page contexts. |
| Can creative and destination be separated? | Creative ID, message, size, destination, and landing-page version are stable. | Do not call a message or format the winner. |
| Can bad exposure be filtered? | Measurable, viewable, invalid-traffic, and frequency fields are visible where available. | Downgrade to served-delivery language. |
| Can outcomes be joined without guessing? | Lead, sale, survey, or matchback exports carry campaign and source-trail fields. | Report observed outcomes with source-trail caveats. |
| Can the comparison survive launch? | The holdout, matched baseline, prior flight, or no-comparison label is locked before results. | Use descriptive language and plan a deeper test next. |
Report shell
Use the same headings in the final readout. The report should make the planned question, actual delivery, response quality, and evidence level visible before the recommendation.
- Campaign decision and threshold.
- Eligible context, delivered context, and any delivery changes.
- Creative, destination, and device delivery by placement group.
- Exposure quality, frequency, traffic quality, and lead or outcome quality.
- Comparison method and evidence level: descriptive, directional, or causal.
- Recommended action and what would need stronger evidence next.
Worked brief-to-readout decisions
The same media plan can produce different recommendation language depending on what the brief locked before launch. Use the table below to keep the final meeting from upgrading a descriptive report into a causal claim.
| Brief choice | What the report can say | What it should not say | Next stronger move |
|---|---|---|---|
| The package, placements, device class, creative IDs, and destination version were locked, but no comparison was planned. | The campaign delivered the intended context and produced observed response under the stated source-trail rules. | That the package caused the response or beat an alternative package. | Add a matched prior flight, matched page set, or holdout in the next brief. |
| Two creative routes were rotated, but one route received more premium placements and a stronger landing page. | The observed response differed by creative route, with placement and destination imbalance noted. | That the higher-response creative is the winner by itself. | Balance creative rotation by placement, device, and destination before reading the next result. |
| Lead quality was the decision threshold, but sales follow-up rules changed during the flight. | The report can describe submitted lead volume and the visible quality mix, then flag the process break. | That the buy created lower-quality or higher-quality leads without a stable follow-up process. | Lock qualification fields, duplicate rules, routing owners, and follow-up windows before launch. |
| A protected comparison, primary outcome, and outcome window were locked before delivery. | The report can estimate bounded lift for the defined population, window, and outcome, including uncertainty and leakage limits. | That the effect applies to every audience, format, or future package. | Repeat the design or expand it deliberately after the first decision is made. |
Pair with
Use this template with the media kit and ad inventory readiness matrix for placement IDs, package IDs, deal keys, and creative sizes. Use the contextual package proof checklist before pricing or trafficking a package, the contextual package proof sheet when the package needs a one-page buyer, operations, and analytics handoff, and the private marketplace package brief template when buyer-facing scope and renewal evidence need to be locked before activation. Use the creative asset acceptance checklist to approve display files, native fields, destination fit, and tracking before launch, then use the landing page launch QA worksheet to check page speed, forms, redirects, source fields, consent state, and routing. Use the campaign data-layer spec to preserve source-trail fields, the private marketplace campaign measurement checklist to set readout boundaries, the private marketplace readout export sample to define row-level delivery and outcome fields, the campaign reporting sample to structure the finished report, and the campaign readout QA checklist before a renewal or budget decision.