Media planning for programmatic: from client brief to a defensible site list
The short answer
Programmatic media planning is the work of turning a client brief into the websites, channels and deals a campaign will buy through automated platforms. A sound plan records the brief as structured fields, checks every candidate site for context, quality, suitability, selling path and real audience, then approves, versions and exports the list with a reason for each site.
Key takeaways
- A plan is only as good as its brief, so capture product, audience, topics wanted, topics to avoid, markets and languages as separate fields.
- Confirm the interpreted brief with a person before any search runs, because a misread brief produces a confident but wrong list.
- Judge each candidate on several sampled articles and six separate checks: context, editorial quality, suitability, seller authorization, audience in the market, and language and network.
- Treat the approved list as a versioned document with an owner, a change log and an export record for each campaign.
- An exported list is an input to activation, not proof of delivery; reconcile planned domains with served domains after launch.
What is programmatic media planning, and why does the site list matter?
Definition
- Programmatic media planning
- Choosing where a campaign will run before it is bought through automated platforms: which websites, apps, channels and deals, for which markets and languages, and why each one belongs on the plan.
Programmatic is now more than half of US internet ad revenue. The IAB and PwC full year 2025 report (opens in a new tab) puts US internet advertising revenue at $294.6 billion in 2025 and programmatic advertising revenue at $162.4 billion, up 20.5 percent year over year. Canada moved the other way in 2025: the IAB Canada 2025 revenue survey (opens in a new tab), released in June 2026, puts direct buying at 72.9 percent of its buying split, private marketplace at 22.4 percent and open RTB at 4.6 percent, in a split that excludes Google, YouTube and social media.
Open buying is becoming list driven too. The ANA's Q4 2025 programmatic transparency benchmark (opens in a new tab), released in February 2026, found that open marketplace buying among its participants became more selective, governed by inclusion lists and performance filters rather than scale. Whether a campaign runs through a deal or the open auction, the domain list behind it is a core planner deliverable.
What belongs in a programmatic planning brief?
A site list can go wrong at the brief, before any search runs. A client email that says "premium lifestyle sites in Canada" leaves every important decision to whoever builds the list. Break the brief into fields that a second planner could read and reach the same list.
| Field | Vague version | Usable version |
|---|---|---|
| Product | Coffee | Whole bean specialty coffee sold online by subscription |
| Audience | Adults 25 to 54 | Home brewers who read about equipment, roasting and cafe culture |
| Topics wanted | Food and drink | Brewing guides, grinder and espresso machine reviews, coffee origin stories |
| Topics to avoid | Anything negative | Health scare stories about caffeine; tragedy and disaster news |
| Markets | North America | Canada, with Ontario and Quebec as priorities |
| Languages | English | English and Canadian French, judged from the articles themselves |
| Channels | Digital | Open web display through a curated deal; YouTube planned separately |
| Constraints | Brand safe | Client tolerance set per sensitive category; exchanges the desk can buy on |
Write topics as descriptions of articles, not single keywords. "An in-depth review of burr grinders for espresso at home" tells a reviewer what a good page looks like. "Coffee" matches a commodity price report, a celebrity cafe photo and a recipe blog equally.
Why should a planner confirm the brief before searching?
Every brief gets interpreted, whether by a junior planner, a spreadsheet of keywords or software. The interpretation is where silent errors enter: a market quietly widened from Canada to North America, a topic to avoid dropped because it was mentioned in passing, a French requirement read as optional. Once the search runs, those errors look like results.
How do planners find candidate websites for a brief?
Good candidate lists come from several directions at once: sites the client already trusts, publishers that ran well on earlier campaigns, sites that rank for the article descriptions in the brief, and specialist publications that trade press and readers recommend. The aim of this stage is breadth. Filtering comes next, and it is easier to remove a weak site with a stated reason than to notice a strong one that was never considered.
Judge each candidate as a whole site, from several recent articles, not from its homepage or a single post. Count the share of sampled posts that match: a lifestyle site where two of thirty recent posts mention coffee once covered coffee, while a small blog where most posts are brewing guides is a coffee publisher. Sampling also protects against a site that changed hands and now publishes something else entirely.
Remove non publishers early: online stores, brand sites, marketplaces, link directories and sites that carry no advertising. They can look relevant to a topic search, but they are not publishers, and they crowd out the sites that are. For finding sites that resemble the ones a client already likes, see the guide to lookalike websites.
What should a planner check on every candidate site?
Keep the checks separate. A site can be perfectly on topic and still be a made-for-advertising site, or be well written and sold only through resellers. One overall "good site" score hides which of these is the problem.
Six checks before a site joins the plan
- Context: most sampled articles match the brief's topic descriptions, and the matching pages are recorded.
- Editorial quality: the articles are original, specific and dated, not templated or syndicated filler; see the guide to made-for-advertising sites.
- Suitability: no sampled page crosses the client's tolerance in any sensitive category, judged by meaning rather than keywords, as described in the brand suitability guide.
- Selling path: the site's ads.txt (opens in a new tab) file and the exchange's sellers.json (opens in a new tab) entry agree, and the planner knows whether the seller is direct or a reseller.
- Audience in market: public data such as the Chrome UX Report (opens in a new tab) shows real visits from the target country, wherever the site is based.
- Language and network: the articles are in the brief's languages, and sites sharing one owner's ad codes are counted as one network.
When a check cannot be completed, say so. A site whose ads.txt file could not be read, or whose audience data is too thin to show, belongs in an unverified group with the reason, not on the excluded pile. Treating missing data as a failure can quietly remove small and independent publishers from a plan.
How does PlacementGraph turn a brief into a vetted website list?
How should a site list be reviewed, approved and versioned?
A site list is a client deliverable. Someone other than its builder reviews it, and the approved version, not a later copy with untracked edits, is the one that reaches the buying platform.
Five steps from shortlist to approved list
- Review the reasons, not just the names. For each site, read the one line reason and open at least one of the matching pages the reason cites.
- Look at what was left out. Scan the excluded and unverified groups for strong sites removed by a blunt rule, such as a foreign domain ending or missing audience data.
- Record every manual change. When a planner adds or removes a site by hand, note who did it and why in the list itself.
- Approve a numbered version. The approver signs off on a specific version, and that version is the only one exported.
- Keep the old versions. When the client asks why a site ran in March, the March version and its reasons should still exist.
How does an approved site list move into buying platforms?
Each destination has its own documented format and limits, and import rules change. The Google Ads placement targeting guide covers Google's campaign types in detail.
| Destination | Documented format | Check after upload |
|---|---|---|
| Google Ads | A CSV for Google Ads Editor with one entity per row and columns such as Campaign, Ad group and Placement (Google Ads Editor Help (opens in a new tab)) | The number of placements imported matches the approved list |
| Display & Video 360 | App IDs and URLs entered one per line, up to 10,000 combined per line item, or a channel for larger lists (Display & Video 360 Help (opens in a new tab)) | The composite list download matches the approved version |
| Curated deal on an exchange | The exchange's own domain upload format plus a deal sheet with geography, languages and sellers | The deal ID, floor and eligible formats match the deal sheet |
What does a complete plan look like for one brief?
Fictional example
A Canadian specialty coffee subscription
A fictional Montreal coffee roaster, Kettlewick Coffee, briefs its agency for a fall display flight in Ontario and Quebec, in English and Canadian French, beside brewing guides and equipment reviews, avoiding caffeine health scares and disaster news. The planner turns the email into the eight fields above and the account lead confirms them, correcting the market from all of Canada to two provinces.
The candidate search gathers sites the client named, sites from a previous flight and specialist blogs found through article descriptions. Kitchen stores and an equipment marketplace drop out as non publishers. A US-based espresso review site stays in because public data shows Canadian visits. A small Quebec cafe culture blog has no readable ads.txt file and goes to the unverified group with that reason.
The reviewer opens the cited articles, restores one blog that a keyword rule had removed for the phrase "caffeine addiction" in a post about cutting back on espresso, and approves version 3. The desk uploads version 3 to its buying platform and records the import count beside the version number.
Guides in this hub
Every guide in the Programmatic media planning hub, each written to answer one question in full.
- GlossaryWhat is an ad impression, and when does it count?What counts as an ad impression under the MRC and IAB guidelines, how a request, a won bid and a rendered ad differ, and which count a plan reports.
- ExplainerWhat is ad inventory availability, and how do buyers confirm it?What avails are, how buyers confirm a site has buyable inventory in the right formats and markets, which assumptions go in the plan, and why sites go dark.
- FrameworkAgency programmatic governance: who decides, who checks and what gets reportedA programmatic governance framework for agency leads: the agency standard, client rules, approval rights, audit cadence, escalation and client reporting.
- ExplainerWhere does AI fit in programmatic media planning, and where does a planner stay in charge?Where AI helps programmatic planners today, what IAB and PwC say, which steps keep a person in charge, and how to brief an AI assisted planning workflow.
- ExplainerWhat do attention metrics measure, and how far should a planner trust them?What attention metrics measure, what the IAB and MRC guidelines of November 2025 require, how attention differs from viewability, and what to ask a provider.
- How-toHow do you write avoid topics in a media brief so a planner can apply them?How to write avoid topics in a media brief a planner can apply: scope words, in and out examples, what stays allowed, section level limits and who signs off.
- GlossaryWhat is budget pacing in programmatic, and how do buyers keep it on track?How budget pacing works in programmatic buying, how even and ASAP pacing differ, why spend can pile onto a few sites in a list, and how to fix underpacing.
- GlossaryWhat is cost per completed view, and how should buyers read it?What cost per completed view means for open web video, how a complete is counted, how completion rate sets CPCV, and why instream and outstream differ.
- GlossaryWhat is dayparting in advertising, and when does it help a plan?What dayparting is, how time zones across the US and Canada change a schedule, when day and hour rules help a small site list, and how to test one fairly.
- GuideHow do you explain site selection to a client, one site at a time?How to write a reason for each shortlisted website that a client can check: evidence to cite, the weakest point to name, length, and how to explain cuts.
- GlossaryWhat is frequency capping, and how should a plan set it?How frequency caps work in Google Ads and Display & Video 360, how they count without third-party cookies, why they leak and how to set a starting cap.
- GlossaryWhat are the standard IAB ad sizes, and which ones does a plan need?The standard IAB display ad sizes, what the IAB New Ad Portfolio changed, which sizes fit phones, and how a size list decides which websites can serve a plan.
- GuideWhat changes when a brand takes programmatic in house?What changes when a brand takes programmatic in house: which planning tasks move, which approvals stay, the skills to hire, and what to keep from the agency.
- How-toHow do you build, size and maintain an advertising inclusion list?How to build a website inclusion list, size it against the budget, load it so platforms respect it, and keep it current with triggers and served domain checks.
- How-toHow do you write a programmatic media brief that picks the sites?A programmatic media brief template where every field drives a site selection decision: what to write, what a reviewer checks and what breaks when it is blank.
- GlossaryWhat is media flighting, and how do planners schedule flights?What media flighting means, how continuous, flighted and pulsed schedules differ, how flights change frequency and delivery, and how to plan them by market.
- FrameworkMedia plan forecast ranges: low, expected and high cases with stated assumptionsA framework for media plan forecasts as low, expected and high cases: the assumptions to state, how to compare forecast with delivery, and when to reforecast.
- ChecklistMedia plan handoff: what the planner gives the buyer, and what comes backA media plan handoff checklist: what the planner gives the buyer, how to separate approved sites from buyable inventory, and what the buyer confirms back.
- ChecklistMedia plan QA checklist: what to verify before a programmatic campaign goes liveA prelaunch media plan QA checklist for programmatic campaigns: list attachment, exclusions, deals, frequency, tracking, naming and who should run each check.
- ExplainerWhat does open exchange programmatic mean for site quality?What buying on the open exchange means for site quality, the controls buyers use to govern it, their strengths and limits, and when a deal is the better route.
- GlossaryWhat are outstream video ads, and how do buyers verify them?Outstream video ads play outside a stream of video content. Learn the OpenRTB 2.6 plcmt values, why they replaced placement, and what to check before buying.
- GuideWhich programmatic advertising KPIs fit each campaign objective?How to choose programmatic advertising KPIs for awareness, consideration and action goals, why cost metrics alone mislead, and how to pair quality with cost.
- ChecklistProgrammatic campaign audit: what to check after the flight endsA post campaign audit checklist for programmatic buys: partner data to request, how to test fees against invoices, and which findings change the next plan.
- GuideProgrammatic client report: where ads ran, what changed and what is not knownWhat a programmatic client report should show: where ads ran, exclusions and why, evidence, changes and open risks, with metrics defined and unknowns stated.
- How-toWhat goes in a programmatic media plan, section by section?A media planning template for programmatic, section by section: objective, sites with reasons, exclusions, flights and budget, reporting and approvals.
- GuideWhat does a programmatic media planner do, from brief to approved list?What a programmatic media planner does: the decisions they own from brief to approved list, the evidence to keep for each site, a weekly routine and key skills.
- ExplainerHow does programmatic native advertising work, and what makes it trustworthy?How native ads are bought programmatically: the OpenRTB Native request, where in-feed and in-content units appear, FTC disclosure guidance and a review list.
- ChecklistWhich programmatic planning mistakes repeat, and how do you fix them?A post mortem checklist of programmatic planning mistakes: unconfirmed briefs, padded lists, stale versions, geo and language gaps and unchecked sellers.
- GlossaryWhich roles run a programmatic campaign, and who decides what?Programmatic team roles defined: planner, buyer, trader, ad operations, analyst and client lead, who decides what, and a RACI table for site list decisions.
- ExplainerHow does programmatic video advertising work?How programmatic video is bought across instream, outstream and CTV, the OpenRTB fields that tell them apart, and how planners pick and verify video placements.
- GlossaryWhat do reach and frequency mean, and how are they calculated?Reach, average frequency, effective frequency and GRPs defined with the MRC formulas, worked fictional arithmetic, and how a site list's size shapes reach.
- ExplainerHow does real time bidding work, one auction at a time?One real time bidding auction in order: ad request, bid request, bid response, win, markup and billing notices, with timings and where a site list acts.
- GuideHow should retailers plan programmatic advertising around peak season?A guide to planning programmatic for retailers: 2026 US and Canadian peak dates, publishers that fit retail briefs, and how to treat store and coupon sites.
- How-toServed domain reconciliation: matching the approved list to where ads ranHow to reconcile planned domains with the domains that served: which log fields to pull, how to normalize URL variants, and what to do about off list delivery.
- How-toHow do you keep version control on a site list?How to version a programmatic site list: a naming pattern, what a change log records, how to link each campaign to its list, and how to roll back safely.
- FrameworkWhat should a media plan do with sites that cannot be checked?What to do when a candidate website cannot be checked: keep missing data as missing, hold the site apart, retry on a rule, and review by hand before approval.
- How-toHow many pages should you read before judging a whole website for ads?How many pages to read before judging a whole website for ads, how to pick them by recency, section and popularity, when to read more, and how to record it.
- GlossaryWhat is a DSP, and what does it decide for a buyer?What a demand side platform does at each step: seats, line items, targeting, bids and reports, how it decides to bid, and which calls stay with the planner.
- GlossaryWhat is an ad server, and what do its reports tell a buyer?What publisher and advertiser ad servers do, how an ad server picks which ad runs, what an ad tag does, and why ad server and DSP counts can disagree.
- GlossaryWhat is CPM, and why is a lower one not always better?CPM, viewable CPM and effective CPM defined, with the arithmetic, how floors and deal prices are set in CPM, and why a lower CPM is not a better campaign.
- ExplainerWhat is programmatic advertising, and what does the planner decide?What programmatic advertising is, how one auction runs from page load to ad, what a bid request says about a site, and what the planner decides at each step.
- GlossaryWhat is viewability, and what does it not prove?The MRC viewable impression rules for display and video in the standard's own words, how viewability is measured, and the questions to ask before a deal.
Frequently asked questions
How many websites should a programmatic site list have?
As many as qualify on all the checks, and no more. A tight list of strong specialist publishers is easier to defend than a padded one. If the list is too short to deliver the budget, widen the brief deliberately, for example with adjacent topics tested on a small budget, rather than lowering the bar for the core list.
Should a plan use an inclusion list or an exclusion list?
Both, for different jobs. An inclusion list defines where a contextual brief may run, so every site on it was chosen for a reason. An exclusion list protects the whole account, including campaigns that run without an inclusion list, but it only blocks domains someone has already found.
Who should approve a site list before it is exported?
Someone other than its builder, with authority to speak for the client's standards, such as the account lead or a senior planner. The approval should name a specific version, and only that version should be uploaded. Record the approver, the date and any sites added or removed by hand.
Does programmatic media planning need cookies or personal data?
Choosing sites by their content does not require tracking individual people. A site list built from page meaning, editorial checks, seller files and public popularity data selects places, not persons. Any audience data layered on top at activation carries its own privacy obligations, which depend on the data and the jurisdiction.
Sources
- Internet Advertising Revenue Report: Full Year 2025 (opens in a new tab)IAB and PwC, April 2026
- 2025 IAB Canada Internet Ad Revenue Survey and 2026/27 Forecast (opens in a new tab)IAB Canada, June 2026
- ANA Q4 2025 Programmatic Transparency Benchmark Shows Media Quality Is Now Driving Measurable Performance Outcomes (opens in a new tab)ANA, February 25, 2026
- ads.txt (Authorized Digital Sellers) (opens in a new tab)IAB Tech Lab
- sellers.json (opens in a new tab)IAB Tech Lab
- Overview of CrUX (opens in a new tab)Google Chrome for Developers
- CSV file columns (opens in a new tab)Google Ads Editor Help
- App, channel, and website targeting (opens in a new tab)Display & Video 360 Help
Researched and written by PlacementGraph Research. Read how we research, source and update these pages. Corrections: diana@placementgraph.com.