What does a programmatic media planner do, from brief to approved list?
Part of Programmatic media planning
The short answer
A programmatic media planner decides where a campaign may run before it is bought through automated platforms. The planner reads the brief, sets the selection criteria, chooses and defends each website, channel or deal, and gets the list approved. The buyer then decides how to buy it. The planner's lasting output is evidence: why each site is there.
Key takeaways
- The planner owns what to buy and why; the buyer owns how to buy it, at what price and pace.
- A planner's real deliverable is a list with evidence, so every site can be explained months later.
- Per site evidence covers fit, editorial quality, suitability, audience in the market and the selling path, each dated.
- During a flight the planner reads delivery against the approved list and decides changes; the buyer proposes and applies them.
- Core skills are judgment about content, clear writing and a working knowledge of supply chain files.
What does a programmatic media planner do?
Definition
- Programmatic media planner
- The person who turns a client brief into the set of websites, apps, channels and deals a campaign is allowed to buy through automated platforms, and who keeps the reasons for each choice.
Job titles vary. In a large agency the planner may sit on a planning team separate from a trading desk; in a small shop one person plans in the morning and buys in the afternoon. The role is defined by the decisions, not the title. Whoever answers the question "why is this site on the plan?" is acting as the planner, and that person needs the evidence to answer it.
The planner's work starts when a brief arrives and ends, for each version, when a list is approved and handed to the buyer. It starts again whenever delivery data, a client change or a site's own change calls the list into question. The planning steps themselves are covered in the programmatic media planning guide; this page is about the role: what the planner decides, keeps and does each week.
Which decisions belong to the planner and which to the buyer?
One workable split gives the planner every decision about where and why, and the buyer every decision about how, how much and how fast. Shared decisions need a named tie breaker, such as the account lead, written down before the flight starts.
| Decision | Planner | Buyer | If they disagree |
|---|---|---|---|
| How the brief is read | Writes the structured version and gets the client to confirm it | Flags fields the platforms cannot enforce | Account lead asks the client |
| Selection criteria | Sets the bar for fit, quality, suitability and market | Says which criteria need a platform setting as well | Planner decides; buyer records the risk |
| Core sites and test sites | Separates the core list from adjacent ideas for a small budget | Sets the budget split between them | Account lead |
| Sites that could not be checked | Decides whether each one waits, is reviewed by hand or is left out | Reports whether the site is buyable at all | Planner decides |
| Buying route | Says which sites need a deal and which can run on the open auction | Requests deals, chooses seats and sets floors | Buyer decides on price; planner on inclusion |
| List size against budget | Reports how many sites qualify | Says whether that list can deliver the budget | Client chooses between a wider brief and a smaller budget |
What evidence should a planner keep for each site?
Evidence is what lets someone other than the planner reach the same decision. A site's reason should survive a client asking about it months later, after the planner has moved accounts. Keep the record in the list itself or in a file named with the list version, never in a personal inbox.
| Item | What to record | Why it matters |
|---|---|---|
| Fit | The URLs of the sampled articles that match the brief, and how many were sampled | Shows the site covers the topic, not just one lucky page |
| Weakest point | The single best reason a skeptic would give to remove the site | Lets the approver weigh the risk without re-reviewing |
| Suitability | Any sampled page near the client's tolerance, with the category | Explains why a site near a sensitive topic was kept or cut |
| Audience in the market | The public popularity signal used and its month | CrUX country data (opens in a new tab) is inferred from IP addresses and is only in BigQuery, so the source must be named |
| Selling path | Whether ads.txt lists a DIRECT or RESELLER account for the planned exchange | In the ads.txt specification (opens in a new tab), DIRECT means the publisher directly controls the account |
| Decision and date | Kept, kept for a test, held for review or left out, with who decided and when | Ties every site to a list version and an owner |
Seller details beyond the summary belong with the buyer. A planner who records that a site is sold directly on the planned exchange, and that the exchange's sellers.json (opens in a new tab) lists the account as PUBLISHER, has enough to defend the inclusion. The full path audit, with seller IDs and log samples, is the trading desk's record.
What does a planner's week look like during a campaign?
Once a flight is live, the planner's job shifts from choosing sites to reading what happened on them. The routine below assumes one active flight and a weekly client call. A planner running several accounts will compress it, but the order holds: read delivery, check changes, decide, record.
| Day | Planner's task | Input |
|---|---|---|
| Monday | Compare served placements with the approved list and note any domain that served but was not on it | The buyer's placement export; Google's Where ads showed report (opens in a new tab) is not an exhaustive list and can omit low volume placements |
| Tuesday | Review listed sites that received little or nothing and ask the buyer whether the cause is bids, deals or supply | Deal impression volume, which buyers can see in Display & Video 360's deal delivery view (opens in a new tab) |
| Wednesday | Re-sample a handful of sites, starting with the ones that carry most impressions | Fresh articles from each site |
| Thursday | Decide changes, write the reason for each and prepare a new list version if needed | Monday to Wednesday notes |
| Friday | Brief the client on what changed and why, then hand the new version to the buyer | The change log for the version |
Some inputs move on their own calendar. CrUX data on BigQuery (opens in a new tab) is released on the second Tuesday of the following month, so audience checks are a monthly task, not a weekly one. Changes that do not need a decision this week go to a holding list for the next version rather than triggering a new one.
Which skills does a programmatic planner need?
Skills a programmatic planner uses every week
- Reads a site as an editor would: can tell a specialist publication from a site that covered the topic once.
- Writes a reason in one sentence that a client, an approver and a buyer all understand.
- Turns vague brief language into criteria a second planner would apply the same way.
- Reads an ads.txt file and a sellers.json entry without help, and knows what each can and cannot prove.
- Knows which platform controls exist for each surface, so criteria that need a setting are flagged for the buyer.
- Treats missing data as missing, and says so, instead of quietly removing small publishers.
- Keeps versions and change logs without being reminded.
Formal credentials exist but are not required for the job. The IAB offers the Digital Media Buying and Planning Certification, which the US Department of Labor's O*NET certification listing (opens in a new tab) describes as requiring a multiple choice exam and renewal every two years. A credential shows shared vocabulary; the evidence record shows judgment.
What does the planner role look like on a real account?
Fictional example
A planner for a Michigan appliance repair service
A fictional appliance repair company serving Detroit and Grand Rapids asks for a fall display plan in English, beside home maintenance guides, appliance troubleshooting articles and local news. The planner writes the structured brief, and the client corrects one field: Grand Rapids is a priority, not an add on.
The planner approves 46 sites in version 1, with six adjacent home cooking sites marked as a test. Three local news sites have no audience signal for the month, so the planner holds them for a manual review rather than cutting them. The buyer sets up a deal with a regional newspaper group and runs the rest on the open auction.
In week two, Monday's comparison shows a recipe aggregator that was never on the list, served through a broad ad group the buyer had added. The planner asks for it to be removed, notes that two listed troubleshooting sites received nothing, and learns from the buyer that neither is sold on the planned exchange. Version 2 moves both to a hold list with that reason, and the client hears about it on Friday.
How does PlacementGraph support a planner's evidence?
Frequently asked questions
Is a programmatic media planner the same as a media buyer?
Not as a role, even when one person does both. The planner decides where a campaign may run and why; the buyer decides how to buy it, through which seats and deals, at what bids and pace. Keeping the decisions separate, even inside one job, makes it clear which reasoning to check when something goes wrong.
Does a programmatic planner need to know how to use a DSP?
The planner needs to know what a buying platform can and cannot enforce, such as list limits, location options and frequency settings, so the plan does not ask for something the build cannot do. Day to day operation of the platform belongs to the buyer, though planners who have built a campaign write better handoffs.
How does a planner work with the client's brand team?
The brand team sets the standards: products, audience, tolerance in sensitive categories and markets. The planner turns those standards into criteria, shows the client the sites that pass and the sites left out with reasons, and asks for a decision when a site sits on the line. The client approves the version, not each site.
What should a planner do when a client asks for a site that fails the criteria?
Show the evidence, name the specific criterion it fails and offer a choice: leave it out, or keep it as a client exception recorded with the client's name and date. An exception kept in writing protects both sides. A failing site added silently weakens the record for the rest of the list.
Sources
- Certification: Digital Media Buying and Planning Certification (DBPC) (opens in a new tab)O*NET OnLine, US Department of Labor
- ads.txt 1.1 specification (opens in a new tab)IAB Tech Lab
- sellers.json specification (opens in a new tab)IAB Tech Lab
- CrUX dimensions (opens in a new tab)Google Chrome for Developers
- CrUX on BigQuery (opens in a new tab)Google Chrome for Developers
- About the "Where ads showed" report (opens in a new tab)Google Ads Help
- Monitor impression volume in your private deals (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.