>

Media Buying Fundamentals

Media buying is the practice of acquiring attention at a price that makes commercial sense. The mechanics differ by channel; the underlying economics do not.

Pricing Models

CPM — cost per thousand impressions. You pay for delivery regardless of response. Standard for awareness and video.

CPC — cost per click. You pay only for engagement. Standard for search and most performance display.

CPA / CPL — cost per acquisition or lead. Risk sits with the platform or publisher, and pricing reflects that.

CPV — cost per view, used in video where a view has a defined threshold.

The model you buy on changes who carries risk. Lower-risk models for the buyer command higher effective prices.

Auction Versus Direct Buying

Auction buying — search, social, open programmatic — prices each impression dynamically against competing bids. Flexible, efficient, no volume guarantee.

Direct buying reserves specific inventory at an agreed price. Higher cost, guaranteed delivery, control over placement and context. Appropriate for high-impact formats and brand-sensitive environments.

Most modern plans use both: direct for premium placements where context matters, auction for efficient scale.

Reach, Frequency and Waste

Reach is how many distinct people saw the ad. Frequency is how often each saw it. Impressions are the product of the two, which is why impressions alone tell you little.

Effective frequency — enough exposure to register, not so much that it irritates — sits around three to ten exposures per period depending on category and creative. Uncapped campaigns routinely deliver forty or more to a small segment while missing most of the target.

Set frequency caps deliberately. Uncapped delivery is the most common and most invisible source of waste.

Building a Media Plan

Start from the objective and the audience, not from the channels you already use. Then allocate by the role each channel plays: some build reach, some capture demand, some retarget.

Budget allocation heuristics worth knowing: do not spread so thin that no channel reaches minimum effective frequency, reserve roughly ten to twenty percent for testing, and set flight lengths long enough to escape the learning phase on algorithmic platforms.

Reading What You Are Actually Paying For

The headline rate is rarely the cost. What sits between the rate and the outcome is the part worth reading.

Ask what share of your budget reaches the publisher. Between you and the inventory there may be an agency, a DSP, an exchange and an SSP, each taking a margin. The difference between gross and working media is frequently large and rarely volunteered.

Establish whose numbers bill. Buyer and publisher count differently; agree the system of record and the acceptable discrepancy before the flight rather than during it.

Read the make-good terms. What happens on under-delivery, on invalid traffic, on a brand-safety incident — and whether the remedy is a refund or more of the same inventory.

Check the cancellation window. A campaign you cannot pause is a fixed cost, and on a direct buy that window can be long.

None of this is adversarial. It is the difference between buying a rate and buying an outcome.

Frequency, Waste, and the Number Nobody Sets

Reach and frequency are planned and then not managed, and the gap between the plan and the delivery is where most media waste lives.

Frequency caps only work within a platform. Three campaigns each capped at three impressions can deliver nine to the same person, and none of the three reports a problem. If you buy one audience across several platforms, the effective cap is the sum, and the only controls are consolidating buying or setting each cap materially lower.

Distribution matters more than average frequency. An average of four can mean everyone saw it four times, or that most saw it once and a few saw it twenty. Those are different campaigns with the same number. Ask for the frequency distribution, not the mean.

Diminishing returns arrive earlier than plans assume, and past a point additional exposure produces irritation rather than recall — a cost that does not appear in any report.

The practical discipline: decide the frequency you intend before buying, measure the distribution you got, and treat a long tail as budget to reallocate rather than as reach.

Sources

What each claim on this page rests on. Entries are typed so you can see which are primary.

  1. officialIAB Tech Lab specifications and the seller authorisation files — the supply-path and seller-verification mechanics referenced here iabtechlab.com

Ask an AI about this page

Opens your assistant with this page as the source, and a question rather than a summary. It will ask what you are building before it answers.

ChatGPTClaudeGeminiPerplexityGrok

Nothing is sent from here. The link carries only this page’s title and address.