What belongs in a connected TV advertising ninety day plan?

What to take away

Why ninety days suits a learning plan

Ninety days is long enough to collect evidence and short enough to stay accountable. It also sits neatly inside a quarterly budget cycle, which keeps finance reporting simple.

A four-week test usually produces too little data to separate a good idea from a lucky one. A six-month test hides weak results behind a long average. Ninety days sits between those two failures.

The aim is to answer one question: which mix of creative, audience and placement earns its place? The Connected TV advertising strategy and planning guide sets out how objectives shape budget and format choices before you brief a partner. Skip that step and three months of numbers will not tell you much.

Weeks 1 to 4: fix the objective and the numbers

Write the objective as a business result rather than a media metric. "Grow trial sign-ups by 8% among 25 to 44s in Greater Manchester" is workable. "Raise awareness" is not, unless you can define what awareness means in your category.

Name the person who can approve a change of direction. If that person is absent when results arrive, decisions slip by a week.

Then set the money and rules: total spend, the cost per completed view you can tolerate, how often results reach the people who can act on them, and what triggers a pause. Do not treat a general TV payback figure as a CTV benchmark. For team-level role clarity, the CIPD's resources on organisational strategy offer general context, not CTV buying guidance.

Name the people who will execute the plan: the advertiser's accountable owner, the in-house or agency media planner/buyer, the broadcaster or platform contact, and the measurement lead. If buying programmatically, name the trading-desk contact too. Record who books inventory, traffics creative, checks pacing, reconciles reports and can pause or reallocate spend.

List the exact service and format before setting the budget. UK options may include ITVX and Channel 4 streaming (BVOD), Sky AdSmart (addressable TV), ad-supported Netflix, Prime Video and Disney+, or Rakuten TV (AVOD or FAST inventory). Confirm that advertising and suitable inventory are available for the intended audience and buying route before including a service in the plan.

Record the buying route for each service: a broadcaster sales house or direct deal, or programmatic access through a DSP such as Display & Video 360 (DV360) or The Trade Desk. Availability differs by service, so confirm access and inventory with the seller before launch.

Agree what measurement covers. Ask whether BARB reporting or CFlight is available for the chosen inventory and what it includes; use supplier delivery reports alongside it. Define completion, reach, frequency and viewability consistently before comparing results. There is no single CTV CPM, completion-rate or viewability benchmark that applies to every service, so obtain service-specific costs and set campaign thresholds before spending.

Add a UK compliance check to the brief. Confirm whether the CAP Code or BCAP Code applies to each placement and who will approve the creative; both are administered by the ASA. If audience data or device identifiers are used, agree responsibilities under UK GDPR and check whether PECR applies, using ICO guidance. Ofcom regulates qualifying TV and on-demand services, but is not the data-consent authority.

Weeks 5 to 8: test creative and placements

Test one variable at a time. A 15-second cut against a 30-second cut tells you something; changing the audience, the length and the publisher together tells you nothing. Keep a log of every change, the date and the reason. Log the hypothesis too, so a later reader knows why a variant ran.

For each supplier, record the frequency cap it can enforce, the period it covers and whether it applies at household, device or another level. Set a campaign-specific numeric cap before launch rather than assuming one cap will carry across platforms.

Write down the sample size you need before you read any result. Reading data too early is a common way teams talk themselves into a bad decision.

Creative practice carries across paid formats, so the IAB UK principles for native advertising creative best practice are a sensible baseline for keeping paid creative consistent while variants run. By week eight you should know which two or three combinations deserve the rest of the budget.

Weeks 9 to 12: scale what survived

Scaling means more money to the winning combination and less to everything else. Move budget in steps of 10% or so, and check delivery after each move. A sudden jump can push costs up and make a good result look worse.

Set the day-90 meeting before the quarter starts, with the owner, a finance contact and whoever signs off budget in the room.

Decide now what "winning" means: a cost per acquisition below your threshold, or a rise in branded search. Then set the next window using the Connected TV advertising strategy framework, which shows how one quarter's evidence becomes the next quarter's choices.

Keep the measurement approach you set in week one. Changing metrics halfway through makes the comparison useless, and the comparison is what justifies the next budget.

Day-90 decision

Use the thresholds agreed in week one to make the written choice:

Match the situation to the decision

Common questions

What should the first 30 days deliver?

A written objective, an agreed budget, a measurement plan and a baseline to compare against. Spending can start in week three, but the paperwork comes first.

How much budget should testing take?

Ring-fence a share you are willing to lose and label it a test cost. For example, a team working with £20,000 a quarter might treat £5,000 as testing and protect the rest for scaling.

What if results are flat by day 60?

Check the basics first: was the audience large enough, was the creative built for the format, and was measurement working? If all three are sound, cut spend and write down what you learned.

Who should own the ninety day plan?

One named person, with a deputy for cover. The owner keeps the test log, calls a pause when costs drift and writes the day-90 decision.