top of page

YouTube Ads Audience Targeting: Build a Useful Test Brief

Writer: DIGITALTHROB
DIGITALTHROB
Oct 18
5 min read

Before approving a YouTube audience test, you should be able to answer three questions: why might these people buy, what will the campaign actually target, and what result would justify another round of spending?


A list of interests answers none of those questions on its own. We start with one buying hypothesis, translate it into settings the account supports, then agree the commercial limits before launch. Here is how that works for a fictional retailer selling an adjustable desk with a cable tray.


Start with the reason someone might buy


The retailer's hypothesis is: people researching an adjustable desk for a cramped home office may value a bundle that changes height and keeps cables organised.


That is specific enough to guide a test. The video can demonstrate those features, the audience inputs can reflect desk research, and the product page can sell the same bundle.


“People who work from home” leaves too much unexplained. They may already have suitable furniture or be researching jobs rather than desks. Write down the buying situation and the reason this offer could answer it before choosing interests.


This brief concerns prospecting. If the real aim is to reach previous visitors or customers, start with when retargeting ads make sense; the audience and the business question are different.


Confirm the campaign and where it can serve


For this example, the retailer chooses Demand Gen with purchases as the conversion goal. The campaign type matters: Google's Video Action Campaign migration guidance explains the replacement of that older conversion-focused campaign type with Demand Gen. An old setup checklist may describe controls the current campaign does not offer.


The retailer selects YouTube in-stream using manual channel controls. Google documents these channel choices at ad-group level, while recommending all Google channels for most advertisers. Here, the team has one approved landscape demonstration and deliberately wants a first read on that surface. Restricting delivery may reduce opportunities elsewhere; it is a choice for this brief, not our default recommendation.


Before approval, the account manager records the saved campaign type, conversion goal and selected channels. If the proposed channel and audience combination is unavailable, resolve that before spending.


Turn the hypothesis into audience inputs


The proposed custom segment uses the search-term option, with these fictional inputs:


  • height adjustable desk

  • standing desk with cable management

  • adjustable desk for home office


Google's custom-segment documentation explains that, on campaigns running on Google properties, this option reaches people who searched for those and similar terms. Its default interest or purchase-intention option interprets inputs differently. The brief should state which interpretation the team selected.


These are audience inputs, not Search campaign keywords. They do not guarantee that each viewer searched an exact phrase, is currently shopping or fits the retailer's imagined customer. Nor do these example phrases carry any claim about search volume or likely performance.


Keep unrelated furniture interests out of this first brief. If URLs are proposed later, Google's documented purpose is to reach people browsing similar websites. Entering a competitor's URL does not buy its visitor list or place the ad on that site.


State how far delivery can extend


Selected audience inputs and the campaign's delivery boundary need separate entries in the brief.


For this Demand Gen trial, the team turns optimized targeting off in the ad group. The purpose is to begin with the selected audience configuration before considering broader automated discovery. Google explains that optimized targeting can go beyond selected signals, using information including creative and landing-page content. When enabled in Demand Gen, it can also extend beyond selected demographic signals.


Disabling it does not verify a viewer's identity or buying intent. It makes the configuration clearer; Google's audience estimates still have limits.


The retailer also proposes eligible first-party customer and previous-visitor segments as exclusions to reduce overlap with existing demand. Confirm that the lists are usable and the exclusions are saved before launch. Google's audience-exclusion guidance records privacy-related limitations, so “prospecting” cannot promise that every viewer is new to the business.


Attach a spending limit and a review rule


The owner now needs an approval record that connects those settings to the offer and the money. For the fictional desk trial, it contains:


  • Offer and destination: demonstrate height adjustment and the included cable tray, then send visitors to the bundle's product page. Price, dimensions, included items and delivery terms must agree. Our YouTube landing-page guide covers that handoff in detail.

  • Measurement: verify purchase tracking and order identifiers. Use order records to identify first-time customers where possible and retain cancellations and returns for reconciliation. Agree the reporting period and attribution basis before comparing results.

  • Spending authority: approve up to $1,200 in media spend, with a named account manager checking pacing and responsible for stopping the trial. This is an owner-approved ceiling, not an instruction to enter $1,200 as a daily budget.

  • Commercial limit: the owner allows up to $110 of media spend per retained first-time order, based on the business's contribution and other cost requirements. That is a fictional assumption, not an industry benchmark.

  • Review dates: review seven days after spending ends, then reconcile cancellations and returns again after 30 days. Choose real timings around the business's buying and return cycle; these dates are illustrative.


Keep the offer, video and destination stable during the run. Pause if tracking fails, stock becomes unavailable or the advertised promise cannot be met. Record interruptions so the result has an explanation the owner can inspect.


Make the next decision from reconciled orders


Suppose the trial spends $1,200. The initial reconciliation finds 12 first-time orders associated with the campaign, giving a media cost of $100 per recorded first-time order.


Two are later cancelled. Across the remaining ten orders, media cost rises to $120 each, before any other acquisition costs. Against the owner's $110 limit, the later result blocks an automatic repeat—even though the early figure appeared acceptable. If those orders remain inside their return window, the result is still provisional.


The review should end with a documented action:


  • Continue within a new spending limit if reconciled economics remain acceptable and the evidence supports another round.

  • Revise one assumption if there is a specific issue worth testing, such as a misunderstood bundle promise. Record the change so the next result remains interpretable.

  • Stop if the economics are unacceptable or the business cannot support another trial.


Low delivery needs a delivery diagnosis; it does not prove the buying situation lacks demand. An acceptable result also does not establish a winning audience. This example has no randomised control, so it cannot show that the segment beats another audience or caused every recorded order.


A useful audience brief gives you a buying hypothesis, a checkable setup and a spending decision you can defend. If you need help turning an audience idea into that brief, our YouTube ads management service connects campaign planning with the evidence needed to judge the next round.


 
 
 

Comments


Discover clics solution for the efficient marketer
More clics

Never miss an update

bottom of page