What this checkpoint covers
The questions draw on the guide to writing a brand suitability policy, the tutorial on building a programmatic inclusion list and the tutorial on responding to a brand safety incident. Every scenario is set at Pluvenhay Rail Tours, an invented operator of heritage rail tours that sells day trips and weekend breaks online and buys display through an agency.
Task
Write a short answer to each question before reading the explanations: the decision and the reason. Count an answer as right only when the reason matches. For every miss, reread the linked section; if the incident questions caught you out, redo the exercise on handling a car seat maker's brand safety incident.
Questions
- The agency proposes to adopt its verification vendor's moderate tier as Pluvenhay's policy, and to relax the safety floor on one deal because the publisher offers a better price. What do you say?
- Pluvenhay's marketing manager wants to exclude all news, because rail accidents are reported in the news, and all user-generated pages, although the brand's most loyal customers talk on railway enthusiast forums. What stance would you write instead?
- A railway magazine deal will not spend, so the trader loosens its unlabelled content setting and mentions it to the marketing manager in passing. What should have happened?
- The agency offers its ready-made "premium travel" inclusion list. How should Pluvenhay source its list, and what check does every domain pass before it earns a place?
- Read the sizing figures below. The list is due to go live next week. What do you do?
- Screenshots arrive in a single week: the ad beside a report of a level crossing collision on a trusted news site; the ad beside an opinion piece criticising steam locomotive emissions on a trusted publisher; and the ad on an unfamiliar domain stacked with auto-refreshing units. The policy excludes tragedy and accidents and allows trusted news otherwise. How do you treat each?
- After the first screenshot, the managing director asks for the keywords train, crash and derailment to be blocked on every line item. What do you do instead, and what happens to any emergency block afterwards?
Pluvenhay Rail Tours: sizing the inclusion list before launch
| Item | Figure |
|---|---|
| Domains on the list after seller checks and policy review | 40 |
| Monthly budget on the governed line items | £6,000 |
| Expected CPM on the listed inventory | £9 |
| Impressions the budget needs | 6,000 ÷ 9 × 1,000 ≈ 667,000 |
| Monthly inventory forecast for the listed domains with the line items' settings | About 400,000 impressions |
| Past monthly spend on the listed domains, from placement reports | About £2,500 |
Answers explained
- No to both. The floor is stated in a sentence, on for every buy and not open to exception, whatever the price. Vendor tiers are a shared vocabulary for the decision, not the decision: the policy says, topic by topic, where this brand stands and why (the safety floor and the tiers above it).
- Keep news, name the risk, and allow named communities. Treating news as unsafe pushes budget towards inventory that reports on nothing. A better stance allows a list of trusted news publishers, excludes tragedy and accidents specifically, because a rail tour beside a crash reads as exploiting it, and allows named, well-moderated enthusiast forums while excluding other user-generated pages. The same table tells the planner which railway subjects suit Contextual TargetingChoosing where ads appear from the content of the page, app or video being viewed rather than from data about the viewer. (topics, news and user-generated content).
- A written exception request, approved by someone else. The request names the deal, the setting, the reason, the delivery problem and an end date; the policy owner or a named deputy approves, never the person asking; it goes in the exception log and the setting is restored on the end date (exceptions, approvals and the people who decide).
- From the objective and the audience, with the borrowed list as candidates only. Start from where Pluvenhay's customers read and from past placement reports read by spend, then check every candidate's ads.txtA file on a publisher's site listing the companies authorised to sell its ad inventory. file and supply path. A domain with no file, or sold only through resellers, deserves a closer look; a failed seller check is a gate, not a low score (gathering candidates and checking sellers).
- Do not launch as it stands. The forecast and the past spend both say the list cannot absorb the budget at that price. Widen it from the next ranked candidates, each with its reason recorded, or lower the budget on the governed line items, and write the expected reach and price into the plan (estimating the reach the list leaves).
- A different treatment for each. The collision report is an incident: the policy excludes accidents, so the settings did not match it; the trader leads with the policy owner told the same hour. The emissions piece is inside the policy: reply, record, and ask the owner whether the policy should change. The stacked domain is a poor placement, handled through exclusions and the list review in the normal routine (incident or poor placement).
- Stop the route and find the cause. Those keywords would remove the railway pages Pluvenhay most wants. Apply the narrowest control the trace supports, ask why until the answer is a policy line, a list entry or a setting, and fix it there. Once the permanent fix covers the risk, lift any broader temporary block so the Brand SafetyControls that keep ads away from content that would harm the advertiser's reputation. response does not leave the buy narrower than the policy intends (finding and fixing the cause).
Where to go next
The guide to planning contextual targeting shows how the same policy table feeds a context map. Then continue to the third module and work through untangling inherited Floodlight activities for a furniture rental firm.
