Why this matters
DV360 campaigns rarely fail on one wrong setting. They fail because reasonable settings were combined on the wrong line items: remarketing sharing a line item with prospecting, a deal starved by an open-auction sibling, a cap the budget makes impossible to respect. The DV360 audiences guide explains where each list comes from and what it costs, the bidding and pacing guide explains how strategies, pacing and caps behave once live, and the DV360 creatives guide explains why the creative set decides which auctions a line item can enter. Here you put the three together on one plan.
The brand and its campaign
Voltwheel is an invented maker of electric scooters, selling online and through small showrooms. It is launching a commuter model and wants people who travel to work in its showroom cities to know it exists, then book a test ride. Its Campaign Manager 360 advertiser is linked to DV360, and Floodlight tags have been firing for some weeks.
Voltwheel: budget, audiences and conversion data
Flight: six weeks. Media budget: £54,000. Cities: Manchester, Leeds and Bristol.
| Audience | Source | Matchable size across the three cities | Cost |
|---|---|---|---|
| Visited the new model's product page, no test ride booked (30-day membership) | Floodlight list, created three weeks ago | 38,000 | Media only |
| Booked a test ride (60-day membership) | Floodlight list | 2,100 | Media only |
| Owners of earlier Voltwheel models | Customer Match upload | 9,500 matched | Media only |
| In-market for scooters and e-bikes | Google audience | Large | Media only |
| Custom audience from commuting and bike-to-work keywords | Google audience | Large | Media only |
| "Urban commuters" | Third-party segment | Large | £1.50 CPM data fee |
Floodlight test-ride bookings from all media: about 20 a week, almost all from paid search.
Voltwheel: inventory and creatives available
- A private marketplace deal with a regional news publisher, fixed price £8 CPM, display and native.
- Open auction clearing prices in recent tests: about £4 CPM for display, £12 CPM for video.
- HTML5 display in two sizes, medium rectangle and leaderboard; three more sizes can be ready within a week.
- Video cuts of 15 seconds and 6 seconds, and a native asset set.
- A "book a ride near you" test-ride creative, medium rectangle only.
Building the line items
Start from the jobs, not the audience list. Prospecting reaches commuters who have never heard of Voltwheel; remarketing persuades people who looked at the model to book; the deal buys a known environment at a known price; a priced segment has to prove it is worth its fee. Each job wants its own bid strategy and cap, so each gets its own line item.
Then test each line item against two constraints. A line item bids only where it has an approved creative in a size the request accepts. And budget divided by the expected CPMCost per 1,000 impressions. gives impressions, and impressions divided by the matchable audience gives the average exposures each identifier would receive. If that figure is far above the cap you want, the budget will either go unspent or buy FrequencyAverage times a person saw an ad. you did not intend. The CPM calculator does the first step in seconds.
Task
Answer before reading on.
- List the line items, each with its job, inclusions, exclusions and format.
- Assign creatives and say what must be ready before launch.
- Choose a bid strategy for each and say when, if ever, it should change.
- Set pacing, and say how you protect the deal.
- Split the budget, calculate impressions per line item, and check remarketing against its list size.
- Set frequency caps at campaign and line item level.
Worked answer
Voltwheel: line item plan and arithmetic
| Line item | Audience | Format | Budget | Expected CPM | Impressions |
|---|---|---|---|---|---|
| Prospecting video | In-market plus commuting custom audience; exclude test-ride bookers and owners | 15s and 6s video | £23,000 | £12.00 | about 1,917,000 |
| Prospecting display | Same as video | HTML5, five sizes | £15,000 | £4.00 | 3,750,000 |
| News deal | No audience layer; exclude owners | Display and native | £9,000 | £8.00 | 1,125,000 |
| Commuter segment test | Third-party segment; same exclusions | HTML5, five sizes | £3,000 | £4.00 + £1.50 data | about 545,000 |
| Remarketing | Product-page visitors; exclude test-ride bookers | Test-ride creative, five sizes | £4,000 | £5.00 | 800,000 |
Remarketing check at £4,000: 800,000 ÷ 38,000 ≈ 21 impressions per identifier over six weeks, about 3.5 a week. A first draft with £6,000 bought 1,200,000 impressions, about 32 per identifier, or more than 5 a week, which a cap of 4 a week would never let it spend. The £2,000 difference moves to prospecting video.
Caps: campaign 10 per week; prospecting line items 3 per week each; remarketing 4 per week; deal 3 per week.
Line items. Prospecting is split by format because video and display clear at different prices, and a shared line item would let the cheaper format take the budget. Owners and test-ride bookers are excluded from prospecting, and bookers from remarketing, because they have already done what the campaign asks.
Creatives. No line item should launch with only the existing sizes; upload the extra sizes, including for the test-ride creative, days ahead so Google's and the exchanges' reviews finish. Use even rotation on prospecting video at first so the cuts are compared fairly.
Bidding. DV360 has no conversion history for these line items, and the account's booking volume is too thin for an automated strategy to learn from. Every line item starts on a fixed bid: the deal at its price, open auction near observed clearing prices, the segment at the display bid so the data fee is the only difference. Once remarketing records bookings steadily, move it alone to maximise conversions with a CPM ceiling.
Pacing. Even flight pacing on the insertion orders, daily pacing on the line items. The deal sits in its own insertion order so open-auction line items cannot reach a shared allowance first.
Segment test. Judge it on media plus data against prospecting display, before scaling it.
Common mistakes
- Remarketing and prospecting on one line item. The cap and the bid suit neither.
- Launching with two creative sizes. The line item bids on a fraction of the requests it could enter.
- Automated bidding with no conversions. The strategy has nothing to learn from.
- A budget the list cannot absorb. Remarketing either under-spends or overexposes the same people.
- Forgetting the data fee. The segment looks cheap until media and data are added together.
Next, test yourself with the DV360 planning checkpoint or return to the intermediate programmatic level.
