Building your first DV360 campaign without learning the expensive way

DV360 is not hard to click through. It is easy to structure badly, and you find out weeks later when someone asks a question the report cannot answer. Getting the hierarchy and the caps right at the start is most of the work.

This assumes you have access. If not, start with how to get DV360 access.

The three levels, and what each one is for

Level	What it holds	What you set here
Campaign	A business goal, grouping related insertion orders	The shared frequency cap across everything beneath it
Insertion order	A set of line items for the same effort	Budget, pacing, default targeting, frequency
Line item	The actual buy: targeting, creative, bid	Everything specific — route, device, audience, creative

Per Google's documentation, campaigns group related insertion orders with a common business goal, an insertion order contains line items related to the same campaign and can carry frequency caps, pacing and default targeting, and line items are where the buying actually happens. Line items in an insertion order share its budget and serve within its dates.

How budgets and pacing actually behave

Two behaviours worth understanding before you set numbers, both from Google's budget and pacing documentation and about budgets:

The insertion order budget caps the combined spend of its line items. If the IO daily budget is lower than the sum of its line items' daily budgets, DV360 allocates the IO budget proportionally to each line item's daily target. So line item budgets are targets, not guarantees — the IO is the real constraint.

Daily spend targets are recalculated as budget remaining divided by days remaining, and adjusted for historical ability to spend and day-to-day supply fluctuation. Practical consequence: adding budget mid-flight does not spread evenly backwards, it raises the remaining daily target. Cutting budget mid-flight does the reverse, sharply.

Frequency caps cascade, and they do not add

This is the single most misunderstood setting. Per Google's frequency cap documentation, if caps are set at more than one level, serving stops once the first cap at any level is reached — campaign, insertion order or line item. An insertion order's cap applies across the line items within it.

What that means in practice:

• Put the cross-channel cap at campaign level. This is what stops the same household seeing your CTV, YouTube and display the same evening.
• Use insertion order caps to protect a specific channel group.
• Leave most line items uncapped unless you have a specific reason, because a tight line-item cap will silently throttle delivery long before the campaign cap binds.

Setting a cap at every level "to be safe" is how campaigns under-deliver and nobody can work out why.

A structure that stays readable

Split line items by what you might want to turn off independently:

• Buying route — curated deals versus open auction
• Device or screen — TV screens versus desktop and mobile
• Publisher group, where supply quality differs meaningfully
• Creative length, when you are running 15s and 30s

One giant line item optimises perfectly well and tells you nothing. Forty line items split by every dimension fragments learning and starves each one of data. The test is simple: could I act on this split? If no, do not make it.

Channel-specific structure notes are in YouTube CTV advertising and how to buy CTV ads.

The day-1 checklist

Before anything goes live:

• Conversion tracking implemented and verified with real traffic — not just a debugger.
• Creative approved in every required format, sized and specced per channel.
• Audience written down, in one sentence two people would agree on.
• Frequency cap set at campaign level, across channels.
• Budget and flight dates set at insertion order level, with a flight long enough to optimise — six to eight weeks, not two.
• At least one curated deal activated, so you are not launching on open auction alone.
• Brand safety and inventory exclusions configured, per brand safety in programmatic.
• Success metric agreed in writing with whoever will judge the campaign.
• Someone named as responsible for week-one checks — delivery, frequency distribution and publisher-level placement.

The last one matters more than it looks. Almost every recoverable problem is visible in week one and expensive by week four.

Week one: what to actually look at

Not performance. Delivery mechanics:

1. Is it pacing? If spend is far under target, something is over-restricted — usually a line-item cap or a narrow audience that does not resolve.
2. What does the frequency distribution look like? An average of three can hide a tail of households seeing twenty.
3. Where did it run? Publisher and app level. This is your quality and brand safety evidence at once.
4. Is the conversion signal arriving? If not, fix it now, not at the wrap.

Only after those look sane does optimisation mean anything. Measurement design is covered in measurement and attribution.

Where SeatNMD fits

SeatNMD gives agencies and advertisers access to DV360 without a direct Google contract and without the spend minimums Google asks of an agency, with access provisioned in days rather than months, across YouTube, CTV, DOOH, audio, display and video. You choose how it runs: Self, where your team operates; Support, the model most clients pick, where you operate with senior backup on exactly the structure decisions above; or Managed, where we run it end to end — and you can move between them without re-contracting. Billing is monthly on our credit lines, and audiences, history and measurement stay yours.

FAQ

What is the difference between a campaign, an insertion order and a line item in DV360?

A campaign groups insertion orders under a business goal and holds the top-level frequency cap. An insertion order holds a budget, pacing and default targeting for a set of line items. A line item is the actual buy: targeting, creative and bidding.

Where should I set the frequency cap in DV360?

At campaign level for the cross-channel cap, since serving stops at the first cap reached at any level. Use insertion order caps for a specific channel group, and avoid tight line-item caps unless you have a specific reason — they throttle delivery before the campaign cap binds.

Why is my DV360 campaign under-delivering?

Most often over-restriction: a tight line-item frequency cap, an audience that does not resolve on the targeted inventory, or an insertion order budget lower than the sum of its line item budgets. Check pacing and frequency distribution in week one.

How long should a first DV360 campaign run?

Long enough to exit learning and optimise — six to eight weeks is a realistic minimum for a genuine test. Two weeks is a spend, not a test. The budget arithmetic is in DV360 minimum spend.

What usually delays a DV360 launch?

Not the platform. Conversion tags not implemented or firing incorrectly, creative missing in required formats, no written audience definition, and deals not yet activated. All four can be worked in parallel with getting access.

---

Want senior backup on your first campaign structure? Tell us what you run today through the contact form.