Google will stop billing for ads that were never rendered in Display & Video 360 and Campaign Manager 360 in February 2027, according to PPC Land. For anyone buying through DV360, that means the gap between impressions you were charged for and impressions that actually appeared on a screen is set to close. It also means your billed counts, spend lines and discrepancy checks will shift on a date you can plan for.
Key takeaways
- PPC Land reports that Google will drop unrendered ads from DV360 and CM360 billing in February 2027.
- Freeze your current baselines now: billed impressions, spend, effective CPM and discrepancy rates by line item.
- Confirm the exact definition of a rendered ad, and which inventory it covers, in Google's own documentation before you change any plan.
- Audit discrepancy tolerances, pacing rules and finance reconciliation first, because they break first when billed counts move.
- Don't cut budgets or rewrite bidding on the assumption that costs will fall. Test, then measure over several weeks.
- Most of the work is internal housekeeping. A partner earns their fee on multi-account reconciliation and measurement design.
What is changing in DV360 and CM360 billing?
The headline from PPC Land is short: unrendered ads come out of billing for both DV360 and CM360, and the change lands in February 2027. That is the part we can attribute to the report. Anything finer, such as the technical definition of rendered, the media types covered, whether it applies to every buying path, or how the change appears in reports, should be read from Google's official documentation and release notes for each product.
Why the pairing matters is easy to see. DV360 is where you buy and pay. CM360 is commonly where you serve and count. Many teams reconcile one against the other every month, and a billing rule that changes on one side of that pair, or on both, changes the arithmetic of the reconciliation even if nothing about your campaigns changes.
An unrendered ad, in plain terms, is one where the buying and serving machinery did its work but the creative never appeared. The page was closed, the slot never loaded, the ad was fetched and thrown away. Under the old arrangement, some of those could still show up as billable. The reported change removes them from the bill.
It is a billing-hygiene change, not a targeting or creative change. Treat it that way. Nothing in a line item's settings needs to be rebuilt because of it.
Who should act, and who can ignore it?
Everyone who spends through DV360 is affected in some way, but the degree varies a lot.
| Team type | Likely exposure | What to do |
|---|---|---|
| In-house team with one seat and monthly invoices | Low to moderate. Invoices and delivery reports will look different after the change. | Record baselines, brief finance, add a note to reporting. |
| Agency running many advertisers on shared seats | Moderate to high. Many clients, many tolerances, many reconciliation sheets. | Audit every client's discrepancy threshold and pacing logic. |
| Team with heavy reliance on third-party verification | Moderate. Verification vendor counts and billed counts may converge or diverge. | Re-baseline the comparison and re-agree tolerances. |
| Team with fixed-budget commitments to publishers or clients | High. Fixed budgets and changed billing can change pacing. | Check how delivery is paced against spend caps. |
| Team on small test budgets | Low. Effects may be hard to see in the noise. | Note the date; no other action needed. |
Who can ignore it? Creative, audience and strategy owners don't have to change what they're doing. Their work isn't touched by a billing rule. They do need a heads-up, though, because a step change in reported numbers in February 2027 will otherwise get read as a creative or audience result.
The people who must act are the ones who own spend reporting, invoice approval, pacing and the monthly reconciliation. In a small team that's one person. In a larger one it's usually three, and they rarely sit in the same meeting.
How do unrendered ads end up in billing at all?
A programmatic impression has stages. A request goes out, a bid is placed, an auction is won, a creative is served, and then the browser or app is supposed to draw it. The count of any one stage differs from the count of the next. Network failures, slow pages, users navigating away and ad slots that never come into existence all create loss between stages.
That loss is the source of the familiar discrepancy you see between a buying platform, an ad server and a publisher. Ten percent or so is a figure practitioners often quote as normal tolerance, but that is a rule of thumb, not a standard, and your own number should come from your own history. The point is that discrepancies of this kind are expected, and many teams already carry a written tolerance for them.
What the reported change does is move one stage of that chain, the rendering stage, from something that was partly absorbed in the discrepancy to something that is excluded from the bill. How much money that represents for you depends on your inventory mix, formats and devices. Some of it will be small. Some of it may not be. We can't tell you which from here, and neither can a headline. Your own logs can.
There's a separate point worth keeping in mind. Rendering is not the same as being seen. An ad can be rendered and sit below the fold for its entire life. Viewability is a different measure, with its own definitions and its own vendors, and nothing in the report suggests this change replaces it. Keep your viewability reporting separate and don't merge the two in a dashboard.
What should you do before February 2027?
Start with the work that is cheap and reversible. Do these in order.
- Read Google's documentation. Find the DV360 and CM360 release notes and billing help pages covering the change. Write down the definition of rendered, the products and media types in scope, and the effective date as Google states it.
- Capture baselines. For the last three full months, export billed impressions, spend, effective CPM and your usual discrepancy rate by line item or insertion order. Store the export somewhere that will not be overwritten.
- Inventory every threshold. List each place a number is compared to a limit: discrepancy tolerances, pacing alerts, spend caps, automated rules, finance variance flags. Mark which ones assume the current billing behaviour.
- Brief finance and account owners. Tell the people who approve invoices that billed impressions may fall relative to delivered ones, and that a lower number is the expected outcome, not an error.
- Annotate your reporting. Add the change date as a visible marker on trend charts so the step is not mistaken for a result.
- Decide who watches the first month. Name one person for the first invoice after the change and one for the first full month of delivery data.
If you are comparing partners to help with any of this, the briefing is easier when the scope is written down. The DV360 RFP template has a structure you can adapt for reconciliation and reporting work.
How do you test and roll out the change safely?
You can't test Google's billing in advance, because you don't control it. What you can test is whether your own processes survive a different set of numbers. Run a dry exercise before the date.
Take last month's export and reduce billed impressions by a plausible range, say a few different percentages, in a copy of your reporting workbook. Watch what happens. Do the discrepancy checks fire? Does the pacing sheet suggest you are under-delivering? Does the cost-per-result chart jump? Any of those that break under a synthetic change will break under a real one, and it is better to find that on a quiet afternoon.
Then agree a rollout checklist with named owners.
| Task | Owner | Timing | Done when |
|---|---|---|---|
| Confirm definition and scope from Google documentation | Platform lead | Now | Written summary shared |
| Export and archive three months of baselines | Analyst | Now | File stored and labelled |
| Review discrepancy and pacing thresholds | Analyst and trader | Before the change | Each threshold marked keep, change or retire |
| Brief finance on expected invoice movement | Account lead | Before the change | Finance has confirmed in writing |
| Add change marker to dashboards | Analyst | Before the change | Marker visible on all trend views |
| Review first post-change invoice | Finance and platform lead | After the change | Variance explained and recorded |
| Review first full month of delivery | Platform lead | After the change | Decision logged on any setting changes |
One firm recommendation: don't change bidding strategies, floors or budgets in the same week the billing rule changes. If two things move at once, you won't be able to say which caused what.
How do you measure what actually changed?
Use a before-and-after comparison with guardrails. The aim is to separate the billing effect from the performance effect, and they are different things.
For the billing effect, compare billed impressions to delivered impressions for the same line items across matched periods. Where a campaign ran continuously across the change date with stable settings, that comparison is clean. Where it did not, note the difference and don't pretend otherwise.
For the cost effect, look at effective CPM and spend per delivered impression, and say which denominator you used. A drop in effective CPM may reflect nothing more than a different billing basis. It does not by itself mean your buying got better.
For performance, keep the outcome metrics you already trust, such as conversions, cost per acquisition or view-through measures, and compare them over a window long enough to smooth weekly swings. Four to six weeks is a sensible minimum for most accounts; short flights may need a different approach.
Two habits help. Keep a change log with dated entries for anything that touched the account, including the billing change. And report the billing effect and the performance effect on separate lines so nobody has to untangle them later.
What shouldn't you do?
A few tempting moves are worth avoiding.
- Don't bank savings. Reduced billing on unrendered ads is not a guaranteed drop in total cost. Budgets, pacing and bids interact, and you may simply buy more.
- Don't rewrite bidding in advance. You don't yet know what your numbers will look like. Wait for data.
- Don't merge rendered and viewable. They are different measures. Reporting them as one will confuse everyone, including you.
- Don't tighten discrepancy tolerances because billing is cleaner. Publisher and ad server gaps still exist for other reasons.
- Don't treat this as the headline DV360 story of the quarter. The Tech Buzz has also run a separate piece on vertical video unification in Display & Video 360. That is a creative and formats question with its own owners and its own checks, and it should not be folded into billing work.
- Don't assume CM360 and DV360 will line up identically from day one. Allow for a settling period and some manual comparison.
Where does an implementation partner help, and where can you go it alone?
Most of the preparation is internal. Reading release notes, exporting baselines, listing thresholds and briefing finance are all tasks a capable in-house analyst can finish in a few days. A partner adds little there, and paying for it is hard to justify.
The case for outside help gets stronger in three situations. The first is scale: an agency or group running many seats and advertisers, where every client has its own tolerance and its own finance process, and nobody has time to audit them all. The second is measurement design, where billed counts, ad server counts and verification vendor counts have to be rebuilt into one view that stakeholders trust. The third is a dispute, where a publisher, client or finance team questions the new numbers and someone with cross-account experience needs to explain them.
If you do go looking, ask each candidate for a specific plan: what they'd export, which thresholds they'd review and what the first post-change report would show. Vague answers about optimisation are a poor sign. You can browse DV360 specialists and use partner search to narrow by region, size and sector. Whoever you pick, keep ownership of the baselines yourself.
The short version: do the reading and the exports this month, hold your strategy steady through the change, and judge the result on matched data rather than on the first invoice.
Sources
- PPC Land — Reports that Google drops unrendered ads from DV360 and CM360 billing in February 2027.
- The Tech Buzz — Separate coverage of vertical video unification in Display & Video 360, listed for breadth.
Frequently Asked Questions
What does it mean that DV360 will not bill for unrendered ads?
An unrendered ad is one that was served or selected but never actually drawn on the page or screen. PPC Land reports that Google will drop these from DV360 and CM360 billing in February 2027. The precise definition and which inventory types it covers should be confirmed in Google's documentation.
Will my DV360 costs go down in February 2027?
Possibly for some campaigns, but the direction and size depend on how much of your delivery was never rendered, and that varies by inventory and format. Don't forecast savings until you have compared your own billed and rendered counts. Treat any reduction as a reconciliation change first and a performance change second.
Do I need to change my bidding or budgets for the DV360 billing change?
Not as a first move. Keep strategies as they are, record your baselines, and watch pacing and effective CPM for a few weeks after the change takes effect. Adjust only where the new numbers show a real problem, such as under-delivery against a fixed budget.
How do I reconcile DV360 and CM360 numbers after the change?
Pull the same date range from both platforms and from any third-party verification you use, then compare impressions, spend and viewability side by side before and after the switch. Agree a written tolerance with finance and publishers in advance. Log the change date in your reporting so a step change isn't mistaken for a performance shift.
