Why Ad Costs Keep Rising While You Still Can't Tell What's Working
Updated: 4 days ago
Updated November 2026.
Two things are true at the same time, and they are connected. The auction you buy in gets more expensive every quarter. And the reporting you get back is built around the media buy, so the lever with the most influence on your results is the one you cannot see.
The short answer: costs rise because more money keeps chasing the same attention. You cannot tell what is working because your data is organized around campaigns and channels, and creative sits outside that structure. The second problem makes the first one worse, because when costs climb the obvious response is to tighten targeting and bidding, and the platforms now run most of that themselves.

The cost increase is real, and the platforms publish it
You do not have to rely on third-party CPC averages for this. Meta reports its own ad pricing every quarter, in its quarterly results.
In Q1 2026, Meta reported that average price per ad increased by 12% year over year, and that ad impressions delivered across its Family of Apps increased by 19%.
In Q2 2026, average price per ad increased by 12% year over year again, with impressions delivered up 14%.
Read those two lines together. The price per unit rose 12% in back-to-back quarters, and Meta sold more units at the same time. Prices and volume rising together is what a market looks like when demand grows faster than supply. Your plan is competing with more money for the same attention, and the clearing price goes up.
Those are Meta's own figures. Treat the CPC and CPM averages that circulate in industry roundups with more caution, because most trace back to one vendor's sample rather than to a platform's reporting.
Four forces behind it
More advertisers bidding through automation
Automated bidding raised the floor. When most buyers in an auction are running machine-driven bids against similar objectives, the easy arbitrage disappears. Everyone finds the same pockets at the same time, and premium inventory prices up.
More content competing for the same attention
More formats, more placements, more advertisers producing more assets. Supply of advertising has grown faster than the supply of human attention, so the cost of simply being seen rises even where the cost of production has fallen.
Less visibility into where the money goes
Automated campaign types optimize across placements and audiences on your behalf. They work, and they also reduce what you can see. When costs spike inside a black box, you have fewer places to look and less to act on.
Creative inefficiency, which is the one you control
Teams churn variants without a test-and-learn structure, keep spending behind assets that have stopped working, and rebuild creative that already ran somewhere else because nobody could see it. Each of those raises cost per result without touching the auction.
Why your reporting cannot tell you what is working
Three structural reasons, and none of them is a discipline problem on your team.
The data is organized around the buy
Ad platforms were built to sell and optimize media. Their data model starts at the account, then the campaign, then the ad set, then the placement. Creative hangs off the end of that as an attachment to a line item. So you can get spend by campaign in four clicks, and you cannot get one video's performance across six campaigns and three channels without building something yourself.
We wrote about that underlying problem in The Data Substrate Matters.
Every platform keeps its own definitions
A view on one platform is three seconds. On another it is two. Attribution windows differ, conversion definitions differ, and each platform reports on its own schedule. Averaging a CPM across TikTok, Meta, Google, LinkedIn and The Trade Desk without normalizing first produces a number that looks precise and means very little. The usual fix is a spreadsheet somebody rebuilds every Monday, which is stale by Tuesday.
The numbers arrive after the decision
Platform reports update through the day. The downstream numbers your finance team trusts land a week later. By the time the two agree, the flight is over and the budget is spent. You learn that a campaign underperformed, and the lesson cost you two weeks of media.
Why these are one problem rather than two
Creative is the biggest lever you control. NCSolutions, analyzing nearly 450 CPG campaigns across digital and TV, reported in 2023 that creative drives 49% of the incremental sales from advertising. Brand came second at 21%, then reach at 14%, targeting at 11%, and recency at 5%.
Hold that next to what happens when costs rise. The instinct is to work the dials: tighten the targeting, adjust the bids, narrow the audiences. Those dials increasingly belong to the platform and run on its models. Meanwhile the one input that still belongs to you, and that carries the most weight, goes unmeasured at the level where you could do something about it.
So rising costs push teams to optimize the levers they have least control over, while the largest lever stays invisible. That is why the two halves of this question belong together.
What to do about it
Size the waste before you hunt for it. Work out what share of last month's spend sat behind creative that missed the goal you set for it. A percentage gives you something to report and something to beat.
Look at the budget split, not only at individual ads. The expensive mistake is rarely one bad ad. It is most of the money sitting behind the weakest stage or the weakest concept, where no single line item looks guilty.
Put every channel on the same definitions. One asset, every channel it ran in, one view, with the metrics normalized rather than averaged.
Measure the elements inside the ad. Knowing that ad 47 beat ad 12 tells you what to pause this week. Knowing that the version shot in a real place beat the version shot in an office tells you what to brief next.
Watch it while the campaign is live. A finding that arrives after the budget is spent is a post mortem. The same finding on day three is a decision.
Write the next brief from what you found. Most teams skip this step, and it is the only one that compounds.
The sequence matters more than the tooling. For the full method, start with Creative Performance Measurement: The Complete Guide, and for the workflow version, How to find wasted ad spend walks through it step by step.
Where mktg.ai fits
mktg.ai reads the creative you have live and connects performance to the elements inside each asset, so the next brief gets written from evidence rather than instinct. It covers TikTok, Meta, Google, LinkedIn and The Trade Desk, and setup is read-only access to your accounts, with no tagging work for your team. If you are comparing options in this category, The Best Creative Analytics Tools in 2026, Compared sorts the main tools by the job each one does, ours included.
Frequently asked questions
Are ad costs rising, or is it just my account?
Rising, and the platforms report it themselves. Meta reported average price per ad up 12% year over year in both Q1 and Q2 of 2026. Impressions delivered rose in both quarters as well, so this is not a shrinking market. If your own costs are climbing faster than that, the gap is worth investigating in your creative mix before you assume it is the auction.
Why can't my ad platform tell me which creative is working?
It can tell you which ad IDs performed, inside its own walls, on its own definitions. What it cannot do is show you one asset's results across every channel you buy, or tell you which element inside the ad did the work. Its data model is built around the media buy, and it has no view of what you ran anywhere else.
Is this a targeting problem or a creative problem?
On the available evidence, mostly creative. NCSolutions put creative at 49% of incremental sales and targeting at 11%. Targeting still matters, and the platforms now automate much of it. Creative is where the remaining judgment lives.
Which metrics tell me whether creative is working?
Start from the goal the ad was built to hit, then hold every asset to that. Click-through rate tells you whether an ad earned attention, and it says nothing about whether it earned revenue. Spend sitting behind weak assets is usually the most useful number to start with, because it converts the problem into money. What Is Creative Intelligence? A Plain-English Guide covers how element-level measurement works.
How quickly can I find wasted spend?
Visibility comes first and recovery follows. The first pass is a waste rate: underperforming spend divided by total spend for a recent period. That can be done in an afternoon once the data is in one place. Moving the money is the part that takes a conversation.
Does this replace my ad platform reporting?
No. Keep buying where you buy. This sits alongside Meta Ads Manager, Google Ads and the rest, reading the same spend and results they report, and adds the layer they do not have: the asset across every channel, and what is inside it.
Is the 49% figure from an independent study?
It comes from NCSolutions, a measurement company, analyzing nearly 450 CPG campaigns across digital and TV, published in 2023. The shares are modeled contributions rather than directly observed values, and the document does not state the period the campaigns ran. It is the most widely cited figure on creative's contribution, and it is worth citing with those caveats attached.

The bottom line
Ad costs are going up and will keep going up, and no reporting change reverses that. What is in your control is whether the money lands on creative that works. Most teams cannot answer that while the campaign is running, which is the only time the answer is useful. The line about half of advertising spend being wasted has been around for a century, and we looked into who really said it. The part worth keeping is that nobody knew which half. That part is now solvable.
If you want to see what this looks like on real creative, book a demo.




Comments