August brought a sharp pullback in programmatic pricing. According to DataBeat’s August 2026 analysis, overall U.S. programmatic CPMs declined 19.9% month over month, with display CPMs down 20.3%.

On the surface, that looks significant. But the year-over-year numbers tell a different story. Overall CPMs remained 25% higher than August 2025, while video CPMs were up 40.1%. That mix of sharp month-over-month declines and strong year-over-year performance suggests publishers should be careful not to interpret August softness as evidence of a broader slowdown.

Programmatic pricing is seasonal, and August often sits in an awkward period between summer budget softness and the heavier spending that comes later in the year. But seasonality should be a starting point for analysis, not an explanation for every decline. The more useful question for publishers is whether performance is moving with the market or whether something specific is happening within their own inventory.

source: Databeat

August softness has historical precedent

Adapex’s own network data provides useful context. Looking at CPM performance from 2019 through 2025, average Q4 CPMs were higher than August CPMs in six of seven years. The median increase from August to the Q4 average was approximately 17.5%. November showed an even stronger historical pattern, exceeding August CPMs in six of seven years with a median lift of roughly 24.5%.

Adapex Network Historical CPM Index

That history doesn’t mean publishers should expect the same rebound in 2026. In fact, 2025 was the exception in the Adapex data, when pricing did not follow the same August-to-Q4 pattern. Market conditions change, advertiser budgets shift, and no seasonal trend should be treated as a forecast.

What the data does show is that an August CPM decline is not unusual. A publisher seeing softer pricing this month should put that decline in historical and market context before concluding that something is wrong.

The blended CPM doesn’t tell the whole story

The more interesting part of DataBeat’s August data is what happened underneath the overall decline. Display CPMs fell 20.3% month over month and were only 1.8% higher year over year. Video CPMs, by comparison, declined just 7.8% for the month and remained 40.1% above last year. CTV was also the most resilient device category, declining 7.6% while desktop and mobile experienced considerably larger pullbacks.

For publishers, those differences matter because a blended CPM can hide where demand is strengthening or weakening. A publisher with a meaningful video business may be experiencing a very different market from one heavily dependent on standard display. The same can be true across desktop and mobile, U.S. and international traffic, or even individual sections of a site.

When overall CPM falls, the next step should therefore be to break performance apart. Looking at CPM by format, device, geography, ad unit and demand source can help determine whether the decline is broad-based or concentrated in a particular part of the business. That distinction becomes particularly important heading into Q4, when increased advertiser spending does not necessarily benefit every type of inventory equally.

Demand path can change the picture

DataBeat’s analysis of Ad Exchange and Open Bidding performance reinforces the same point from another angle. Across the App, Web and AMP inventory it analyzed, Open Bidding generated higher CPMs than Ad Exchange, but the size of the difference varied considerably.

For App inventory, DataBeat reported a $1.32 Open Bidding CPM compared with $0.83 for Ad Exchange, a 59% premium. On Web, the difference was much smaller: $1.63 versus $1.47, or 10.9%. For AMP, Open Bidding generated $1.07 compared with $0.79 for Ad Exchange, a 35.4% difference.

Those numbers should not be interpreted as evidence that one buying route is universally better than another. Ad Exchange may provide greater demand scale or fill, while Open Bidding can generate stronger pricing where competition is deeper. The value of each path depends on both the demand it contributes and what that demand ultimately generates in revenue.

That is an important distinction because CPM alone measures price, not total monetization performance. A higher CPM paired with substantially lower fill or fewer monetized impressions may not produce more revenue. Conversely, a lower CPM can still be valuable if it brings incremental demand that would otherwise go unfilled. For publishers, evaluating demand partners and buying paths therefore requires looking beyond which one posts the highest CPM.

When CPM falls, look at the whole revenue equation

The same principle applies when diagnosing publisher performance more broadly. A 15% decline in CPM sounds concerning in isolation, but its impact depends on what happened to traffic, fill, impressions, and RPM at the same time.

If pageviews increased 20% and Page RPM remained stable, the business may be performing well despite lower auction pricing. If CPM increased but fill deteriorated enough to reduce Page RPM and revenue, the higher CPM is not particularly helpful. And if blended CPM declined while video revenue grew significantly, the topline metric may obscure an important area of strength.

This is why CPM should be evaluated alongside traffic volume, traffic mix, fill rate, Page RPM or Session RPM, and total revenue. Publishers should also compare month-over-month performance with the same period last year. A July-to-August decline can look very different when August is still materially outperforming the previous year.

Once those topline relationships are understood, publishers can go deeper into device, format, geography, ad unit and demand source to identify where the change is actually occurring.

Seasonality is context, not an excuse

There is a difference between recognizing a seasonal pattern and explaining away underperformance. DataBeat’s August numbers and Adapex’s historical data both provide evidence that softer late-summer CPMs can coexist with a healthier underlying market. But that does not mean every publisher experiencing a decline should simply wait for Q4.

If CPMs are falling roughly in line with broader market trends while RPM and revenue remain healthy, the movement may primarily reflect seasonality. If CPMs are declining much faster than the market, fill is weakening, traffic composition has changed, or Page RPM is deteriorating, there may be a publisher-specific issue that deserves attention.

The same analysis can uncover opportunity. Video and CTV held up considerably better in DataBeat’s August data, while its Ad Exchange and Open Bidding comparison showed that the value of demand can vary significantly by inventory type. Those are signals publishers can use to understand where demand is proving more resilient and where their inventory may have additional monetization potential.

August’s 19.9% CPM decline makes for an eye-catching headline. But the more useful story is underneath it. For publishers heading into Q4, understanding why performance moved and which parts of the business moved with it is far more valuable than the topline CPM alone.