Low fill rate is one of the most frustrating symptoms for a publisher: you see traffic on your site but not all of your audience generates revenue. Understanding what causes this problem and which levers actually work is the first step to fixing it. Some depend on you, others on factors outside your control, and some require a change in your monetization structure.
What is fill rate and why it matters more than CPM
Fill rate is the portion of your impressions that actually sells. If your site generates 10,000 impressions in an hour and 7,000 of them receive an ad, your fill rate is the ratio of those 7,000 to the 10,000 total.
The reason fill rate matters more than CPM is simple: a high CPM with a low fill rate can generate less total revenue than a moderate CPM with a high fill rate. An ad that doesn't sell generates no money, regardless of what its price would be if it did.
Many publishers focus only on CPM and neglect fill rate, or assume it's beyond their control. That's not entirely true. There are causes you can address, and there are changes in monetization strategy that can transform your results.
Why fill rate drops: technical and demand issues
Low fill rate comes from two families of problems: technical and demand-related.
On the technical side, the most common issue is that ads don't load in time. If your Header Script is too slow, the auction delays and buyers leave. If an ad is poorly positioned or out of view, the buyer doesn't see it and doesn't bid. If there are conflicts between ad scripts, one overrides the other and the impression is lost.
On the demand side, the problem is that buyers simply aren't interested in your inventory. This happens for several reasons: your geography doesn't match the advertisers who pay more, your content is on a brand safety exclusion list, your traffic profile doesn't fit what they're looking for, or there simply aren't enough buyers competing for your space.
Geography, content, and seasonality: what drives fill rate?
These three factors have enormous weight on fill rate and almost none are directly under your control.
Your audience's geography is the most determinant. A site with traffic mainly from Europe or North America will see a different fill rate than one with an audience in emerging markets. Advertisers concentrate budgets in certain countries, and if your audience isn't there, there's no demand competing for your inventory.
Content matters because buyers apply exclusion lists by topic. A site about finance or sports will see more demand than one in a category advertisers avoid for brand safety reasons. It's not a judgment on content quality; it's that programmatic demand concentrates in certain topics.
Seasonality is real. The same site performs differently in a quarter with high advertiser budgets than in January. It's not something you can change, but it's important to know so you don't confuse seasonal decline with a structural problem.
How to tell if your fill rate is really low
The most useful comparison isn't with another site, but with yourself over time.
If your fill rate was higher three months ago and is lower today, something has changed. It could be a technical change on your site, a shift in your traffic, or a drop in demand for your topic. If it's always been the same, then that's your baseline: whether it's low depends on your context.
Some contexts are better than others. A high fill rate on a European news site is different from one on a niche site in an emerging market. The metric only makes sense if you compare it to yourself or interpret it within your specific context.
- Check if anything changed in the last 30 days: a redesign, a change of ad provider, a traffic spike or drop
- Compare month to month in the same period as last year to isolate seasonality
- Observe if fill rate drops in certain formats or positions, or if it's widespread
Low fill rate: waterfall versus simultaneous bidding
This is where the technical structure of your monetization has the biggest impact on fill rate.
In the waterfall model, the ad network asks each demand partner in order: first the first one, and if they don't respond with a sufficient price, it moves to the next. The problem is you almost never get the best available price. The second on the list might be willing to pay more, but you never know because you already accepted the first partner's response.
In the simultaneous bidding model, multiple buyers bid at the same time for the same impression. This is what header bidding does: it sends the request to multiple demand partners in parallel and accepts the highest bid. The simultaneous bidding structure allows more buyers to compete, which can improve fill rate compared to a waterfall, depending on your specific context.
The difference in fill rate between one model and the other can be significant, especially in markets where demand is fragmented among several buyers. If you use a single network or a waterfall, you're leaving money on the table.
What you can change today without waiting for a new partner
Although some factors aren't under your control, there are actions you can take right now.
- Review your ad positions: an ad above the fold always performs better than one hidden at the end of the article
- Reduce ad density if it's excessive: when there are too many ads together, buyers reject the impression due to brand safety policy
- Make sure your Header Script loads fast: any delay in the bidding script is a delay in the auction
- Verify your ads.txt is correct: if buyers can't find authorization, they won't bid, and your fill rate drops automatically
- Test different formats: a native format might perform better than display in the same position, or a sticky better than a fixed ad
How we work on it at ADEQ Media
Low fill rate is a symptom that there's available demand but it's not reaching your inventory. At ADEQ we approach it from two angles.
The first is technical: we place your Header Script, verify your ads.txt is correct, and adjust ad positions during the first 15 days. Those 15 days aren't a formality; it's the period when we see with real data what works on your site and what doesn't. If a position underperforms, we move it. If a format is intrusive, we remove it. It's continuous work, not just initial setup.
The second is demand-side: we use header bidding with multiple demand partners competing simultaneously instead of a waterfall. That means each of your impressions receives simultaneous bids from multiple buyers, not one after another. In verticals like Forex and betting, we also negotiate direct campaigns at fixed CPM, which is demand that doesn't depend on what bid came in that day.
If you want to see what the result would look like in your case, fill out the evaluation and we'll prepare an estimate with no commitment.
What could your inventory earn?
We analyse your site and tell you what can be improved. No commitment.
Evaluate your site