Key Metrics in Your Ad Server Account
In programmatic media buying ad servers are the most reliable tools for comprehensive measurement and reporting on your digital campaign A worthy ad server offers real time ad campaign data, highly customizable build-your-own reports, and a tailored interface. The suite of reports available through an industry leading ad server will greatly exceed those available through a supply-side system, and of course it will consolidate and de-duplicate the campaign measurements across all the supply.
In this article we'll be giving an extensive review of the metrics to consider before your first ad campaign kicks off. We will also delve deep into the main reporting numbers crucial for advertisers, publishers, and networks.
Basic Metrics
Most ad server reports are flooded with numbers which represent the increment of impressions for a particular metric. Every single ad will have 5 different numbers attached to it:
- Requests
- Impressions
- Viewable Impressions
- Clicks
- CTR
Requests are the number of publisher's attempts to show an ad to the user counted by the ad server.
Impressions are the industry's standard "trading currency", since the majority of cost models are performed on cost per thousand impressions (CPM). Impressions are a count of the number of ad requests coming from a user's browser for a single ad. Viewable impression only registers an impression when the ad is in view on the users screen. It's vital to understand at this point that the ad might not get filled. An impression is an opportunity to show an ad. A Viewable impression occurs when an ad is actually shown.
There can be a vast difference in the number of counted impressions on the ad server versus the number of viewable impressions.
Some argue that the industry factors in the difference between the cost of an impression, and the cost of a delivered or viewable impression by keeping a low cost where a delivered impression or viewable impression may be provided at a higher charge because of the delivery verification that surrounds it.
Clicks show the number of users who clicked on an ad. As a component of the click redirect, the ad server is pinged and the master database of the ad server appends a row stating that a particular cookieID clicked on an ad at a particular time from a particular website.
CTR composes of the number of clicks divided by the number of impressions to determine how many viewers successfully became "clickers". Usually CTR is expected to be a very low number from the start. For example it can be lower than 1% of the total impressions for standard desktop banners. On a big enough scale this is still a considerable number of interested users, but it is difficult to determine what caused the click itself. Sometimes the creative may be resulting in the click action and sometimes it may have just been down to the correctly targeted audience. If your campaign CTR is lower than expected, consider optimizing the aforementioned parameters.
Revenue metrics
After checking delivery reports briefly to make sure that the campaign started without a hitch, you may need to take a closer look at real-time reporting statistics. Real-time analytics allows you to make decisions immediately right after the user engaged with an ad.
Viewability
It is the percentage of ads which were actually seen by a user. A viewed impression is considered valid for being at least 50% on screen for at least one second. If a site has 100 ad impressions, of which 50 are counted as viewable, the viewability rate of the website would be 50%

Viewability is not that easy to measure. It works with JavaScript tags, whicht means it is more complex to execute on Instream and Mobile ads. Viewability is also much more difficult for the Publisher to control. Many Publishers are well aware of the fact that a fair amount of their inventory may never be seen, but won't be aware of it until the ad is served. Such Publishers might argue that their pricing for the inventory is adjusted to reflect the possibility of wastage from viewability.
To avoid this, we advise both publishers and advertisers to negotiate the preferred percentage of viewability before going live. Some technologies can help avoid this wastage by making a request to the ad server, but only when the placement is coming into view, rather than loading the ad tag when the whole page loads.
eCPM rate
This is how much money you earn from the advertiser for every 1000 filled impressions. The formula is easy:

For example, if you earned $8 from 10,000 viewable impressions. Your CPM is $0,8, as seen in the calculation:
8 / 10 000 *1000 = $0,8
Optimization: There are factors that can impact a publishers eCPM such as lost impressions, low fill rates, and discrepancies. For a publisher it's crucial to know exactly how much they are earning from an ad network. Ad Servers do the eCPM optimization based on the estimated conversion rate and CTR.
Fill Rate
This term is used to quantify the number of ads publishers return compared to the number of ads advertisers requests. The formula here is very simple:

A 100% fill rate is a hard-to-achieve golden standard meaning that every time there was an opportunity to show an ad, it was shown. A high fill rate generally increases a website's eCPM and earning potential. However, missed ad opportunities happen if the user leaves the website before the ad loads, page is timed out, or network issues between the ad server and publisher.
Conversions
Conversions are any targeted actions that you want visitors to take on your site. Conversion is a measurement of campaign success, particularly where the user is driven to make a purchase online.
Optimization: Conversion Rate Optimization (I2C) increases the Banner Weights within a Campaign based on Actions such as conversions, downloads, installations, etc. I2C optimization works for all methods of action tracking available in Epom Ad Server.
Discrepancy
Discrepancy is the difference between the number of ads that the publisher's ad server has sent and the number of ads that the advertiser's ad server received. Some people calculate discrepancy differently, but the simplified formula is:

Discrepancy measures the loss of impressions between a publisher's ad server and the advertiser's server. As you will typically get paid based on the advertiser's numbers, it's important that your numbers match up. Don't expect a perfect match!
Discrepancy of up to 10% is considered normal, but lower is always better. Basically, the percentage of discrepancy means that a proportional percentage of your impressions are wasted without being monetized. It's possible to see a negative discrepancy, but as a rule, negative discrepancy indicates a configuration error or timezone problem. Read the tips to reduce ad discrepancy.
Hope these small guidelines on ad server metrics will help you start your journey safely.
Have more questions about ad server metrics? Contact the Epom representative and we'll be glad to help.