...
| IMPORTANT! Under delivered ads. The only change is the impression goal to be met. It does not make ad delivery any faster. The optimization of under delivered campaigns is made through the weight and/or priorities, spaces and targeting criteria management of all ads involved. For more information, visit the “Delivery” category. |
Even Distribution/redistribute if over delivered 
In this mode, ad traffic is evenly distributed throughout the scheduled time. If for any changes the ad is over delivered, a new traffic distribution will be calculated so that the remaining traffic is evenly distributed throughout the remaing time.
Recommended option.
Delivery Schedule:
Delivery schedule when an ad is under delivered: 
Even Distribution
The system delivers the impressions the ad has assigned between the start and end dates evenly. This option does not re-calculate traffic distribution if any change is made to the number of impressions. Thus, it is not a recommended option since, if for any reason, the ad is over delivered, the system will maintain the same traffic distribution, and will probably stop for hours or days until new traffic is assigned.
Not recommended option since, as it does not re-calculate traffic distribution, ads can end before the end date set.
Even Distribution in Two Halves
This type of distribution allows to manage the delivery of a number of ad impressions for the first and second halves of the ad. Halves are calculated based on the ad total duration, its weekly schedule, and multiple start and end dates, if this is the case. The system then multiplies the total amount of traffic for the percentage corresponding to one half, and assigns traffic evenly within that half.
When selecting this option, you can set the percentage of actions to be delivered in the first half. This value can be set by default from the account Global Preferences menu.
Recommended option to avoid ad under delivery.
Adaptable Distribution
This option allows traffic distribution to adapt to web traffic time load only in websites with more than 15,000 daily impressions. In websites with less traffic, distribution performs as “even distribution/redistribute”.
If an ad is associated to more than one site, at the time of distributing traffic, the system takes into account a distribution average of all associated sites. Site impressions are not considered so as to get an accurate average. This may cause some problems if the ad is associated to sites with very different traffic distribution ratios. For example: a site with a rush hour from 10 am to 6 pm, and another one with a rush hour from 7 to 11 pm.
The ad conclusion -the time when the system starts to calculate “blocked” hours- is considered as the effective “working”48 hours before the ad ends. In other words, hours where the ad was indeed active (inactive hours for weekly schedule or multiple dates are not included).
In these last 48 hours, the system looks for ads with higher priority, sponsor type, without freq caps and action limits per hour/day/week/month. To set an example, if 10 hours and 10,000 impressions are left, the system will assign 1,000 impressions per hour. With the new feature, if the system realizes that the ad will be blocked 5 hours out of those 10 hours, then it will assign 2,000 impressions per hour during the first hour, and then it will assing a lower amount of impressions if the ad is blocked during the remaing hours.
To determine if an ad is blocking another one, only ads that block fully are considered. In other words, ads that do not allow others to run even 1 impression. For example: if an ad runs in 10 spaces with a priority level of 2, and another one is sponsor type with a priority level of 1, and runs in 9 spaces out of those 10, then it is considered that the ad is not blocking it.
Blocks are determined based on country targeting only. No other targeting criteria are considered. For example: a sponsor ad with priority 1 targeted to iPad blocks a CPM ad with priority 2 unfiltered.
Results obtained from an ad with this type of distribution:
