New Online Advertising Cluster Technology Created By

Share Article a leader in online advertising solutions released a new proprietary algorithm in May 2010 which allows it to target ads to clusters or groups of users based on user behaviour of the group. This unique technology allows Simply to deliver very targeted high performing campaigns for its advertisers. Simply has done studies which show increases of 35%+ using this new cluster technology for ad display. Online Advertising

...studies have shown a further +35% eCPM delivery, making Simply one of the top performance ad networks in the website advertising market.

At the forefront of online advertising in 2010 is, a division of Italian company Dada. In May it released two new algorithms which improve the targeting of ads on its network. leverages its strong partnerships with Yahoo’s Ad network (Right Media) and Google (Doubleclick Ad Exchange) as well as its own proprietary network of more than 6,000 publishers to display their advertisers’ ads on search engines and on the content network. But to understand why this is different from using Google or Yahoo directly, requires a closer “under the hood” look at the technology it is using to deliver ads for its advertisers.

Until now Simply used a mix of three different strategies to get the best click through rates for advertisers, and ad delivery is optimized in real time depending on which strategy is working best. The strategies which work best get a higher “weight”.

  • Contextual targeting – which analyses the content of pages on the network to deliver the most relevant ads
  • Behavioural targeting – which profiles web users based on their navigation behaviour determined by anonymous cookies on their browser
  • Yield targeting – which optimizes the campaigns in real-time to place ads on sites which are performing best in terms of click through or conversion

And these already deliver 107% higher eCPM than market average levels, according to SimplyLab results.
However in May this year, introduced its new Cluster technology, and studies have shown a further +35% eCPM delivery, making Simply one of the top performance ad networks in the website advertising market.

How Simply clusters users into groups

Each user is assigned to a “carrier” and a carrier is assigned certain characterstics depending on what sites they have visited or clicked on, and what searches they have done. Keywords which are less common (e.g. tyrannosaurus) will be more significant for clustering than more common keywords (e.g. hotels) and therefore assigned a higher weight.

The next step is to identify users similar to the carrier, or users who have performed similar actions in the same period, and by doing this, about 40 different clusters were identified. These groups are updated on a daily basis. Users within a cluster show similar types of website interaction, and by using techniques to measure the statistical significance of the clusters, Simply could confirm the validity of the new algorithm.

Once the clusters of users are identified, the best time slot is calculated in which to run a campaign. For example the algorithm can obtain a higher eCPM for some groups in the morning than at other times of the day, so ads can be delivered to this cluster of users during the optimal time period.

What does this mean for online advertising?

This new way of ad targeting means advertisers no longer need to experiment or blindly guess which sites or which times of day will work best for their campaign. They have a higher chance of running a successful campaign, where the system already knows when and where to place ads for optimal performance. This is particularly useful for smaller businesses who may not have the budget necessary to continuously experiment until they find the “right” websites to advertise on.

The latest developments are intended to help publishers increase their eCPM and also give advertisers better returns through better targeted online advertising campaigns, without the advertiser having to do anything different. To test the service go to


Share article on social media or email:

View article via:

Pdf Print

Contact Author

+44 (0)845 363 6172
Email >
Visit website