[ad.agio] uses a combination of machine-learning, statistical analysis, knowledge modelling techniques, data/text-mining and sophisticated database technologies to gain valuable data information about ad views and users’ behaviour across interactive channels. [ad.agio] automated and semi-automated optimization technology works on price, position, targeting, and behaviour.

• LOCATION BASED MODELS: [ad.agio] learns optimal campaign allocation allocation over multiple location (ad-space) to maximize the click-through rate and/or the conversion/lead rate. It does so by analyzing creative’s exposure, users response, and location traffic.
• eCPM Competion: [ad.agio] identifies the most performing ad-unit according to position and price, and automates the process of ad-serving to maximize revenue from your inventory. Based on CPC/CPA/CPL VALUE, [ad.agio] delivers the highest revenue generating campaign providing the best value for publisher’s unsold inventory. CPC/CPA/CPL campaigns value is counted according to the Effective CPM, which measures the weight of those campaigns in terms of CPM. Thus, it is possible to compare campaigns value regardless of their type.
• USER CLUSTERING: [ad.agio] classifies consumers in several segments according to their past behaviour. Each cluster contains people with a similar profile. To keep track of users' interest over time, clusters are updated continuously.
• ASSOCIATIVE RULES/COLLABORATIVE FILTERING: [ad.agio] collects preferences from a large number of users to build associative rules on their preferences. If many people interested in creative A have clicked also on creative B, as soon as a user exhibits some interest for A [ad.agio] increases automatically the chances to show him B too.
• 1-TO-1: [ad.agio] can perform a fine-tuning ad-serving optimization based on each single user’s past behaviour. E.g. banners previously seen/clicked may be discarded in following imps to the same user.
• CONTEXTUAL BASED: through sophisticated text mining algorithms [ad.agio] is able to recommend any content based on text semantic;
• TEMPORAL ANALYSIS: Objects seen/clicked are weighted based on their age