If yours is anything like a typical e-commerce website, the vast majority of your daily visitors are leaving without making a purchase. And a small percentage of the users who do transact likely go on to become your best customers, accounting for a large amount of your overall profit. This, coupled with the fact that it’s far more efficient to keep a good customer than it is to acquire a new one, should prompt an obvious conclusion: you’ve got to discover these best customers! And for the shoppers on your site right now, you’ve got to predict which may become your best.

Below are some of our key areas of innovation which make these predictions possible:

• Massive Data Collection

  What: Lightweight JavaScript tags which load quickly & collect information about every action your users are taking on your site.Why: Analysis of massive amounts of data enables unparalleled precision in ad personalization. This huge amount of data is the input to our predictive Customer Quality Scoring models.

• Customer Quality Scoring

  What: Using almost unlimited computing power in the cloud, our predictive behavioral models evaluate thousands of data features (e.g: time spent on site, referring traffic source, on-site search terms, type of products browsed, whether the shopper is new or repeat, etc.) and train on millions of examples.
  Why: These models are key to TellApart’s ability to “tell apart’ high quality visitors from the rest. With so much data, our models can predict customer value with an order of magnitude greater accuracy than models created by hand.

• Personalization

  What:A dynamic personalization engine for ad creatives serves custom ads to shoppers and makes optimization decisions with every ad impression, and real-time product importing enables us to work with retailers who sell hundreds of millions of unique products.
  Why: Personalized display ads yield the highest visitor response rates, often 10x greater than regular banner ads.

• Real-Time Bidding

  What: Real-time auction-based bidding for each ad impression (calculated in about 10ms), billions of times per month.
  Why: RTB allows us to differentially value every unique impression on every unique user—enabling us to be much more intelligent in our ad buying, targeting & optimization. TellApart is integrated with the largest sources of RTB inventory in the US.

• Elastic Scaling

  What: A cloud-based infrastructure that allows us to easily scale up computing power to accommodate sudden traffic increases caused by seasonality, one-day sales or other spike-inducing events.
  Why: We scale with our clients, as their traffic & revenue grow, our services are there to amplify their efforts.