How we do it
Combinatorial real time modeling
Online shopping represents a unique environment in which consumers reveal a lot about themselves. All online actions – every page view, every search query, time spent on individual pages, decisions to add items to the shopping basket – reveal something about the individual consumers – their interests and preferences.
Through a complex combinatorial modeling platform, MintFinder develops a personalized recommendation set for every consumer on every page in real time. MintFinder utilizes all sources of available data – order history, click stream history, product catalogs and meta data to construct the recommendation sets. These recommendations expose your customers to products that they are most likely to buy, resulting in increased sales.
Once an initial model is developed for your website, it is refined and improved through the continual analysis of your customer click stream data. It is important to remember that all of the customer shopping data analyzed by MintFinder is completely anonymous.
Implementation is easy.
Once the initial analysis is complete, MintFinder provides an opportunity to preview the recommendations before they are presented to your shoppers. This is to assure that the appearance and quality of the recommendations have your approval before they go live. We appreciate the importance of site aesthetics on the customer experience. In every case, we will work with you to develop recommendation set designs that seamlessly integrate into the look and feel of your site.
On-site and off-site personalization
In addition to displaying recommendations to your customers while they shop, implementing MintFinder gives you the ability to maintain contact with your customers after they have left your site. MintFinder offers a full-featured email marketing service. This service delivers personalized product recommendations via targeted email campaigns and messages. The ability to send personalized follow-up emails based on your customers’ most recent shopping history gives you additional opportunities for follow-up sales, and can be set into motion based on a number of different rule sets. We pride ourselves on flexibility, so we can also integrate with your existing email service provider if necessary.
How do you know if MintFinder is actually WORKING and increasing your sales? Simple. The A/B Split test. MintFinder’s impact is continuously monitored through the use of a control group. This control group views your website as it exists without MintFinder recommendations in place. During the trial period the control group is comprised of 50% of your web site visitors. The other 50% of your web visitors view your site with the addition of MintFinder personalized recommendations. The A/B split test allows you to directly compare key metrics – conversion rates and average order values – between the two groups. The A/B split test is not affected by any other marketing efforts that you may execute in parallel with MintFinder because both the A and the B groups are affected equally.
The A/B Split test results can be viewed at any time by logging on to the MintFinder dashboard. Our simple-to-read dashboard offers you continuously updated statistics and individual, yet anonymous, shopper details. We make all of the statistics fully transparent in real time. The MintFinder impact can be independently verified through third party analytics such as Omniture, Coremetrics and Google Analytics, so you can always be 100% certain that MintFinder is at work and is increasing your sales.