DMNews chats with Amy Kennedy, VP of marketing at

DMNews speaks with’s VP of marketing Amy Kennedy about the e-commerce site’s product recommendation tool and its evolution

Did have a recommendation function before RichRelevance?

We had a homegrown tool built by our IT team. It was better than no tool at all but it was limited in what it could do. It made recommendations based on a limited set of information and was not dynamically following the customer’s browse path or purchase history. It wasn’t a strong performer. We contracted RichRelevance, a personalization and recommendations provider company, to manage our recommendation engine.

Why is a recommendation tool particularly important to

A lot of wine is bought day-of for that evening, so we’re competing with a brick and mortar store that has a wine expert. Recommendations are the answer to that online. In addition, in a wine store customers are getting the advice of a single person, whereas RichRelevance takes into account your purchase history as well as that of people who buy items like you, and aggregates all that data into the recommendations. The tool expands people’s horizons and puts different products in front of them they might not have thought about otherwise.

What were your goals when you re-launched the recommendation tool?

Our immediate goals were to improve the recommendations that we had and see that they were driving revenue. We’re seeing an increase in basket size, but more importantly basket value.

Any other results?

The numbers speak strongly: units per order and average order value both increased. We’ve enabled ourselves to present the customer with more options on some pages that originally had a singular point of view. When customers got to a product detail page, they were faced with information on a single wine they’ve chosen. If you didn’t want that wine, you were at a dead end. The two recommendation modules on those pages now present both alternative and additional options and paths, and that has increased sales.

What’s next for as far as optimizing the e-commerce experience?

We’re launching a recommendation module via e-mail and exploring other modules that leverage the core data, enabling us to have a better one-to-one consumer conversation in different mediums. Mobile marketing is also definitely on the horizon.

Related Posts