When thinking about data feed optimization (DFO), the first step we talk about at SingleFeed is proper engine set up and feed delivery.  Did your products actually go live on the shopping engines?

The second step in DFO is the quality of the data feed.  Are you including more than just the basic product attributes?  Are you categorizing all your products properly?  Do your titles and descriptions include proper keyword phrases?  And lots lots lots lots more!

But as we move into the holiday shopping season where clicks and click costs increase, merchants need to give some thought to a third step: Quantitative Data Feed Optimization.  While that sounds fancy, I’m just talking about doing some basic math to make sure you’re not wasting money on the shopping engines.  For example, if you get 400 clicks/month on engine X for product Y and it’s costing you $0.40/click, and you’re not recording any sales through those clicks, then maybe you should suppress that SKU from that shopping engine.  Each merchant will have a different profit margin, but the basic rule applies that if a product is not performing at all on a shopping engine, the maybe you shouldn’t send it to that shopping engine.

Some data feed management companies jump right into this type of quantitative optimization without first concentrating on the quality of the data feed, which means that they are suppressing SKUs unneccesarily and usually cutting your catalog in half.  That’s wrong.  However, if you’ve put in the time and effort to provide the shopping engines with great data, and you’re not seeing the results, don’t be scared to suppress some SKUs.  Some products just won’t work on all the shopping engines.  That’s fine.  It gives you the opportunity save money or shift that ROI negative spend to a better performing product.