We mixed things up a bit at the Shopping Search Tactics session at SES this year. I did a quick beginner level intro to shopping search, Jake Berry of Marchex gave an intermediary level guide to shopping search, Brian Mark of Toolbarn discussed how he and his company handles data feed marketing, and then I talked some more about data feed optimization.

Here’s the beginner level presentation outline…

How to view shopping search engines:
-Consumer: Knows the question shopper is asking, Simplifies the process of finding the right merchant (high customer feedback rating, best price)
-Merchant: Delivers a highly targeted shopper (someone with the intention of buying - although not necessarily online), Another way to get listed high in organic results, Another way to get listed high in paid results

The different engines:
CPC based: Shopzilla (BizRate), Shopping.com (Dealtime, PriceTool), Yahoo! Shopping (Yahoo! Tech), PriceGrabber, NexTag (Calibex), Become, Smarter, CNET Shopper, Pricewatch, Streetprices, Sortprice
Free: Google Base/Froogle
Mixed/Crawlers: Pronto, Shopwiki, TheFind
Affiliates of the CPC based: PCWeb, USA Today, Camcorder Info, Sprenzy, AOL Shopping, MySimon, Ask Shopping, MSN Shopping

Some data feed fields:
Name, Description, Price, Link, Image URL, SKU, Product Type, Category, Become Category, PriceGrabber Category, Shopping.com Category, Shopzilla Category, Gender, age range, size, color, sale price, shipping price, shipping weight, actor, producer, memory, tax percentage, promotional message, keywords, format, coupon, coupon expiration date, monitor_size, ski_size. In other words, this can be a lot of work.

How to create:
-Download the feed specs from each engine you want to use
-Pull from catalog as much information as possible
-Manually enter the rest, manually optimize the catalog pull
-Categorization does matter
-Take baby steps. If you have 50,000 products, this is going to be a painful process. Just as you have someone dedicated to SEO/PPC, eventually you might have someone dedicated to feed management. You might want to start by listing your 1000 most expensive, highest margin products

When creating a data feed, think about:
-Read the directions
-These guys act just like search engines
-Add as much information as possible
-Automated XML solutions aren’t optimal
-Test, test, test

This is not easy (if you want to succeed):
-Define success metrics (CPA, ROAS, Overall Sales)
-SKU level/Engine level/Basket approach
-Tracking – Google Analytics - Omniture
-Testing – Pick 3 engines, Budget of $500
-Optimizing – Refine listings
-Repeat

Final notes:
-Don’t expect the shopping engines to provide excellent customer service unless you are spending over $50K/month
-Data feed optimization (DFO)/Comparison Shopping Optimization – categorization matters, bidding strategies
-Monitor costs very closely
-This may not work on all engines and for all products (you must test, monitor, refine, and test some more)