A Guide to AdWords Applied

I’m a little late posting it, but last week’s visit to the Web 2.0 Expo threw off my schedule in getting this column live (let alone getting it to MediaPost). It continues in the extended entry.

A Guide to AdWords Applied
By David Berkowitz

“IF THE Google model is so successful, it seems
reasonable to think that everybody should be in the Google business.” — Pixsy
CEO Chase Norlin

The quote’s taken out of context from an email I received from
Norlin last week, and that context will be good fodder for a future column. He
allowed me to borrow it to set the stage for what Guidester is doing for online
consumer electronics retailing, and explain why Guidester should prove to be a
beacon for future innovations in search, online publishing, and ecommerce.

Guidester provides search technology for retailers’ sites, making
it a sort of Google Custom Search Engine for ecommerce. That analogy can only be
extended so far. I’ll illustrate it with an example Guidester CEO Joe Chin
showed me over lunch last week, though I’ll use sample advertisers that aren’t
intentionally reflective of Guidester’s clients:

1. A consumer, we’ll call him Devin, arrives at Buy.com. Guidester plays no
   role in bringing Devin there. Buy.com can do that any way it knows best, whether
   it’s through a search engine or a Sunday circular.
2. Devin’s in the market for a digital camera. Devin clicks the “digital
   camera” tab on Buy.com’s site, and Buy.com here allows him to check various
   boxes and refine what he’s looking for. Devin is brand-agnostic, and he searches
   for all models under $500.
3. The Guidester-powered search delivers an array of search results, with
   models from Panasonic, Sony, and Canon coming up as the most prominent listings.
   Guidester used three factors to order the search results: the click-through rate
   (CTR) for each listing, the cost-per-click bid that the electronics manufacturer
   submitted, and the retailer’s preference on which listings should rank higher.
   In other words, it’s an AdWords-esque model that the retailer applies to its own
   site.

The way Guidester applies the model is even more fascinating when
exploring some of its nuances and ramifications.

Redefining Natural and Paid Search

It’s fun when you can be in this business for awhile and still
feel naïve. At first, what Joe Chin was telling me felt like heresy. How could
these search results be sold to the highest bidder? Yet it’s a perfect model for
retailers, as there’s no need for purely “organic” search on a retailer’s site.
How else should the results be ranked? Guidester offers answers to that question
too, as the performance of each listing (its CTR) is one of the major factors,
and if there are no advertisers for a particular search (or a “browse,” as this
may be better described), then the performance is all the more important.
Additionally, retailers can exert a degree of control. If Buy.com thinks Sony’s
a better brand than Casio, then it can offer its own vote. Yet if users click
Casio’s results more, or if Casio submits a high enough bid, then Casio can come
out on top.

Meanwhile, Guidester comes with built-in trademark protection. If
Devin has a brand preference and wants to see Sony models under $500, advertiser
bids aren’t factored in. Guidester only applies its three-factor approach when
the consumer lacks brand preference.

Bringing New Advertisers to the Table

It’s hard for manufacturers to participate in the AdWords
marketplace in many situations, as on a search engine, they’ll wind up competing
more with their distribution partners than they will with competitors. A
retailer’s site, however, is perfect territory to sell to the highest bidder.
It’s similar to bidding on slotting fees for shelf space in a retail store.
Here, however, if consumers keep bending over to reach for items that aren’t at
eye level, those other items will gain higher prominence.

Applying AdWords Beyond Retail

Google’s Custom Search Engine allows publishers to offer preferred
sites to appear in natural search results, yet the performance of each search
listing doesn’t matter, making Guidester more intelligent than the CSE. It’s
interesting to think of how else this model could be applied.

For instance, say Google News, or another news aggregation site,
allowed publishers to bid for a more prominent ranking on its homepage for any
relevant stories. Would news sites take the bait? While some would denounce the
model, it’s conceivable that many others (especially all those who already spend
healthy sums with AdWords) would participate in the marketplace. Granted, if you
look at how media companies are working with YouTube, many news publishers would
presumably bid for Google News rankings and sue Google at the same time. Don’t
you love this business?

The applications of the Guidester and AdWords models can be
extended just about anywhere, though it’s worth following the Guidester approach
of deciding under which circumstances advertisers can bid and which ones allow
for more “natural” listings.

More and more though, the idea of advertisers bidding for
placement feels more natural in its own right.