Live From NAA: Local SearchTakes Center Stage

I'm at the Newspaper Association of America Marketing Convention in Las Vegas this week. The general chatter in and around the conference grounds has a palpable air of online strategy development. These things have also had a thematic dominance of many sessions I've attended (albeit a self selected sample).

Today, a 90 minute session focused on the very broad category of local search. The always entertaining and insightful Greg Sterling did a good job keeping the conversation on track (and to the continued amusement of the audience, instituted an edict against sometimes-meaningless buzzwords like “leverage” and “solutions”). In a discussion with Greg following the session, he agreed that this topic has tentacles that reach into newspapers, yellow pages, shopping, social networking, video, mobile and of course online search.

Today the focus was not surprisingly on how these many areas apply back to newspapers – particularly with the strategies involved in bringing better search functionality to newspaper websites. An important anecdote came from an audience member who explained his local newspaper's choice to pull the plug on a one-box search feature when it proved ineffective after a few months.

Tim Judd, CEO of Search Initiatives answered that Google has raised users' expectations of search, so if you can't create a comparable search technology, you are destined to fail. Jim Michaels, Director of New Media for The Evansville Courier & Press answered this same question by saying that you have to be patient and less inclined to look for immediate traffic and revenue benefits.

Planet Discover's president Terry Millard meanwhile asserted that you need the marketing muscle behind new functionality or new websites in order to promote them. “It's not just that you build it and they will come,” seconded Teresa Lawlor, Marketing Director of MediaNews Group Interactive.

These challenges all get to organizational issues, particularly in newspapers where the luxury of being able to wait four years before monetization is realized (which is what Google did) is not present.

There are also technical challenges. Yes, Google has set the standard in search, but the discreet and segmented nature of newspaper content can make it difficult for their online operations to create the local search experiences they would like to. The goal in many cases is to integrate disparate forms of data – historically siloed in different search buckets – to create user-centric unified local search results.

This would involve, for example, returning search results for “BMW” with related news (auto reviews), listings (dealerships) and classifieds (private party listings). The range of search queries in news however is broader and more complicated than that of classifieds which in turn is more complicated than that of business listings, which are neatly bunched into a relatively small volume of search headings.

“In automating unified search, you will get roughly 60 to 80 percent accuracy in contextually relevant results,” said Planet Discover's Millard. The rest requires an investment of manpower to manually police the accuracy of the data and search results. This requires an upfront investment to create the foundation of a search index. This investment level and maintenance resources lessen over time, but that initial hump is difficult to execute and scares many newspapers according to Millard.

In terms of human power, there are also opportunities to allow users to police data with different “social search” models that layer on community interaction to go the last mile beyond where the algorithms take you. There are also more “passive” social search engines that don't require user tagging and reviews – sometimes a barrier to user adoption - but instead increase relevance in search results by using past click stream analysis or grouping users into certain categories of interests. Such platforms from Eurekster and Collarity are available in private label form for newspaper or vertical sites to plant on their websites.

Branding is also an important challenge: Is it better to build upon the current newspaper brand equity, or to build out separate brands without the “baggage” of the legacy brand?

According to Jay Small, there are a few scenarios where a new brand is advisable. One is when it is possible to obtain a city.com domain (i.e. Boston.com), which has shown effectiveness in gaining traction. It can also be valuable if you can build something that is significantly removed from the functionality and brand of the legacy product. This, like some online directories or standalone vertical search destination sites, is harder to market but can be a much better branding strategy and value proposition in the long run.

Along these lines, Terry Millard reiterated that you need to put in the proper marketing investment to expose the new brand, and that in these situations it's “all about marketing”. Teresa Lawlor agreed.

“To flip over an existing newspaper brand and integrate personalization entertainment, social bookmarking, and mobile functionality, this is a new sandbox in which you can do these things,” she said, “but you need the marketing dollars to see it through.”