The Future of Local — Revisited

I’ve written about this several times over the course of the past couple of years: the future of online geotargeting is demographic targeting. I raised the issue again briefly today in a panel about local search at the SMX East show. That’s what prompted me to do this post.

The historical discussion of local search and local online advertising has largely been about small businesses: e.g., plumbers and pizza. Over the last two or three years I’ve tried to expand the local conversation to encompass brands, product sales and to generally reflect the relationship between the Internet and offline transactions. But in many ways the most interesting future for local is something else.

Stepping back for a moment, direct mail is arguably the single biggest advertising/marketing category in the US:

Source: comScore, Advertising Age, Magna Global, Newspaper Association of America; Radio Advertising Bureau; Simba; Outdoor Advertising Association of America; Interactive Advertising Bureau (IAB)

Direct mail’s online cousin is search. However, offline direct mail is quite a bit more provocative a tool than search in many respects. Online a search engine generally knows my query (and maybe has some behavioral data) but doesn’t know much else about me.

Demographic data from public and private sources are tied to zip-level targeting in offline direct marketing. Elaborate databases exist containing layers of information about people who live in particular zip-codes throughout the US. But that type of demographic targeting isn’t possible right now online at any level of scale or accuracy. One might be able to get a diluted version of demographic targeting — a bastardized version of cable-TV or magazine targeting — but it’s pretty weak.

Once online location detection becomes more precise, however, census and other related household composition, education and income data become available to enable direct-mail type targeting, at the zip or neighborhood level, on a national scale. Hopefully I’m making myself clear. (I wrote about this previously.)

As I’ve said, location is coming to the browser (cell-tower/Wifi triangulation) probably next year.

So once Google or Yahoo know where I reside with greater precision than IP-based lookups, then I can be profiled as part of a market segment or group (using existing data) based entirely on where I live. But the radical thing is that this could be done on a national scale with relative ease. Marketers could thus buy audiences or market segments on a national level through geotargeing. Display could also following the same model.

In the search context, imagine an advertiser-facing interface with a bunch of check boxes or buttons that correspond to income, household composition, race/ethnicity, education and so on. Say I’m an automaker or cellphone carrier and I have a clear profile of who my customers are. I could target them according to their group characteristics (income, etc.), based on location.

After I as a marketer selected the desired segments, the system would show me all the zips or geographies across the US where my customers tend to reside, based on the available data already resident within the platform. I then buy only those zips or areas where my potential customers reside. It’s much like today’s IP-based geotargeting — only much richer.

Indeed, the underlying demographic data already exist. The only missing piece in my equation is being able to zero in on people in their homes and neighborhoods when they’re online, as direct mail marketers do offline. That’s where the browser with Wifi comes in. More precise location detection in the browser unlocks the targeting possibilities I’ve described. Location or geotargeting thus transforms into a potentially effective method of doing demographic targeting online — in an extremely scalable way.

Sure there are challenges and complications: what about Internet usage at work for example? People don’t typically live and work in the same locations and profiling people in the workplace is more difficult. What about getting location from the browser to the ad platform? I believe these challenges are quickly surmountable.

Demo-geotargeting could also be done with graphical ads or, potentially, video as well as search. Under any of these scenarios, resident of “community A” might see a different ad than resident of “community B” on the same page or in response to the same search. And search keywords and/or behavioral information potentially provide additional layers of data that can improve targeting further still.

Whether for branding, direct response or some hybrid of the two, this coming capability will dramatically change the way that marketers think about geotargeting. And that may begin to attract more of the many billions in offline, direct marketing dollars.

This new still-hypothetical form of “local-demographic” targeting won’t eliminate the more conventional forms of geotargeting. It will, however, provide a very powerful complement to them.


18 Responses to “The Future of Local — Revisited”

  1. Mozilla Geode: Location for Firefox « Screenwerk Says:

    […] Geode: Location for Firefox I was just talking about location in the browser this morning. Just a few minutes ago I stumbled across Mozilla Geode, a Firefox plug-in that captures user […]

  2. Justin Skogen Says:

    We can provide industry leading Geo Targeting for your site(s)….
    Feel free to contact me at your conveneince to learn more…

    Justin Skogen

  3. Jon Says:

    I see your point Greg, but…

    The reason Direct Mailers use demographic data is to give them clues as to what the prospect might buy, and at best that is a “good guess” based on aggregated data about what people with similar demographics have bought in the past.

    Where online advertising like Adwords scores higher, IMHO, is that it displays the ad depending on what the prospect is actually searching for (or which webpage they are reading) right now. Which gives much stronger clues as to what they might buy right now.

    Of course, if you combine the two, then you are really cooking!

    Cheers, Jon

  4. Greg Sterling Says:


    That’s precisely what I’m suggesting: a combination of keyword to indicate intent with profiling based on residential location.

  5. Jon Says:

    Ah, perhaps I didn’t see your point as clearly as I thought I had 🙂

    Thanks for the clarification

    How about a new buzzword – “geo-demo-SEM” ? (maybe not!)

    Cheers, Jon

  6. Greg Sterling Says:


  7. Yesterday, Today, Tomorrow Says:

    […] The Future of Local — Revisited by Greg […]

  8. Stever Says:

    Adwords already has demographic bidding options but its tied to the content network and the demographic data of a particular websites users. But as soon as IP targeting can get down to a scale of zip codes then demographic targeting combined with searcher intent, on the search network, will make any marketer cream their jeans.

  9. Greg Sterling Says:

    The demo targeting based on content is not really demographic targeting; it’s largely unreliable. And your second point is precisely what I’m describing in the post.

  10. Will Scott Says:


    I think one of the things you miss in your analysis of direct mail is the ability to target publications with whom you have a shared user base.

    So more than slicing and dicing a generic list, you can effectively get into the conversation stream with the data of a trusted participant.

    Direct mail is too often associated with a shotgun approach, a la Publishers Clearinghouse and infrequently associated with the sniper fire of Conde Nast or Harlequin.

    So here’s the tie-back to online… Email. We don’t talk enough of email as part of the marketing mix but I would suggest it’s an unmetered component which soon will rival direct mail in ad spend.

    It’s not all pharma spam — it’s often targeted, by invitation and welcome.


  11. Greg Sterling Says:

    I’m sure you’re right Will. I’m not a marketer and so don’t know all the angles.

    Agree re email. I think, along the lines you discuss, that there are many possibilities that browser location awareness opens up.

  12. Greg Sterling Says:

    Current demographic targeting in adwords is quite limited and weak:

  13. Report: Yahoo! Adds Zip-Level Targeting « Screenwerk Says:

    […] Trulia). Urban Mapping also has a neighborhood-level ad targeting offering. (I’ve also written quite a bit recently about how location in the browser will affect online […]

  14. Darrin Clement Says:

    Isn’t IP-based targeting pretty inexact? I mean, in my case, I’m in Vermont but my IP address shows up in New Hampshire. (We’re close to the border.) If there are errors at the state level, I’m not sure how IP targeting is reliable at all. Maybe I’m missing something?

  15. Greg Sterling Says:

    IP targeting is somewhat hit-and-miss

  16. Search Beyond Search Engines « Screenwerk Says:

    […] thing not discussed in the piece is my theory about search becoming a demographic targeting tool and thus a much more complex advertising vehicle when location awareness becomes more […]

  17. Loki Brings One-Click Location to Publishers « Screenwerk Says:

    […] Brings One-Click Location to Publishers By Greg Sterling As I’ve written many times in the past, as location on the PC becomes more precise lots of interesting things happen for both users and […]

  18. Offline Local Marketing Says:

    Just when I thought internet makrketing was the bomb…

Comments are closed.

%d bloggers like this: