Return your Census Form for Improved Targeting!

If you’re a cit­i­zen of the good ‘ole US of A, you already received your 2010 Cen­sus in the mail. For those of you read­ing this from over­seas, every 10 years, our gov­ern­ment asks us to self-identify, to pro­vide basic infor­ma­tion about the peo­ple in our house­hold in a Decen­nial Cen­sus. The infor­ma­tion we all pro­vide is used for myr­iad pur­poses, every­thing from allo­cat­ing dol­lars from gov­ern­ment pro­grams to pro­tect­ing the rights of minor­ity vot­ers, to…targeting. Busi­nesses have long used Cen­sus data to deter­mine mar­ket share, find pock­ets of highly qual­i­fied can­di­dates and to mea­sure local and regional suc­cesses.
His­tor­i­cally, one of the com­pa­nies that has ben­e­fited most from the cycle of refresh­ing Cen­sus data is Nielsen Media Research Group, the same folks that cre­ated the con­cept of DMAs (Des­ig­nated Mar­ket­ing Areas) that are used in Site­Cat­a­lyst geo-segmentation reports. While offline mar­keters who work in the Direct Mail/Database Mar­ket­ing field are keenly aware of DMA data, many online-only mar­keters haven’t fully unlocked the power of this report­ing.
The 2010 Cen­sus will take nearly two years to tab­u­late and pub­lish, but when it’s avail­able, mar­keters around the coun­try will see which DMAs have grown and which have shrunk rel­a­tive to one another. In Site­Cat­a­lyst, the DMA report is valu­able, but rather sta­tic. Obvi­ously, we know what are the largest cities and mar­kets in the USA, and vis­i­ta­tion from those DMAs gen­er­ally will not change much over time on a macro level. Regard­less of where you’re based in the US, unless your busi­ness focus is highly regional, your top 15 DMAs prob­a­bly include many of the same ones shown in the report below.

Inter­est­ing infor­ma­tion, but hardly sur­pris­ing. Or action­able for that mat­ter. So even though Easter was last month, my cab­i­net is still loaded with Marsh­mal­low Peeps (the Peeps Bun­nies, to be hon­est. The “old skool Peeps” have too high of a marshmallow-to-sugar ratio). But any­how, here is one of my favorite “Easter Eggs” in the Site­Cat­a­lyst user inter­face. This will be espe­cially valu­able if you do spot-market adver­tis­ing, want to deter­mine lift within a mar­ket, or under­stand where you have good mar­ket pen­e­tra­tion and where it’s lack­ing. All this is just a few clicks away. First, run the Vis­i­tor Pro­file > Geoseg­men­ta­tion > US DMA report that’s shown above, then nav­i­gate to the top right of your screen and click on the high­lighted area that says “Vis­i­tors Per Capita”

Voila! Now you have an inter­est­ing and action­able report that will look some­thing like this:

DMA Pen­e­tra­tion report in SiteCatalyst

So…what in blazes does this mean? To help explain, I’ve added shad­ing to the report. The mar­kets at the top (Boston, Austin, Wash­ing­ton DC) show higher than aver­age mar­ket pen­e­tra­tion. Those that are shaded are where this fic­tional com­pany under-indexes against US panel and Cen­sus data. You can inter­pret the data for Boston to mean that out of every 10,000 peo­ple in that DMA, 61.4 of them vis­ited the site in the given time frame. That is nearly 2X the national aver­age, as rep­re­sented by the field show­ing Boston’s pen­e­tra­tion is 99.9% above the national aver­age. That’s not the area code in paren­the­sis next to the DMA name, but rather the DMA num­ber that Nielsen assigns. You’ll note that New York has dropped from 1^st on the ranked report to 10^th on the pen­e­tra­tion report, but NYC still over-indexes com­pared to the national aver­age, show­ing better-than-average vis­i­ta­tion from that cru­cial mar­ket.
Except for Kansas City and Chicago, this fic­tional site has sub­par pen­e­tra­tion in the Mid­west. Mar­kets like Cleve­land, Colum­bus, and the Twin Cities show lower-than-average vis­i­ta­tion. If these mar­kets are impor­tant to this site’s over-arching strat­egy, they need to invest locally. So, what hap­pens if they do invest? Can we track that in Site­Cat­a­lyst too? Really, would I have asked the ques­tion if the answer were ‘no?’ Am I actu­ally hav­ing a dia­logue with myself right now? Yes? Sorry. Back to our example…this com­pany sees they are weak in Colum­bus, Ohio, and so they run a quick test cam­paign from Feb­ru­ary 14^th through the 18^th tar­get­ing vis­i­tors from the Colum­bus mar­ket using dis­play ads on major search engines. Did it work? Let’s see.

To get the view below, first use the fil­ter on the top right to choose the DMA num­ber for the mar­ket you wish to track. Columbus’s is 535 (but of course, you already knew that, right?) On the result­ing screen, switch over to the “Trended” view and you’ll see a report sim­i­lar to what’s shown below.

Now we can see what hap­pened when mar­ket­ing spend was pulled from the Colum­bus mar­ket on the 18^th of Feb­ru­ary; vis­i­ta­tion tanked! The geo-targeted mar­ket­ing had been dri­ving traf­fic, and when it stopped, there was a dis­cernible decrease in Vis­i­tors, roughly 100–200 per day fewer than dur­ing the cam­paign.
Proof pos­i­tive that geo-targeting works. And while you may have already seen how well Omni­ture prod­ucts like Test and Tar­get exe­cute geo-targeting, it’s not always as appar­ent for offline, spot-market cam­paigns, but now you see a way to get these data. I should note that by default, Site­Cat­a­lyst only reports vis­i­tors (which are the same as daily unique vis­i­tors) for each Geoseg­men­ta­tion cat­e­gory; if addi­tional met­rics (such as weekly or monthly unique vis­i­tors) are needed, a VISTA rule can be used to copy geo­graphic infor­ma­tion into props/eVars.

So other than geo-targeting through dis­play or con­tex­tual search, how else can you track regional pen­e­tra­tion this way? Just a few ideas include:

• White Mail/catalogs
• Spot cable buys
• Outdoor/out-of-home
• Mobile pro­mo­tions
• Local mag­a­zine buys
• Feeder mar­ket analysis
• Guer­rilla mar­ket­ing endeavors

So make sure you return your Cen­sus forms (it’s the law!), send me your unwanted Peeps (I pre­fer stale to fresh) and if you don’t want your marsh­mal­low treats, at least have the decency not to do this to them. Oh, the horror!