Where vs. Who: The Battle for Internet Marketing Dominance
Published: March 12, 2012
Author: David Rodnitzky
Search engine marketing works well because it is so accurate at telling marketers the intent of potential customers. When someone types “best mortgage rates” into Google, marketers have a very good idea of what that person wants, and thus what ad to serve them, what landing page to show them, and how much they should pay to target that customer. To put it another way, the query is a marker on a map of customer intent. Marketers can discern the stage of the buying cycle, the product that is wanted, price-sensitivity, geography, and many other factors based on the words a customer uses on a search engine.
We can also determine some information about “who” the customer is, but it is generally quite limited. Someone who types in “New Maserati” is probably wealthy, and someone who types in “The Structure of Scientific Revolutions” is probably a science nerd, but we don’t know if the person looking for a Maserati is also interested in Italian cinema, likes to collect matchbox cars, and recently got a job on Wall Street.
The “who” is much more available through Facebook than through Google. Since its inception, Facebook users have readily (and honestly) provided tons of “who” data about themselves; age, gender, marital status, sexual orientation, interests, and geography are all easily mineable by marketers. The problem with marketing on Facebook is that there are much fewer “where” markers that tell us that someone wants to buy a specific product or service. For example, someone who says they are “engaged” on Facebook may not be looking for a wedding photographer, now or ever; someone who went to the University of Iowa may not be looking for work in LA, or looking for work at all:
Like Peanut Butter and Chocolate
As you might imagine at this point, the “holy grail” of online marketing is the ability to understand both the “who” and the “where.” Knowing that someone searched for “New Maserati” on Google, and also knowing that they work at Facebook and went to graduate school, versus someone who did the same search and is 18, unemployed, and likes matchbox cars can obviously make a huge difference on your bid (or whether you bid at all).
This merging of data already starting to happen in display advertising. Using behavioral targeting and DSPs, you can find inventory on ad exchanges for a particular type of Web site or article (the “where”) and also learn some behavioral data about the user who is visiting that site (the “who”). You can do the same on Google Display Network (GDN), by overlaying placement or semantic targeting with interest-based behavioral targeting.
On Google proper or on Facebook proper, however, you are currently left with the choice of just ‘where’ (Google) or ‘who’ (Facebook). This is about to change. Google’s strategy to integrate ‘who’ into search is Google+. As TechCrunch author John Constine noted, whether you use Google+ is irrelevant to Google:
Google scrambled to build Google+ because it watched Facebook and saw users were willing to volunteer biographical data to their social network, and that data is crucial to serving accurate ads users want to click. Search keywords and algorithmic analysis of your Gmail and other content weren’t enough. It had to start the journey to identity after shortsighted years of allowing users to sign up without asking who they really were.
Facebook’s approach will be to allow advertisers to access the “Open Graph”, enabling advertisers (eventually) to mine data on Facebook users as they interact across the entire Internet.
The open graph doesn’t exactly give you the same “where” data that Google does, in that people only “like” or “read” or “friend” someone, but it certainly gives advertisers a lot more information than the static demographic/psychographic information that can be gleaned from a user’s Facebook profile.
My prediction is that Facebook will take this one step further, allowing advertisers to combine Facebook profiles, Open Graph data, and status-update semantic matching. Thus, if you are 40, work at an investment bank, read an article on CNN about a new Maserati, and post an update on Facebook that said “I’m thinking about buying a new Maserati,” Facebook advertisers could use all of these data points to place a highly targeted ad.
This is an exciting time to be an online advertiser and a bit of a scary time if you hope to have any privacy at all when you surf online. The battle to know everything about everyone online is raging!
– David Rodnitzky, CEO