Putting The People Back Into Paid Search

Last week at ad:tech I presented a new approach to SEM called, "Putting the People Back in Paid Search." The thinking is that putting too much focus on the keyword instead of the person behind the keyword can prevent you from capturing the insights needed to fully optimize your paid search and other marketing campaigns. 

Power to the People

As Facebook marketplace ads gain popularity among search marketers for their ability to deliver high volumes of targeted exposure and traffic via real-time bidding, we're reminded that interests matter just as much as intent.  

With Facebook, ads that can be targeted to individuals based on their stated interests and creative units can (and must) be tailored to match what advertisers know about each person. Furthermore, bids can (and must) be based on the predicted value of engaging that person with one's brand. 

With paid search, ads are targeted to keywords, a subtle but important distinction. Keywords are not people. Rather, keywords represent the intent of a person at a moment in time. Nonetheless, marketers have become adept at tailoring search ads (and bids) to meet the anticipated need of each individual based on the keyword he/she is searching. 

To date, intent has beat out interest in terms of driving direct return on advertising spend. I spent a lot of time dissecting why this is the case in chapter 3 of my book  -- but, for now, suffice it to say that "mindset matters." However, relative to intent, interests represent more volume (after all, there are only SO many people out there searching for things that are relevant to your business) and opportunity for long-term brand engagement (after all, a search is fleeting).

Best of Both Worlds 

What if you could have your cake and eat it too? What if you could target paid search ads based on interests? And what if you could set bids based on the projected lifetime value of each customer or prospect?

Achieving this would require an understanding of the person behind the keyword -- his/her (and figuring out if it's a "him" or "her" is certainly part of the equation!) demographics, lifestyle characteristics, motivations, etc.

Enter Acxiom, a global marketing services company that has data on everything and everyone. Acxiom has developed PersonicX, a series of 70 audience clusters that roll up into 21 life stage segments. Based on the insight that life stages are proven predictors of consumer behavior, Acxiom has classified people into manageable groups more likely to respond to similar marketing messages.

An example of such a cluster is "Sitting Pretty." Acxiom has identified 19 million households that share these characteristics -- financially secure couples, on the verge of retirement, living in outer suburbs, and enjoying a luxurious lifestyle. These folks fall into the 45-65 age bracket and boast a net worth of $250-500,000.

For a company like Extra Space Storage (disclosure: Kenshoo customer), "Sitting Pretty" is a highly valuable audience segment, as these people are very likely to have more stuff and, therefore, a greater need of extra space for storage (hence the name, btw).

But how can Extra Space know if someone searching for its brand or a related term (which are often one and the same, btw... hence the name) is "Sitting Pretty" and thus representing a high lifetime value customer? If it were able to distinguish these searchers from others, surely, Extra Space would bid higher and tailor ads and landing pages accordingly.

Don't Call Me Shirley!

Indeed, that's just what Extra Space found itself able to do. After a thorough analysis of over 185,000 transactions, the Extra Space customer base was mapped to Acxiom PersonicX clusters, with up to 1,500 demographic elements appended per individual.

The next step for Extra Space was to match the clusters with specific paid search conversions and the keywords that triggered them. Now, Extra Space can understand the people behind the keywords and not just optimize campaigns based on demand metrics (clicks, conversions, leads, CPA) or even margin KPIs (profit, ROI, marginal ROI)  but actual lifetime value as projected at the time of the search query, based on the density of a given keyword to an audience cluster.

Putting this into play with paid search, Extra Space can determine if certain low-performing keywords represent high-value segments and optimize creative, landing pages, and bids to capture these conversions rather than just cutting them out of the campaigns completely.

Going one step further, Extra Space can determine which clusters perform best via paid search and look for other marketing channels to engage these people, such as display or even offline ads. Acxiom is integrated into comScore, allowing marketers to find web publishers with high compositions of various PersonicX clusters and prioritize them for display ads or sponsorships. 

It's Marketing, Stupid!

Every now and then we need a reminder that the M in SEM stands for marketing. As we geek out over line items in reports, it's easy to forget that the data we're looking at represents real people, not just keywords. And, when we're marketing to people, it's critical to consider everything we know about them as people when creating ads, on-site experiences, and assigning value to them based on their potential, as people, to become customers and brand loyalists.

Integrating Acxiom's PersonicX clusters into your paid search efforts is one of many ways to put the people back in paid search. To be sure, there are other companies such as Experian and Targus that provide the ability to map data to individuals or households and take action. And there are many emerging outlets that boast similarities to paid search (for example, Facebook marketplace ads and display ad exchanges) but rely on understanding people and their interests rather than keywords and their intent.

As Glenn Fishback put it during an interview for my book, "Sometimes, you have to step away from the spreadsheet." That's sound advice whether you're optimizing your marketing campaigns or your marketing career. 

Next story loading loading..