Remarketing For Search: Approaching The Promise Of Digital Marketing
Good fishermen know where the fish are… so why is it that good marketers don’t always know where the best consumers are? Despite massive advancements in targeting, tracking and user insights, John Wannamaker’s adage — “Half the money I spend on advertising is wasted; the trouble is I don’t know which half” — may still […]
Good fishermen know where the fish are… so why is it that good marketers don’t always know where the best consumers are?
Despite massive advancements in targeting, tracking and user insights, John Wannamaker’s adage — “Half the money I spend on advertising is wasted; the trouble is I don’t know which half” — may still be true. Even in the hyper-targeted and over-analyzed online marketing space, there are still many unknowns, but there’s light at the end of the tunnel — and it’s bright!
The best advertising is relevant, in the moment, and served to an individual who has interest in a product or service they have not purchased. Digital marketers use keywords, contextual relevance and statistical algorithms blended with a touch of intuition to spend ad budgets as efficiently as possible.
No matter what ad technology you’re using, the future of performance marketing is in the direct relationship with the consumer. And to own that direct relationship, you – the seller or purveyor of service – need to know exactly what the consumer wants and when they want it… possibly before the consumer knows.
That’s easier said than done. If challenged to deliver on that proclamation, I previously would have run a series of day part analyses and built correlation models between purchase behavior and query intent to identify trends and create a best-fit strategy.
Despite their imperfections, these strategies are time tested and work more often than not. The art within the science is minimizing the times it doesn’t work.
One major step in that direction is through building lists. Lists have been at the core of direct response marketing since direct mail and catalogs were invented and, more recently, for the purposes of remarketing based on consumer behavior. The latest evolution of lists comes from Remarketing Lists for Search (RLSA).
Google released the product in beta back in June 2012 and formally released it in June 2013. The idea is simple: similar to remarketing, you create lists of users who have executed certain actions (visited specific pages, signed up for emails, etc.) and remove users who have converted.
The difference is that rather than serve these lists of users ads until they purchase or for a flight of 30 or so days via display ads, we can now target these users though paid search (query/keyword based targeting).
This opens up a wealth of opportunities for efficiency and targeting in search:
- Increase impression share for users who are more qualified
- Show unique messaging for first-time viewers versus repeat viewers
- Tailor on-site experience for users that have already shown interest in certain products
- Stay top-of-mind for users in the consideration period
You get the point. For the first time in the paid search space, we can truly get creative by merging the known (what someone has done) with the correlation models (keyword/query-based intent) to back out revenue-per-visit at a user level.
But this is only the beginning. AdWords allows advertisers to pass custom variables into remarketing lists, which can be used to build more refined, custom lists. With a little elbow grease and creativity, we can imagine building out lists of qualifying users based on actions and further separating users by perceived value based on predictive analytics.
A user’s actions are worth a couple thousand impressions. We may be losing incremental levers of control through enhanced campaigns, privacy concerns, and the imminent loss of third-party, cookie-based tracking in search, but small features with big implications keep our heads up.
Mr. Wannamaker may have to rethink his quote — not only should we know which half of our budget is wasted, but we should know in real time the likelihood of a click being wasted.