Personalization Strategies Are Driving More Effective Ads, leveraging data to deliver a highly relevant ad to the right person at the right time
I watch a lot of hockey. Which means I have to watch a lot of ads for cars and pickup trucks. Historically, some advertisers throw tons of cash to target demographics watching sporting events. Namely for hockey, TV is a medium with a wide reach that can attract Males 25-39.
However, surely this is a ridiculously broad category - I’m not in the market for a Chevy, and have seen the ‘Real People’ campaign so much, I can probably recite all the lines (as an aside, these spoofs made me feel a bit better about life)
Digital Advertising provides a platform to get highly specific and cater ads that are specifically relevant to me. But clearly it is easier said than done.
What is Marketing Personalization, Exactly?
Harvard Business Review was talking about One-to-One Marketing way back in 1999. Here they define it as “being willing and able to change your behavior toward an individual customer based on what the customer tells you and what else you know about that customer.”
Elaborated in the book, The One to One Future, the concept describes managing customers rather than products, differentiating customers not just products, measuring share of customer not share of market, and developing economies of scope rather than economies of scale.
Insert in Box
Today, companies are far more data rich than 20 years ago. Given that we can collect, organize and derive far more meaning from data, we can now deploy marketing strategies where we leverage data to deliver individualized messages and product offerings specific to our customers’ needs. The popularity of Account Based Marketing is an extension of this thinking.
Instead of broadly lumping in customers into buckets like Male, 25-39, we can create segments that organize based on demographic, contextual, and behavioral data. Instead of making our best guess about classify segments around pretty arbitrary features like gender and age, we can now leverage machine learning to understand trends that perhaps weren’t so obvious.
Growth of Marketing Personalization
Brand owners recognize the potential. Most marketers worldwide claim to have some level of personalization strategy in place.
According to a recent Monetate study, almost three-quarters of retailers in North America have listed “Personalization” as their #1 priority for 2017. But many of those brands also indicate it could take up to three years to implement their personalization strategies.
And it seems those who are willing to put in the effort are seeing results. Linus Gregoriadis, U.K. research director at Econsultancy claimed that ‘Businesses implementing personalized marketing have seen an average increase of 19% in total sales’.
Daniel Faggella runs CLVboost, a boutique email marketing consultancy in Massachusetts surveyed over 50 machine learning marketing experts and found that while most executives voted “Search” as the AI marketing tool with the highest profit potential today, “Recommendation and Personalization” topped the list for ROI potential in the coming five years.
The More Segments, the Merrier
Lotame, a leading Data Management Platform, argues that coming up with a handful of customer segments is over simplifying things and recommends building up to 100 different segments to experience with.
"Data management is facilitating an explosion in the extrapolation of the media planning capability. Properly diversifying a data portfolio means evolving from media plans built on a single piece of creative split across five line items, and moving into an environment with hundreds of line items.”
"Those segments cannot only be leveraged for testing and rapid refinement but can also be aligned with hundreds of pieces of creative to get to the true one-to-one communications...”
Segment based on Behaviour
Segmentation can also occur through consistent patterns on how consumers are behaving. Being able to cluster segments based on consumers’ habits, can help advertisers understand which campaigns and activities are driving results across a range of different groupings. This ideally will lead to more productive campaigns, saving a huge amount of wastage that occurs with overly broad segments.
Large companies are getting into the act too. DMP Rocket Fuel incorporates IBM’s Watson into their predictive marketing platform ‘to better understand the content where ads will land’. This is likely to establish some sort of consumer intent. This type of information is often referred to as ‘Thick Data’, defined by Teradata as qualitative information that provides insights into the everyday emotional lives of consumers.
In fact, according to Yahoo, the Receptivity of Emotions study, even reaching customers in the right mood could make digital ads 40% more effective. Understanding when the timing is appropriate to show an ad for a mutual fund as opposed to a bag of chips is incredibly valuable. Especially in the context of B2B sales where approaching someone with a more complex offering is likely to be much more successful when the person is in ‘work-mode’.
Further, leveraging this ‘Big Data’ can really illuminate consumer behaviour and provide insights into why consumers behave in a particular fashion, what preferences they may have, and ultimately identify why certain trends stick.
Challenges with Marketing Personalization
But, as with most marketing technologies, matching the lofty promise with actual results is keeping a lot of executives busy.
The solutions are only as good as the data they’re built off. Ad fraud has been discussed ad nauseum, but if you have questionable data integrity, you may be reluctant to trust that the trends established in the many segments are not being unduly influenced by bad data. One hopes that with increased volume, accuracy can be ratcheted up, but easier said than done.
And the industry needs to be sensitive not to inspire a consumer backlash. Personalization done badly can be ineffective in the least, progressing to annoying or worse disturbing. Lisa Lacy from The Drum put together a great post entitled ‘5 Tips to Personalize Your Marketing (Without Looking Like a Creep)’
Making broad sweeping generalizations about people based on a few attributes (whether it be age, race or gender) is usually a pretty terrible idea, but we constantly do it in segmentation. Never say never, but I don't see a Chevy Silverado in my future, despite seeing those ads each commercial break during the hockey game. Using data properly, and employing personalization strategies that better tailor the right information, to the right person, at the right time can help engage a target audience who is already tired of too many terrible ads.