It seems impossible to read a marketing blog that doesn’t bring up personalization, 1-to-1 conversations, cohorts, and segments. But what is “segmentation” really? Segmentation is arranging your customers into smaller groups to give them a more personalized experience with your brand. But how do you do this once your customer base breaches into the millions?
Despite all the recent buzz around hyper-segmentation, segmentation is not a new concept. Initially emerging to better serve buyer groups, segmentation has evolved past simply incorporating demographics like gender, age, geography, etc. Let’s take a look back at how one of the most iconic brands in history embraced segmentation.
1886 – Coca-Cola uses a single product to target a mass market
1950s – Coca-Cola expands to international markets and tailors the product ingredients to fit international palettes.
1982 – Coca-Cola begins to segment by gender, targeting women concerned with their sugar intake.
2005 – Coca-Cola expands segmentation and starts to target men concerned with their sugar intake. Note, the removal of “diet” and the darker branding.
2011 – Coca-Cola takes segmentation further by marketing the #ShareACoke campaign where consumers can personalize individual cans and bottles. An attempt at getting to each individual (hyper-segmentation).
Today – One of Coca-Cola’s more recent campaigns targeted around behavioral segmentation. (Not their best—just sayin’).
Hyper-personalization, not segmentation
Like Coca-Cola, many brands are segmenting their consumers into buckets. And truthfully, segments work today but consumer expectations are rapidly evolving. As more brands enter the ring to compete for a consumer’s attention, the brand that gets heard will be one staying ahead of what marketers are doing today. Some of today’s most iconic and famed brands like Amazon and Netflix have predicted a bucketless world and have reacted by moving from segments to hyper-personalization.
A great example of how Netflix is personalizing content and context is subtle, unintrusive, and valuable to the consumer. If you are one of the 93 million Netflix consumers, you know that there are tens of thousands of content options on Netflix but somehow all your favorites end up on your own personalized home screen. Netflix’s algorithmic recommendations ingest so many behavioral signals, they are able to curate content into a personalized homepage, just for you.
Context, on the other hand, is tougher. Context is basically brand messaging at the right time for the consumer on the right channel and on the right device. We’ve all been victim to a barrage of dismissable push notifications or 20 emails at the beginning of the day that get deleted immediately. For Netflix, context is partly built into their personalized homepage because the intent to watch is inherent (you wouldn’t be on their homepage if you weren’t looking for something to watch). For those who have the Netflix app and are die hard bingers, you’ve likely experienced tailored, useful messages from Netflix about a new season release for a favorite show that turn you into one happy consumer as well as a brand advocate.
This is the marketing dream—to have happy consumers who are brand advocates. So if the formula is simply personalizing content and context, why isn’t every marketer doing it? The reason is scale. Hyper-personalization is really really hard to scale.
Using artificial intelligence to personalize at scale
Let’s take a look at a few top industries in the B2C space. There is an overwhelming consumer base that is projected to increase in the coming years. From a marketing point of view, creating a hyper-personalized experience for each consumer is not only intimidating, but nearly impossible.
- In 2015, there were roughly 205 million digital shoppers with 2017 projections at 217 million online shoppers. (Statista)
- In 2016, 38% of U.S. consumers use the web, mobile, or apps to get news. That’s 121 million media readers and growing.(journalism.org)
- Since 2012, more than 148 million consumers booked travel through a combination of web and mobile. (statisticbrain)
- The 2014 Sports Media Report noted that of the 167 million sports fans, the majority are using desktop, mobile, or tablet to watch their teams. More so, roughly 7 million are using a second screen while watching live games. (sportsvideo.org)
Data ingestion powered by Artificial Intelligence allows for millions, actually—billions—of consumer paths to be created for a fluid cross-channel journey. Below is a unique use case made possible through data ingestion and AI decision-making.
Kelly, female, age 28
Kelly is an active Shoppal user on the web. Downloading the Shoppal app would increase her engagement. Understanding Kelly’s preferences allows the marketer to know to send Kelly an email during the weekday at 3pm, when she is actively checking her inbox.
Once Kelly downloads Shoppal, she is shown an in-app message welcoming her and offering tips to navigate the app. After 3 interactions, she is reminded with a personalized push notification to build her list of favorited items.
Knowing that Kelly prefers to make purchases on her laptop and not through the app, Shoppal shows her a browser notification Sunday evening when she has made purchases in the past. Being reminded about her favorite items, she adds to cart and checks out.
When you consider the journey listed above, you can identify and almost empathetically imagine yourself going through your own consumer path. Had this consumer been broken out by segments or audience buckets, they most certainly would have been lost at step 1 to 2.
Consumers are no longer passively interacting with brands and accepting what is handed to them—they are proactively seeking brands that offer them unique value that has been specifically crafted for them, fills them with value, and is relevant in the moment. This is a high bar to get over for marketers.
To avoid the mistake of bucketing your consumers into segments, and possibly missing out on a large number of opportunities, you must adapt content and context to each consumer in real time.