Will Marketers Finally Give Up On Personalization?

Mastering the practice of personalization can seem overwhelming for marketers, leaving them to wonder whether the cost and effort is worth the trouble. But giving up on the dream of one-to-one marketing would be a shame.

By: Stephen Shaw
Read time is 8 minutes



When the era of Big Data dawned in the early 2000s with the explosive growth in web traffic, marketers swooned at the potential for personalized messaging: at long last, the dream of one-to-one marketing seemed within reach.

Yet as the years rolled by personalization remained a buzzword: the practice never caught up with the rhetoric. In a McKinsey survey last year, just 15% of CMOs claimed they were “on the right track with personalization”, stymied by a chronic lack of resources. Meanwhile, customers have come to view personalization as a parlour trick, urging them to add one more item to their online shopping cart.

In recent years, however, a new generation of AI-powered technologies has made it easier than ever for marketers to personalize the customer experience. Which is why in 2019 the Association of National Advertisers named personalization its “Marketing Word of the Year”. But then at the end of last year the research firm Gartner came out with the glum prediction that five years from now 80% of marketers will have abandoned their personalization efforts due to a lack of ROI or the “perils of customer data management”.

Personalization is hard to get right – even with the latest technology. But the reasons go far beyond the technical complexities of mastering personalization. It has more to do with marketers treating personalization as an afterthought. Viewed as a basic customer expectation, personalization becomes too important for marketers to give up on.

Being Better

Look over the shoulder of any digital native absorbed in their phone these days and it is easy to see why marketers are struggling to connect with them. Headphones on, eyes glued to their screen, they live in their own personal media bubble, frenetically scrolling, tapping, texting and swiping, messaging their friends, checking their newsfeeds, sharing their pictures, maybe even watching a video or playing a game.

Jamming sponsored posts in between their social feeds or stuffing e-mail in-boxes with “best guess” offers or blitzing them with push notifications just forces them behind a privacy wall. The same is true of every other invasive tactic – web site pop-ups, infobars, digital ads, video pre-rolls, the list goes on. So people have learned to screen out the street noise – skip and block ads – “cut the cord” – insulate themselves from blatant commercial pitches – and seek sanctuary on ad-free media platforms.

As long as marketers persist in making campaigns the basis of their planning, their brand messaging will never be able to compete against the constant barrage of media distractions. Instead, they should find ways to be useful in the moment, whenever personalized assistance is needed, based on individual need and circumstances. And that means shifting the marketing mindset from “What can I sell you today? to “How can I help you?”. As Seth Godin says, “Marketing is driven by better. Better service, better community, better outcomes”.

Personalization should never be considered synonymous with targeting. It should not be limited to “you may also like” recommendations. Instead, it should be woven into the fabric of the total customer experience, helping to make it “better”. Better at delivering advice. Better at answering questions. Better at making the experience seamless across devices and touchpoints. Better at delivering timely service and support. Better at anticipating customer needs.

Personalization should never be considered synonymous with targeting.

Take the example of the calorie-counting mobile app LoseIt!. It stands out from the galaxy of weight watching brands by offering real-time analytical reporting such as pattern detection. For example, the app will figure out how specific food items influence the amount of weight gain or loss: “We’ve noticed that on days you incorporate muffins, you tend to keep your total calories lower”. The reason LoseIt has grown to 30 million members is its commitment to being “better”. Every new feature is born out of a singular crusade: to fight the obesity epidemic by making people more conscious of their food choices.

The biggest barrier to personalization is that so few companies take a holistic view of the customer experience. Most are still organized around product and channel silos. And while customer journey design can make it easier for people to interact online, a different planning approach is needed to individualize the experience, making it more relevant and useful. Marketers have to give up transactional thinking (“How do we get a user to click and convert?”) in favour of asking, “How do we make the experience better for the user?”.

A Unified Experience

The first step in the process is to map out a “value matrix” which defines how the company will apply personalization in the service of customers. That matrix can then be used to prioritize the opportunities based on expected outcomes (such as higher satisfaction and retention).

For example, different treatment strategies can be applied to first-time versus repeat customers; frequent versus occasional users; single versus cross-category buyers; passive members versus brand advocates; as well as to different loyalty segments, lifestyle cohorts, demographic groups, and so on. The tricky part is coming up with the right set of nested segments (e.g. high value + long tenure + highly engaged + brand loyalist) deserving of their own personas. Rules can then be created to customize the experience for each segment, knowing their shared attitudes, beliefs, interests and habits.

At an individual level, the strategy should revolve around specific moments in time, when a timely intervention – say, a recommendation, an alert, a status update, a reminder, a newsfeed – can make the customer journey friction-free. This is where personalization technology can play a huge time-saving role, relieving marketers of the heavy analytical load involved in generating the conditional rules (e.g., “if-then-else” statements, “and/or” logic). The toughest part, of course, is to ensure a unified experience, applying a consistent set of rules, regardless of the device or touchpoint.

Once the personalization strategy is set, marketers can move on to cross-channel planning, figuring out how to deliver the best possible experience based on everything they know about customers – their observed behaviour, stated preferences and implicit intentions derived from their web sessions (“How much time did they spend looking at that product page?”).

The website is usually the best place to start by showcasing the most relevant products or solutions according to the browsing history of repeat visitors and customers. That can be done by setting up dynamic blocks of content offering individualized product recommendations; swapping out generic text and images in the “hero carousel” to suit each persona type; streamlining the navigation to accommodate a particular segment; promoting relevant content based on reading history; and finetuning search results according to known interests.

By factoring in both shopping behaviour and historical buying habits, e-mail content can be fully personalized, informing subscribers of their loyalty status and privileges; introducing ancillary products related to a recent purchase; providing advance notice of local sales events; announcing new merchandise in their preferred categories; or letting them know how soon to expect a resolution to their service ticket submission. Similarly, a mobile app can use location data to deliver geofenced notifications; support in-store navigation; or use bar code scanning to provide on-the-spot product comparisons and ratings.

Zero Latency

None of this is easy to do without a single unified customer profile – one central place to store all of the data – plus the ability to act on it in real-time. But as Gartner observed, customer data management is usually a black hole for most companies due to systems fragmentation. To make personalization something customers truly value, the individual-level data must be instantly accessible, otherwise, the appropriateness and timing of the real-time response is compromised.

Too often, when a customer visits a brand web site, opens a follow-up e-mail, uses the mobile app, reaches out to the call centre, subscribes to a newsletter, or clicks through to a landing page, each interaction is captured in a separate database, often without an identifier flagging that individual as the same person. In that all-too-common scenario, personalization is limited to each channel, failing to take into account all of the other interactions, even though one may have led directly to the other. The exact stage of the buying journey is unknown (are they researching, evaluating, ready to decide?) – multiple buying signals get lost – true intentions go unnoticed. This failure to piece together the full picture also causes unnecessary redundancy and frustration (“You’re asking me to give you that information again?”).

Identity resolution addresses this blind spot by linking an individual with both their terrestrial and digital identities, associating, for example, a household address with an e-mail, IP address, and one or more device IDs. An identity graph stores all of the possible identifiers in a single database. That way, a persistent customer identifier can be used to link interactions across channels, helping to connect, for example, a web site visit to a recent call centre inquiry. The graph is constructed with the help of a specialty vendor (like LiveRamp) which uses matching formulas – both deterministic (absolutely positive) and probabilistic (highly likely) – to come up with a universal master key.

The other massive barrier to overcome is designing a data architecture which supports zero latency (go back to that digital native with their manic thumbing). It requires converting the “big data” running through the digital bloodstream of the company into actionable prompts like instant call-outs, auto-notifications and triggered alerts. But the response cycle time of most companies today is still measured in hours, never mind milliseconds, a relic of an earlier era when daily or weekly batch updating was the usual practice.

“Fast-twitch” Ecosystems

Today companies at the leading edge of personalization have developed “fast-twitch” data management ecosystems that ingest streaming data in real-time at the moment of interaction. Rather than rely on traditional relational data stores, data warehouses, data lakes or CRM systems, they have opted for Customer Data Platforms to skirt the delays and bottlenecks that can occur pulling data from different “systems of record”. A CDP is a “master data store” which assembles unified customer profiles by pulling together every bit of available data from multiple sources using a single API, including active web session events like “checkout in progress” or “order paid for”, and makes all of it available on-demand to “systems of engagement”. For example, a web site visit by an addressable customer, tracked using a JavaScript plug-in, will set off a sub-second chain of commands that lead to a contextualized response as the web session is in progress.

Today companies at the leading edge of personalization have developed “fast-twitch” data management ecosystems

The job of message generation is typically controlled by a standalone personalization engine where the engagement rules are stored and activated in real-time. Based on the customer data passed through by the CDP (e.g., CID, segment code, event flag, last order, etc.), the associated personalization rules are invoked, instructing the web content management system to serve up a specific page or block of content. To make rule development scalable, machine-learning algorithms are used which comb the data to find groups of people similar in their choices and behaviour; select offers with the best chance of converting based on known affinities; and recommend content based on viewing history. The data can be as granular as time spent looking at a specific page or longer- than usual hovers over content, actions that might reveal possible preferences or interests.

As companies grapple with the growing customer demand for more individualized attention, marketers will have no choice but to master the practice of personalization. Some marketers may find it too hard and give up, as Gartner says. But now that one-to-one marketing is finally technically feasible, that would mean giving up the opportunity to “be better” at meeting the expectations of customers.

tn


Stephen Shaw is the chief strategy officer of Kenna, a marketing solutions provider specializing in delivering more unified customer experiences. He is also the host of a monthly podcast called Customer First Thinking. Stephen can be reached via e-mail at sshaw@kenna.