jeep driving on a red rock
PHOTO: NeONBRAND

After several years of businesses chanting the personalization mantra, the state of the market is ... mixed.

This conclusion came from research my company, Magnolia, conducted via in-depth interviews with close to 100 organizations from a range of industries, including finance, retail, travel, media and public services. We found that truly well-done personalization is ramping up in some organizations but is still far beyond the reach of others, while an elite group of digitally mature organizations have exceptionally active and advanced personalization initiatives.

Marketers are clearly recognizing the personalization road is bumpier and windier than they had expected. What are some trends emerging in personalization, and where do we go from here?

Manual Personalization Reaches Its Limits

One thing is certain: manual personalization doesn’t scale. This is not really big news, but many have seen the manual route as a good first step to get started with personalization.

By “manual personalization,” I mean the planning and serving of content to people in various market segments based on what we know about users. Regardless of whether we talk about implicit or explicit personalization approaches, this method requires marketers to manually define what content should be served to which segment.

Why does this fall flat? For one thing, it’s tricky to collect information about users. Research compiled by WebHostingBuzz showed 86 percent of users reported being bothered by having to create new accounts and 77 percent of users believed a good alternative was to log in to sites via social media. However, users seem hesitant to adopt the social login option: Email management service MailChimp, for example, found just 3.4 percent of its users signed in via social login.

Yes, manual personalization has its limits. One of our research interviewees, a European retailer, deals with so many content variants and visitor segments that it’s no longer possible for human editors to do targeting and match content to visitors. So it implemented weak artificial intelligence (AI) system that picks the right content items for visitors from a content pool.

Related Article: Personalization Isn't the Goal, Conversion Is

Marketers Turn to Tagging Content Assets and Using AI

There are two sides to this approach. On one side, you must be in control of a large pool of content with plenty of well-structured, reusable content assets. On the other side, your AI services gather data implicitly on user behaviors such as click paths or page views. You then marry the two. Leaving it to the machine to bring together visitor, content and experience carries a risk of delivering somewhat odd experiences with sometimes unforeseen results. If you’re looking for a sleek catsuit for your next Halloween party, you don’t want teasers for cat food and so forth.

Personalization Pushes Us to Become Better at Content Production

Look behind the processes and you’ll see content cannot easily be used everywhere. Our research showed creating channel-specific adaptations of a base story is slowly becoming the norm. That means marketers produce a foundational piece of content, which then gets adapted for channel or platform idiosyncrasies. It is possible to produce truly reusable content, and that certainly has been done successfully from time to time — but only by a few.

Another respondent, a fashion retailer, put it this way: “We currently create content for individual channels and not really for omnichannel. What we want is efficient content production that is truly omnichannel, for use in any country, channel or device.”

Some organizations we meet are starting to train their editorial staffers differently — to write with different purposes in mind, to reuse much more. They treat their marketing channels as ongoing products and train their channel marketers to grab the pieces best suited to assemble different experiences.

The content problem keeps a number of businesses away from personalization, at least for the time being. The common cry goes like this: “We struggle with running just one website. We need more content before we can think about personalized content. We’ll focus on localization first and get all our content translated. Our tools for serving personalized content are too difficult to use.”

Related Article: Refine Your Personalization Efforts By Ditching Your Tech-First Tendencies 

Explicit Personalization in the Travel Business

On the other end of the spectrum, some are chasing the holy grail of one-to-one personalization. They know how they want to personalize and are keen to smooth the customer journey. We’re even seeing companies establishing and maintaining specific yearly personalization road maps. Companies in the travel industry are among the leaders, and their initiatives offer lessons on what could make for effective personalization.

Tap Into a Large Number of Logged-in Users

Compared with companies in other sectors, companies in the travel industry have had more success getting users to log in — e.g. to buy tickets. Travel is also a sector with strong loyalty programs, where there are incentives to engage users and build up profiles and preferences over time.

A Sabre report showed about 79 percent of business travelers rated login features as valuable, in particular if the tool knows who they are, what company they work for, their loyalty status and their preferences. Travel sites are able to dig into a treasure trove of data, especially when they work with social logins and third-party payment providers.

Related Article: How AI Will Usher In a New Age of Personalized Commerce

Define Logical, Natural Rules for Personalization

Travel lends itself to the creation of rules (platinum customers get this piece of content or that special offer, for example). And customer journey mapping tends to be more obvious in the travel industry. For example, after booking their flights, travelers usually select their seats and meals, and then they can easily be nudged to look for the supplementary components of their trip (hotels or rental cars, for example) before proceeding to payment.

According to Kissmetrics, JetBlue set up trigger emails around those customer touchpoints based on two factors: the destination city and the customer’s loyalty program status. It found that, compared with standard promotional emails, trigger emails had higher open rates and conversions, and thus generated more revenue.

Spot Unique Opportunities for Personalizing the Customer Experience

In another example from the travel industry, a cruise operator was looking for ways to use wearable devices and internet of things (IoT) technologies to enrich passengers’ journeys during their time on board ships. This effort included displaying content unique to the ship the passengers are on, or sending timely alerts based on where people happen to be (“There’s a quiz game starting on the deck you’re on”).

Related Article: The Value Exchange Equation Is Shifting

Comply With Privacy and Data Protection Laws

Even in a dream industry for personalization like travel, there are still significant challenges. One is meeting legal requirements and determining whether it is possible to abide by privacy and data protection laws if you’re using individualized, but not individually identifiable, data.

One airline company cited the need to adhere to regulations in dealing with personally identifiable information, given it could potentially be doing a lot with the amount of data it collected. As is true of so many other companies, the airline’s investments at the moment go into stitching data together in ways that are coherent but not personally identifiable.

The airline said the most difficult part is tracking the same individual across platforms. It is exploring different ways of doing that and is striving for the 360-degree customer view. Needless to say, this is a complex task and demands solid integrations between different tools and systems, from customer relationship management to analytics.

Still Offer Personalized Service, In Spite of Problems

Looking through our research on personalization, a common theme is the number of variables is too high to manage properly, and therefore, things frankly get messed up on the touchpoint experience. Customers may find information about the currency or the departure airport is wrong. Or they may have to re-input credit card details. Or a system may forget an address. The customer experience is spoiled when things like that go wrong, and businesses across the board are naturally striving to spot such errors before complaints flood in.

Bumpy Road With Starry Lights

The road to effective personalization is rocky. Most businesses are struggling with basic personalization, which isn’t a bad start at all. Anything that gives customers more relevant experiences, tailored to their preferences and needs, is better than broadcasted generic messages trying to reach everyone. 

The common notion that personalization isn’t actually happening simply is not true. The starry lights in our research shine very bright.