Personalization: Serving relevant content for the optimal customer experience
Updated: June 2, 2022
Personalization is a byword for digital experiences. Here’s a bit about the tools and tactics used to build tailored customer experiences.
What is personalization?
What is personalization?
Personalization enables businesses to build closer, more profitable relationships with their customers through empathy. The actual process of personalization involves tailoring something — like information, an offering, or an overall experience — to match the priorities and preferences of an individual customer or entire customer segment. In anticipating what these individuals want and need, businesses make customers feel valued, creating a more positive user experience.
In less clinical terms, it’s those social media advertisements featuring shirts that look similar to styles you’ve been browsing online elsewhere, the local concerts your music streaming software serves based on your location, or the articles a news outlet recommends reading next based on those you’ve already perused while sipping your morning coffee. It’s the well-informed act of putting something attractive to someone directly in front of them. Customer preferences might be explicitly communicated by a customer by a pick list or a quiz, for example. They may also be inferred based on characteristics, behaviors, and context, typically derived from past actions of other similar customers.
In the scope of digital, personalization is a multifaceted project. To be truly successful with this tactic, brands must consider not only what customers want, but what customer persona they fit into, where they are in the buyer’s journey, and what digital channels, devices, and touchpoints they frequent. By meeting customers where they are in each of these categories, brands are more likely to win over and retain customers. In short, it’s all about serving the right content to the right person at the right time through the right channel.
Technology research and consulting firm Gartner takes its definition of personalization one step further to encompass the value offered. It considers personalization to be “a process that creates a relevant, individualized interaction between two parties designed to enhance the experience of the recipient.”
What does good personalization look like?
Effective personalization goes beyond monogrammed mugs, name-bearing keychains, and email marketing with subject lines sporting the recipient’s first name. Personalization is not about simply knowing a customer’s name but helping them achieve their goals. The best personalization strategies remove the burden customers bear to contextualize them and identifies the necessary steps to achieve them.
Brick-and-mortar shop owners create personalized experiences by conversing with and observing customers. Through such interactions, they become familiar with the desires of individual shoppers and use that information to tailor real-time suggestions. Some may even go so far as to purchase inventory with a specific customer in mind and then showcase those hand-selected products to the customer the next time they come in. Whether or not the customer had an immediate need for that item, the shop owner's commitment to their satisfaction lays the foundation for customer loyalty.
Personalized digital experiences translate this philosophy across digital touchpoints like email, social advertisements, personalized microsites, landing pages, web banners, pop-ups, product recommendations, apps, chatbots, and even entire web pages. McKinsey predicts that select companies will attempt to take personalization a step further into the technological future, incorporating augmented reality. The management consulting firm predicts that in-person shopping experiences will be replaced with personalized virtual shopping experiences where customers can “try on” cosmetics and clothing curated specifically for them. Personalization that considers a consumer's emotion or state of health might also come into play as more and more customers use wearable devices and voice assistants.
Why are personalized experiences and content important?
Fulfills customer expectations
Personalization is one way companies show they value customers. When brands consider who a customer is, what they want, when they’re likely to make a purchase, and the channels they shop on, the customer feels seen, heard, and cared for. New research indicates that customers not only have a positive perception of personalization but that the majority are disappointed when it's not offered — especially younger consumers for whom consumption in every shape and form has always been largely digital. A recent McKinsey study found that, “71% of consumers expect companies to deliver personalized interactions. And 76% get frustrated when this doesn’t happen.” With brands experiencing a crowding of competition, customers have ample alternatives should the experience you serve fail to address them or their preferences.
Speeds up the customer journey
Personalization, especially in the ecommerce space, streamlines and speeds up how customers progress through the buyer’s journey — from awareness to consideration and, finally, decision-making. Through demographic information and past browsing and buying habits, companies can place the products and services customers are likely to be interested in within reach, removing the need to excessively search or navigate through the site. The brands in true alignment with their audiences might even be proactive with their recommendations, retailing knowledge of routine events, subscriptions that might need renewing, and more. For example, online gift and greeting card retailer Moonpig keeps track of which holidays customers purchase cards and gifts for, recommending the perfect present or card when that same holiday rolls around the following year.
Creates a stickier experience
What do Amazon, Netflix, and Spotify, all have in common? In addition to leading the pack in their prospective industries, these online marketplaces and streaming platforms are leaders in personalization. They’re favorites among shoppers and subscribers not just because they offer quality, variety, and reliability, but because they take the guesswork out of shopping, watching, and listening. When personalization is done right, customers can find more of what they love in less time. In customers eager to consume, personalization aids in retention and demotivates individuals from searching for alternative experiences, products, and services offered by competitors. Netflix’s findings support this notion, with the typical subscriber halting their search for something interesting to watch after roughly one minute. The majority of subscribers, about 80%, prefer to tune into what’s recommended instead. While personalization might require new tools and time spent on strategy, setup, testing, and iteration, brands should view these as investments in marketplace relevancy that make for more enticing, stickier brand experiences.
How do you deliver personalization?
With personalization aiding in customer satisfaction, conversion, and retention, the strategy is at the top of many to-do lists. But, before companies consider which type of personalization they’d like to implement, it’s worthwhile to address the following critical components of personalization:
- Which customer segments you’d like to target
- What content you are tailoring
- What ability you have to track its effectiveness
Understanding where a business stands in each of these areas can direct the type of personalization strategy it should implement and what technologies are needed to support it.
What are the different types of personalization?
When personalization entered the digital marketing scene, a single type existed: one-to-many or rules-based personalization. As purpose-built software became available, this type of personalization became easier to manage and a new type of personalization, algorithmic/machine learning, emerged. Below, we provide an overview of each approach, including its pros and cons. Then, we explore how the two are sometimes combined to create a third category of personalization, hybrid personalization.
Rule-based personalization takes an if/then approach to personalizing experiences. This type of conditional logic means that “if” a customer performs an action or meets certain criteria, “then” they will receive a specific type of content. In this approach to personalization, brands begin by manually dividing customers into well-defined segments. Then, they match those segments to specific products, content, or experiences based on the group they fit into or the attributes they possess. A/B testing plays an important role here as initial segments and conditions are based on assumptions that need to be continuously tested and iterated on.
Rules-based personalization can be as simple as bucketing customers into three categories, for example, new customers, returning customers, and past customers. Or, it can get more granular, with brands including a customer’s location, device, or other data points in the segmentation. The more granular the efforts are, the more time and testing will be needed to best tailor conditional statements. For this reason, the most effective personalization deals with simple variables.
Added values and drawbacks of rules-based personalization
|Easy to adjust content - Because segments and content are manually set up and linked, teams can track performance metrics and easily adjust content and segmentation to improve program optimization.||Difficult to add new customer segments - Each time customer personas change or expand, content and conditional statements must be manually added, adjusted, and assigned in alignment. Developers and content creators might struggle to keep up these tasks as growth and scaling occur.|
|Straightforward - With rules-based personalization, customer segments, the content being tailored, and the rules determining that tailoring are relatively clear. This clarity makes them easy to share with stakeholders.||Constricts customers - Not everyone fits neatly into a single customer segment — those driving the personalization effort or the software must decide which group is the closest fit. In these instances, the content a customer receives may do a poor job resonating with and converting them.|
|Affordable - This approach skips expensive software, meaning less financial investment up front. The trade-off is an investment in time. Rules-based personalization demands extensive critical thinking during setup and constant customer segment management over time — the financial implications of which can be seen in payroll.||Time-consuming to set up - This type of personalization requires planning and strategy that’s best informed by cross-functional stakeholders. Getting sales, marketing, engineering, and customer service teams together, and agreement can move personalization launch dates far into the future.|
Algorithmic/machine learning personalization
Algorithmic personalization, machine learning personalization, and predictive personalization all describe a similar approach to personalization, which relies on machine learning and data-driven algorithms to deliver data-informed content to customers. Personalization engines and data analytic tools work together to identify individual customer needs and then make real-time calculations to deliver the most appropriate content. This one-to-one personalization method relies heavily on automation, data, and technology.
There are several standard machine learning personalization models, which businesses can choose to implement – they may even decide to pair several in support of highly targeted experiences. Basic models use sequence detection to predict what customers might want or do next based on previous behaviors, patterns, or characteristic classifications that act as key personal identifiers. More advanced models track an array of signals to consistently "learn" about customers and serve tailored experiences that aim to attract attention now and then maintain it into the future, maximizing customer lifetime value. With these more advanced models, every action a customer takes, or doesn’t, informs how the personalization engine serves that individual and other customers in the future.
Added values and drawbacks of algorithmic personalization
|Advanced targeting - With access to diverse, high-quality data, experiences can be tailored to individual customers over large and finite customer segments, expanding the types of products, services, and experiences being served.||Expensive - While the ROI can be rich if done successfully (for example, Netflix reduces churn by one billion dollars annually with its personalized recommendations) the tools necessary for carrying out machine learning personalization can quickly eat up budgets.|
|Less maintenance, more scaling - Once personalization, analytic, and content tools are set up, little maintenance is required. Unlike rules-based personalization, algorithmic personalization recognizes, indexes, and tailors experiences for customers when they enter the site, with no manual segmentation needed.||Data reliant - With so much reliance on accurate, high volume, and heavily updated customer information, machine learning personalization is at the mercy of growing data and user privacy laws. Europe’s General Data Privacy Act and California’s Consumer Privacy Act are only a sample of what’s to come.|
|A/B testing isn’t winner takes all - When running A/B tests with algorithmic personalization, it's possible to assign variations to customer segments and individuals, rather than deploying the winning variation to everyone, tailoring the experience to the individual rather than the masses.||Potentially intrusive - Customers want to feel seen and heard — within reason and on their terms. With massive amounts of data to base one-of-a-kind experiences on, machine learning can sometimes cross the line between caring and creepy.|
Many of the drawbacks that come with rules-based personalization are offset by the benefits that an algorithmic approach offers, and vice versa. Hybrid personalization recognizes the balance that these two can create and combines certain aspects of them to deliver individually tailored content that also considers what customer segment someone is likely to fit into and if that should override any content being delivered via algorithms. With hybrid personalization, brands pair business rules with algorithmic personalization, which makes for more intentional, less invasive experiences.
This curated approach to personalization also lets companies build out personalization motions that might suit specific campaigns or business needs. As an example, Apple might use hybrid personalization to recommend its new iPhone to a customer that just bought a MacBook on a mobile device. This rule would override the company's go-to algorithmic recommendation which might suggest that the customer purchase something to go with their new computer, say a mouse or keyboard, based on sequential buying patterns seen in similar customers. In setting up this hybrid personalization rule, Apple can promote its new iPhone to more individuals temporarily without affecting algorithmic suggestions that exist elsewhere in the customer experience.
When launching personalization efforts for the first time, businesses may be unsure which approach to select. Smaller businesses might be financially positioned to implement a rules-based approach and then, over time, migrate to a hybrid or algorithmic approach as they’re able to make more sizable technology investments and have greater customer data to pull from.
What data is essential for personalization?
No matter which personalization approach a business pursues, data is a large contributor to success. Data is bucketed into four categories: demographic data (i.e., age, gender, education, income, interests, location, marital status, profession, etc.), behavioral data (i.e., web browsing, search history, social media profiles, device information, etc.), transactional data (i.e., purchase history, payment information, contact information), and attitudinal data (i.e., net promoter score, overall satisfaction, brand trust, customer loyalty, etc.).
The type of data one business collects and uses to inform personalization may differ from that of a company in another industry or even a competitor — what’s needed depends on objectives. Still, all data regardless of the type should be of high quality and moderate quantity. Companies should be judicious about the data they collect as all data needs to be managed in accordance with stringent data privacy laws and protected against breaches.
In the past, self-reported methods of data collection, such as interviews, surveys, questionnaires, and customer observations were the primary method for learning about customers. Today, most companies reserve interviews, surveys, and questionnaires for collecting attitudinal data. Demographic, behavioral, and transactional data are largely collected online via cookies and account tracking instead.
What tools or software are needed to create personalized digital experiences?
While the primary differentiator between rules-based and algorithmic/machine learning personalization is their reliance on manual vs. technology-driven customer segmentation, both approaches will require some combination of the following purpose-built tools.
Much of personalization relies on customer tracking to collect information and analytics tools to evaluate and identify patterns presented within that information. It is likely that companies looking to embark on personalization already use one or several analytics tools to organize data and measure success. In such cases, it might not be necessary to introduce another tool, but rather, to reconfigure the one already in use to accommodate additional data points.
Customer data platform
A customer data platform (CDP) captures partially or entirely structured customer data from diverse sources to build dynamic customer profiles. While some CDPs segment customers or come with machine learning capabilities to predict their next moves, such capabilities alone are not enough to support personalization. Customer relationship management systems(CRMs) may also be used, but because they contain just general information — contact information, location, transactions — they are best suited to rules-based personalization projects.
Because personalized experiences are made up of variations of image, text, video, product information, and other assets, this initiative requires companies to manage a vast library of content. A content platform stores content and structures it for reuse and distribution across any channel or device. Traditional content management systems (CMS) organize content in page format, meaning personalization requires teams to build out one-off pages. This demands much copying, pasting, and editing from content teams. A modern content platform like Contentful breaks content into reusable components that can be configured and reconfigured based on who a customer is, the device they're accessing that content on, and so on. In this way, content is more easily versioned for various customers.
Personalization engines, or personalization software — which may be stand-alone, suite-based, or an entire platform — unite customer segments, data, and content to build out unique customer profiles which inform the messaging a customer is served after taking a specific action. As Gartner puts it, “Personalization engines apply context about individual customers and their circumstances to select, tailor and deliver messaging such as content, offers and other interactions through digital channels in support of three use cases — marketing, digital commerce, and customer experience.”
How companies create personalized experiences with Contentful
Contentful’s content platform stores and structures content in a way that allows it to be accessed and distributed across any digital touchpoint. It also integrates seamlessly with personalization engines and other technologies companies use to tailor their customer experiences. Here’s a look at how two companies create personalized experiences with Contentful.
The primary goal of many businesses — office supply retailer Staples Canada included — is to maximize profits, which depends largely on lead gen and conversion. Through personalization and rigorous A/B testing in Contentful, Staples found it could optimize these efforts. Since re-platforming to Contentful and Shopify in 2019, the brand has built out personalized homepage banners in the hopes of engaging and retaining more customers. Instead of integrating a personalization engine, Staples utilized Contentful’s content platform and data analytics tools to segment customers and implement rules-based personalization. Within the business user segment, Staples Canada’s tailored banners saw a 51% increase in click-through rates.
The decision of which new tools to adopt in support of personalization will rely on two things: a company’s industry and the tools it already uses. When Contentful customer, direct-to-consumer pet food retailer, Pets Deli embarked on its personalization journey, it combined the power of long-loved ecommerce platform Shopify, with Contentful, Ninetailed, and Google Analytics. With these four tools, the brand was able to test and identify what pricing options converted the most customers. In eliminating antiquated discount codes and implementing personalized pricing, Pets Deli’s conversion rates doubled and bounce rates decreased by 10%. Pets Deli credit Ninetailed with maintaining site speed and performance throughout these efforts. The brand experienced no delays in site speed or changes in performance as is common with dynamic content changes.