November 2018
Text by Alberto Ferreira


Alberto Ferreira is a user experience researcher and globalization services expert with over ten years of hands-on experience in UX on the client and agency side with some of the biggest companies in the world, including Sony, BBC, and Mars. He is the author of Universal UX: Building Multicultural Experience, and writes and speaks frequently on topics ranging from Agile to persuasion design.



One size does not fit all: Personalizing content and e-commerce experiences

Personalization is a key word in the modern consumer market. Presenting users with information, services or products that might be interesting to them at a particular moment and in a particular place can greatly increase your revenue. And you might be surprised how even small initiatives can have a big impact.

The age of the "mass market" as a consumerist concept is over. The old absolute has given way to the rise of the personal, the relative, and the intimate. The 19th and 20th centuries saw the rise of a middle class and the breaking down of old barriers to social emancipation. Old prejudices and misconceptions were slowly broken down, and as a result, women and minorities became vital cornerstones of the economic zeitgeist. By the 1950s, children and teenagers entered the consumerist mainstream. Later, the expansion of mobile technology allowed billions of people to become "potential customers" for global brands.

But, for all of its social impact, the concept of a "mass market" was lost along the way. The idea of “one product for all” was discredited when consumers became so varied, international, multicultural, and differentiated by their beliefs and buying power. This became an intrinsic part of the new wave of design and communication: Today, we know that a "well-designed" product will never befit the variety of preferences, applications, use cases, and contexts it will be exposed to. From pumpkin-spice latte drinking millennials to Gen-Xers, vloggers, vegans, and carnists: We live in fractured societies where each group represents an established market with its own targeted needs. As writer Chris Anderson (The Long Tail) stated, "our economy and culture are shifting from mass markets to millions of niches". 

However, the story doesn’t end here. There is a deeper phenomenon on the rise, guided by digital technology. The looming challenge ahead of us is not about targeting a group, but targeting the individual. This is the next logical step: Once a brand knows enough about you, it can show you – and only you – what you need and want. This phenomenon is deemed personalization, and its implications are profound for e-commerce and technical communication alike.


The new wave of personalization

The brand experience has become fluid and continuous, and several companies already have a business model based on anticipating users' needs and requirements using their own data. This is the model that enables content delivery services like Spotify to produce a "Discover Weekly" bespoke playlist based on the songs you listened to. Other services like YouTube and Netflix analyze what users watch and make recommendations based on content similarity and popularity. The comedies you watch, the cooking shows you follow almost religiously – all of this data is used to profile you against the wide variety of the service's user base.

On a basic level, information on users can be collected using session IDs and assorted browser information. Even when users are not logged into a website, each action they take on the website is tracked and identified uniquely. As long as users do not clear their cookies or start a new session from a different IP, analytics behind the website will be able to identify who accessed the website and what they did during the session, and leverage this data to present suitable content and recommendations.

There are three fundamental types of personalization: 

  • Explicit: Personalization based on user settings and preferences, which is generally used to customize a website or service to the specific user preference. This is the most common form of personalization, and is sometimes called customization. Common applications include app and website settings set by the user, like profile preferences, content visibility, and preferred products.
  • Implicit: Personalization based on an app’s or website’s user data to guide the user to products, content, and other elements.
  • Hybrid: Personalization based on a mix of elements such as time, position, and IP address to understand the use case. Geotargeting is the simplest form of gathering information, as it can automatically redirect users to the right regional websites when accessing your platform, or change the language automatically based on browser and operating system settings. Geotargeting can also be used for recommendation engines. Airbnb and FourSquare use geotargeting to present activity guides and recommend places to stay based on nearby areas. 

Some services can leverage a combination of or all of these types of personalization into one combined solution. 

Figure 1: Different personalization scenarios can be applied to your customer experience. Strategize your methods according to business value and user satisfaction.



Personalization tools

Personalization is usually dependent on a dedicated technical infrastructure, which means it can be costly and it carries an implementation effort. Most tools available on the market are based on site behavior and user profiling, depending on the type of personalization required. Some of the tools available are complex hybrids between analytics, CRM, and delivery systems. These tools support the ability to deploy A/B testing in addition to email, mobile, or in-page survey tools. 

Such tools include Evergage, Monetate, and Adobe Target, among others. These platforms can be embedded in your CMS or sit on top of your front-end delivery layer. For websites and apps alike, these tools should be able to add or change elements in pages and screens depending on how the user behaves or his profile. They can focus on specific scenarios, such as cart abandonment or CX optimization, or be more general.

CRM features are also common as they can sometimes also manage communication elements such as e-mails and notifications. A case in point is an email that is sent to a user who did not complete a purchase. If one reminder email is not enough to entice the user, triggers such as offering a discount can be added.

Other personalization engines such as Qubit and Dynamic Yield tend to have a marketing side, based on conversion. However, the most interesting aspects of personalization deal with the ability to offer the user the most relevant or useful product or proposition. 

In technical communication this implies delivering just-in-time documentation, adjusted to the user’s needs. Some of the current solutions include a combination of topic-based technical writing and structured content. Combining these with actual user-centered content can greatly improve the overall user experience.

Promoting relevance

Despite their promise, personalization and customization are not universal yet, even as devices like Alexa and Google Home start to pervade every single nuance of our life. But you don't need to develop a talking orb or acquire an expensive software solution in order to dive head first into personalization. 

The key principles in personalization are relevance and empathy. According to a 2013 Janrain study, up to 74 percent of users are frustrated when they are confronted with irrelevant content. In the years since this study, the attention span has shortened further and the web has become an even more convoluted and distracting place.

Therefore, if you are showing different content to different audiences, this content should be as particular to the audience as possible. Despite the fact that personalization is an extensive concept, some of the possible applications tend to fall into one of the following categories:

Recommendations and sorting: According to a McKinsey study, 35 percent of Amazon’s revenue is generated by its recommendation engine. This clearly illustrates the benefits of having a personalized set of recommendations. But it is not just about the bottom line. Users feel more engaged and more taken care of when digital touchpoints provide relevant choices during their interactions. These are usually algorithm-based and thus require a hefty infrastructure. Implementation patterns include:


  • Suggest related content based on similarity and popularity. Keep in mind relevance, as most users will not be happy to get the same recommendations over and over again.
  • Propose new products that are shopped for by other users with a similar profile.
  • Automatically set sorting preferences.
  • Suggest items for repurchase based on previous shopping behavior.

Content: Content filtering has been around for years, from DITA's structured content approach to role-based content. Ideally, for documentation and technical communication, a structured approach can help to distinguish between audiences and therefore deliver the appropriate content to the right user. With the help of these content tools and a thorough understanding of the segments' attributes and values, you can use different content to establish a relationship. Simple tricks include using the user's first name in an email subject or addressing an action after a CTA call has been triggered in a playful manner.

Getting closer to the customer

There are a lot of statistics floating around on the Internet regarding personalization, and we should study them with an appropriate amount of skepticism. Any personalization effort starts with segmentation, and skipping this is sure to wreak havoc on the entire project. Start by asking questions:Do you have different audiences that you are trying to reach? Do they respond differently to different content? If you are working in e-commerce, which areas are most critical to personalize first and which could be of greatest use to the company as a whole?

There are a number of steps that you can take to address this as an integrated project before even beginning to determine how to differentiate experiences offered to different users. 

  1. Start by identifying the value and feasibility of personalization. Look at analytics data across your various channels. Identify the key areas where the user can be served different content. Is your CMS even capable of serving differing content? A simple way of analyzing this is by small A/B tests where you experiment with differences in the content served. Think about personalization in view of the entire journey, not just one or two pages. Collect any hypotheses and questions that may arise in this process. 
  2. Strategize your cross-channel approach. For example, Android phones natively provide recommendations to their users for improved commutes, nearby restaurants, and other offers. iOS users, on the other hand, rarely open these, so a different approach is needed.
  3. Understand the customer and his actions. 
    a. Establish user scenarios across those multiple segments where personalization would have a real impact. Research is very valuable here. Hold interviews with users, send out a survey to the customer base, use in-page surveys to ask about specific behaviors while the user is going through different stages. 
    b. Cross-check the high-level findings with your own priorities. For example, a scenario could be “A field engineer with an entry-level profile is logging into the knowledge base.” From the analytics on most often checked articles, and the results of your own research, you should be able to determine what type of content would be most useful in this case.
  4. Understand the user journey to obtain a specific outcome. Analyze user expectations carefully in each stage.
  5. Map these scenarios to actual content in each journey. Think about the information architecture of your site. How you present something is as important as what you present. 
  6. Once you understand the key targets for receiving personalized content, you need to map the type of content they should receive. Will you show a warning to one segment and a cross-sell opportunity to another? Personalized content can be as granular as one short string or a full content panel of text. Target users based on their expectations as much as on their needs. Personalized content is not just about selling, but supporting the user’s journey with actual worthwhile content.
  7. Remember to only specify content for audiences that contains actual value. It’s easy to start building micro-content with 20 different versions of the same string. This will not help you in the long run and, most importantly, it will not help you to focus on the scenarios that will best serve the needs of the user. Small changes in tone will never be as impactful as an overall content strategy.
  8. Experiment and measure iteratively. Conversion is not the only valid metric. Engagement, time on page, and even satisfaction ratings can help you to track how effective your changes are. Adjust your metrics accordingly.

Personalization is not a straightforward process and it demands a hefty investment in both planning and overall effort. But when even a simple type of personalization like "visitors who viewed this also viewed" can deliver up to 68 percent extra revenue (according to Smart Insights), it is hard to dispute its importance.

Regarding investments in email personalization, there is evidence that dynamic content, incorporating contextual information like weather or nearby locations, is attractive to users. Similarly, personalized product recommendations in key sections of the user journey have a huge impact on the uptake. Don’t stop at addressing your users by their names. Leverage their needs to present them with things they would be interested in regardless of context. Even pages based on static templates may have room for minor changes in CTAs and informational text with genuine and measurable impact. Regardless of the technical infrastructure, and the extent of your personalization strategy, everything starts and ends with the user. Demonstrate value in exchange for the users’ personal information and always promote tangible benefits, and the long journey to the purchase will become much more pleasant – and engaging.

Figure 2: Personalization can be embedded in every single aspect of the business, Metrics from user research are vital to finding the best approach.


Are there risks?

We live in an era of repercussions and reverberations. Overfiltering is a risk for users who might never get exposed to anything outside their comfort zone. Twitter is one such example of hyper-personalization, where users are continually exposed to content that they at one point considered relevant and interesting. It is the responsibility of brands and technical communicators to leverage personalization in a way that augments the user experience, rather than limiting it artificially. Personalization doesn't only build loyalty and re-inforce the relationship between brand and user, it can also ensure that a digital touchpoint becomes recognizable, approachable, and relatable.

Content managers and communicators have long been tasked with understanding users and their needs. As personalization pushes interaction and content to the fore, the technical communicator becomes a key part of this new wave of digital experience. And in these first few tentative steps of connecting big data, user experience, and technical communication, we are merely hinting at the potential of what is to come.