Digital interactions are the backbone of people’s relationships with many organizations. Whether it’s through a website, mobile app, social channel, or company portal, these channels are often the primary way people now engage with companies.

Compared to other interaction channels, digital provides a highly flexible and efficient medium for capturing experience data (X-data), making it hugely beneficial to an organization’s Experience Management (XM) efforts.

Five Types of Digital X-Data Collection Mechanisms

How should organizations think about instrumenting their digital interactions to collect this valuable X-data? Start by understanding five of the key digital collection mechanisms:

1. Active, Behavior-Based Listening Posts: An active feedback request that is triggered when predefined criteria are met (such as time on page or cart abandonment) to capture information at a particular moment in time
2. Passive, Always-on Listening Posts: A persistent feedback channel that is present on every page where visitors can volunteer to share about their experiences
3. Embedded Feedback: A request for feedback embedded directly within a page (thumbs up/down, 5 stars, Yes/No, etc.) 
4. Cross-Channel Initiation: An option at the end of a survey for respondents to provide their email address and receive a follow-up communication from support
5. In-app Notifications: A brief feedback request specifically designed for capturing real-time insights during in-app experiences

1) Active, Behavior-Based Listening Posts

Organizations can capture people’s perceptions of their digital experiences in real time by sending them feedback requests based on their current behavior and activities, like when they abandon their cart or visit a specific page on an internal company portal. This type of active feedback request (sometimes known as an intercept) is a valuable research tool for collecting core XM metrics – like satisfaction, ease of use, task completion, or likelihood to return – as well as key drivers like navigation, content effectiveness, speed, and visual appeal. It also offers organizations the opportunity to collect demographic and psychographic information that may reveal meaningful trends between different segments.

However, due to the active nature of this type of active X-data collection mechanisms, they can seriously disrupt the visitor’s experience if presented too frequently. Organizations must therefore use them sparingly and keep feedback requests succinct – never asking for information they already have. Avoid deploying these requests on the first page of a visitor’s digital experience, on important landing pages (like the homepage), or on transactional pages (like a checkout page). Instead, best practices for these active feedback requests include presenting them only when someone has spent enough time on the site or app to provide meaningful feedback, escalating from discrete to indiscrete requests based on response rates, and using targeting logic to prevent repeated display.

2) Passive, Always-on Listening Posts

Another listening post organizations can deploy to provide visitors with an easy, two-way line of communication is always-on listening. Unlike behavior-based listening, always-on listening – also known as passive listening or persistent feedback – is less obtrusive, and can be deployed on most pages and devices, and requires minimal logic or sampling. It is particularly valuable for diagnosing specific pain points and reporting bugs. However, because visitors often use always-on listening to vent their frustrations, organizations need to recognize that data collected through this post will skew negative compared to other channels, and it is therefore not the best venue for asking research questions to capture core metrics like satisfaction, ease of use, etc.

This listening post often takes the form of a “Feedback” button displayed along the side of the page which, when clicked, reveals a survey with questions asking about the nature of the visitor’s feedback – such as website feedback, reporting an issue, or contacting customer service.  The touchpoint typically also includes an open-ended text field for additional clarification, and some brands decide to include an option for visitors to provide their phone number or email address so as to continue the conversation in another channel.

3) Embedded Feedback

Embedded feedback provides a happy middle ground between passive and active requests for feedback on digital channels. This mechanism embeds the feedback request directly within the page, which allows visitors to answer questions without deviating from their given path. This type of feedback request is particularly useful for understanding the effectiveness of a specific piece of content. Content owners have historically had to rely on rudimentary mechanisms like traffic, search rank, or time on page to understand – or, more accurately, guess – how effective their content might be. This embedded feedback mechanism not only allows content owners to move beyond these legacy metrics, it also generates prescriptive insights into what type of content is driving brand consideration and usefulness while simultaneously surfacing page-level trends.

Although organizations often use embedded feedback to evaluate content effectiveness – usually by asking something like, “Was this content useful? Yes or No,” – this mechanism can also be deployed to ask a slightly longer list of questions that dig into a particular aspect of a digital journey. For example, embedded feedback can be added to an order summary page to ask customers to rate their purchase journey or added to a sign out page to ask employees to assess their account management experience after logging out. Regardless of where they are embedded, it’s important to make sure these feedback requests match the look and feel of the rest of the page, yet are not so well integrated that visitors are liable to overlook them.

4) Cross-Channel Initiation 

Sometimes a survey isn’t capable of collecting the full extent of a visitor’s thoughts and feelings about their experience. That’s where cross-channel initiation comes in. This mechanism allows people who feel that their feedback wasn’t adequately captured by another listening post to share more about their experience with someone on the support team. So while not every respondent will take advantage of this listening post, cross-channel initiation can be an enormously valuable tool in instances where visitor feedback requires more time to fully explain or when closing the loop is crucial to maintaining the relationship.

Often this type of collection mechanism is used at the end of a passive, always-on feedback button. The survey will end with a question such as, “Would you like someone from our team to contact you about this feedback?” and if the respondent selects “yes,” they are asked for their first name and email. This triggers the creation of a ticket on the backend, which then will be routed to the appropriate support agent, who will then follow up with the respondent to understand their feedback or concerns.

5) In-App Notifications

Organizations typically shy away from an intercept-based approach to soliciting feedback through mobile devices as screen sizes are small and no one likes to have some aggressive popover take over their phone. However, in-app collection mechanisms have advanced in recent years, and there are now a few options that are especially effective at requesting feedback in a polite, unobtrusive way.

The first – an in-app notification – is a small creative type that is particularly valuable for catching users as they exit the app. This type of notification looks very similar to other reminders people see on their smartphones and is useful for capturing real-time abandonment scenarios, which can be tricky on mobile devices as people tend to exit apps quickly and immediately move on to the next thing. The second in-app collection mechanism discretely rises up from the bottom of the screen and gathers real-time insights into people’s needs and preferences. Organizations can then use this information to improve app usability and onboarding, identify which features are working and which aren’t, and keep a pulse on their brand reputation before issues show up in App Store reviews.

Recommendations for Deploying X-Data Collection Mechanisms

To make the X-data you capture through these digital collection mechanisms as useful as possible, we suggest you:

• Limit survey length. Longer surveys negatively impact completion rates regardless of the medium. In a transactional environment – like a website or app – this effect is even more dramatic. So keep the number of questions you ask limited. We suggest a maximum of 10 questions for overall site experience, 2-4 for journey-specific research, and fewer than 5 questions on in-app surveys.   
• Develop a sampling strategy. Unfortunately, only a small percentage of visitors will volunteer feedback through these collection mechanisms (about 2-3% of visitors click on active intercepts and 0.2-0.3% click on passive intercepts, of which only about 50% complete the survey). So to ensure you are gathering enough X-data to generate reliable and valid results – an especially important consideration for active in-journey and overall site experience research – you will need a strategy for collecting feedback from a representative sample of your visitors. This requires identifying how many unique users visit your site or undertake a specific journey and then calculating the correct sample rate. To ensure the right populations are providing feedback, you can use contextual or embedded data – such as time spent on site – to adjust which visitors see which collection mechanism.
• Make requests appealing. People are more likely to respond to your feedback requests if they both look attractive and if the messaging in them sounds compelling. So, for example, make sure that the visual appearance and tone of all collection mechanisms match your brand and feel like a natural part of your site. And when you’re soliciting feedback, phrase your invitation in terms of the number of questions you’re asking rather than the time you expect it to take. So rather than saying, “Do you have two minutes to provide us with feedback?” instead say, “Please answer these three questions to help us improve your site experience.”
• Focus on understanding specific journeys. People don’t just interact with organizations through digital channels for the sake of it – they are inevitably trying to achieve some higher-level goal, like gathering information, managing their account, getting support, or making a purchase. So rather than asking questions about individual interactions or moments, ask targeted questions during specific journeys aimed at understanding visitors’ perceptions of their entire journey. This in-journey research will allow you to optimize the site or app to help people accomplish their ultimate objectives as efficiently and painlessly as possible.
• Combine X-data with operational data (O-data). Another reason digital channels offer such a rich environment for producing XM insights is that they automatically generate contextual and embedded data through background analytics. To make the X-data you collect as actionable and predictive as possible, you should combine it with this automatically generated O-data. Examples of relevant O-data here include browsing data like page views, search terms, and time spent as well as personal information like demographic data, account management, and annual spend.

The bottom line:  Digital X-data is critical, so collect it strategically.

Juliana Smith Holterhaus, Ph.D., is a Senior Product XM Scientist with Qualtrics, specializing in Digital XM

Isabelle Zdatny, XMP, CCXP, is an XM Catalyst with the Qualtrics XM Institute