About a month or so back, I wrote a fairly popular essay on customer experience and the central role that empathy plays. I also said these platforms (and any martech stack) have the potential to help transform customer experience, but that many of the tools enabled by these suites are often aligned to a superficial understanding what we think customer experience is.
But I did promise some more content around what DXP platforms or your martech stack can do to help, rather than hinder the process.
Forrester has an excellent podcast called “What it means”, which has been hitting it out of the park lately. Coincidentally, I would suggest this latest episode is particularly relevant to the topic.
“[Forrester] Vice President and Group Director Melissa Parrish explains on this week’s What It Means. Marketers tend to believe that consumers want more precise targeting and more personalization — when, in fact, many prefer some level of anonymity. Consumers may see highly targeted or personalized messages as intrusive or even distressing, such as when targeted ads presume information about health status.”
The goals or marketing are usually not the goals of your customers
Hopefully you listned to the full podcast, because it’s filled with excellent and relevant examples of personalization gone awry. At its core, it resonates with what I said about empathy and using that as the guiding principle for customer experience (long-term goal), not customer acquisition (short-term outcome at the potential expense of long-term goals).
It’s important to remember that DXP tools can be used for good and bad, so it’s important to have some guiding brand and experience principles in place, particularly where overall customer experience and marketing/sales goals may be in conflict.
Look closely at the KPIs you have and determine if they are properly aligned to CX goals.
We constantly hear of KPIs gone awry; such as attempts to reduce call time (leading to agents dropping customers within the middle of problem solving) or in the case of Wells Fargo, agents fraudulently signing up customers for fake accounts en masse (leading to over a $ Billion in fines). Organizations should be auditing customer interactions and KPIs in order to ensure there are no unintended consequences to CX as a result of misplaced incentives.
Resist the temptation to use dark patterns
A “dark pattern” is a usability function which is designed to fool the user into doing something they did not intend. Some examples from Harry Brignull include :
- Trick questions
- Hidden costs
- Privacy Zuckering (making options too complex to understand)
There is a sub-type which I would call a needy or “nagging pattern” (or “confirm-shaming”) – which is designed to either guilt or annoy a user into clicking “okay”. This is my own personal annoyance lately as it’s not just a sneaky practice, but often belittles your user with dialog boxes stating things like “No, I don’t want to save money, because I am clearly a moron”.
Fortunately DXP and WCM systems in general do not actively encourage dark patterns and tend to push towards clarity of purpose and compliance for various modules such as forms.
Recognize that all content is marketing content
Often, when in the discovery phase, potential customers will spend a great deal of time on what are usually considered “post-sale” activities such as evaluating the quality of documentation, support and community.
A proper content strategy should take into account the entirety of customer experience across all systems they may interact with – not just marketing content in the “top-of-funnel” activities. The “content hub” concept starts to address these requirements in various different ways – either coordinating these functions and content across systems, or acting as the central source of truth, unifying content and experience across channels and applications.
Privacy is not just for compliance – it is a competitive differentiator
The great thing is that most DXP platforms provide very clear mechanisms and functionality for ensuring European GDPR compliance. Some examples:
- Sitecore and GDPR
- Adobe GDPR resource site
- Episerver – Solutions and responsibilities for GDPR compliance
- Acquia and GDPR compliance
You should consider leveraging the GDPR features in your platform regardless of geography – it’s easier (both technically and in compliance – especially for multi-national sites) but also provides a clear indication of your privacy principles. And while customers may not have the time to fiddle in the labyrinth of privacy settings of Facebook, at a high level they will choose to spend at vendors who make this a hallmark of their brand ethos (such as Apple and their strategy of differentiation against Android).
The similarly structured California CCPA will be a forcing function for most sites anyway, coming into effect shortly (Jan 1, 2020).
Among WCM vendors, Crownpeak has also been making privacy a central plank of both their product strategy and messaging.
Identifying drop-off points (but look to the larger context)
Always be aware of the bigger picture – just because a button has a higher click-through rate, it doesn’t necessarily mean to translates to better experience.
There was once a website that had a button which read “Instant Demo” – but when you clicked on that, you got a form which made you fill out some details, at which point a salesperson would get back to you……. sometime…. later. Not so “instant”, but the sales director saw much higher conversion rates compared to “Schedule a demo”, so the wording stayed. I never knew if the overall brand experience suffered because of this disconnect between what was promised and delivered (my assumption is yes), but this was mostly because we didn’t seek to find out.
In my mind, once trust starts to erode (even on these little things) it’s much harder to rebuild.
As the Nielsen Norman group states:
Measuring the live impact of design changes on key business metrics is valuable, but often creates a focus on short-term improvements. This near-term view neglects bigger issues that only qualitative studies can find.
The goal of personalization should not be to push more to a client, but to de-prioritize irrelevant content and streamline experiences.
Often we hear of simple scenarios for personalization – to either explicitly (i.e. based on known information such as a user profile) or implicitly (usually through content tagging analysis to determine interest) using that information to make suggestions for related content or personalizing the hero image on the home page based on persona.
For example, if an insurance company has various products, commercial, residential, etc. – by looking at browsing activity, you can put a happy picture of a family playing soccer on the hero image, instead of an executive touring a new factory build. The idea (of marketers, anyway) is that you are prioritizing imagery which your persona can relate to, in order to provide a more personal connection.
But rarely do we see personalization used to actually alter the user experience to make user goals easier to accomplish, which should be the real goal. Rather than focusing on emotive imagery, it would make more sense to transform the experience to reduce the navigation elements and content unrelated to the user (especially for organizations spanning multiple disparate personas such as Finance or Healthcare) – you want to reduce the clicks it takes for users to accomplish their top tasks.
Look at the data you already have – chances are you don’t need more of it (or supposed miracle tools like AI), you just need to be smarter with what you already have.
Here’s a great example of these previous principles in action. I am a big fan of Strava – I use it regularly to track my rides and other activities. When I am training actively, I even pay extra for some of their offerings, such as heart rate zone tracking.
Now, let’s look at their home page (above the fold):
Clearly you can see that a massive portion of the screen real estate is devoted to signing up a new user, through three different methods, no less (Facebook, Google and email).
The annoying thing for me, is that as an existing user, I have two tiny “Log In” prompts. And Strava already knows that I am a frequent user, by virtue of the fact that I have a cookie on my machine. So for my “active user” persona, the amount of screen real estate devoted to sign up vs. log-in should be reversed entirely. I already know that Strava is the #1 app and have it on my phone and watch.
A DXP should be able to determine at the time that we are building the rules for personalization, what percentage of users will see one variation vs. another. Using something like “previous visitor” as the profile, you would easily know what percentage of users are repeat visitors or not in advance of making those changes. (A really good one should be suggesting this type of action preemptively, but I’m not sure any platform is quite there yet, despite the promises).
Similarly, if you did put this rule in place and saw a drop in sign-ups, ideally you would be able to know at a glance the cause and effect of various changes (and rollback quickly). If this usability change was responsible for a drop in new accounts, you would be able to know easily (at which point you can balance the needs of marketing and customer acquisition against overall brand and known user experience) – but a DXP should allow rapid decision-making and testing of these scenarios (assuming you take a wider view).
In addition to cookies and existing web browsing data, there are also other linked systems such as CRM and support which provide very clear indications of customer intent and issues.
More data is not necessarily better (continued)
DMP data in a web experience is particularly hard to trust as it relies on the categorization of other content creators outside your control (for example, as I live in Quebec, due to my geography being conflated with my language by ad networks, I get many ads and associations with French content, despite being an anglophone).
As a test, if you use Twitter, visit this link: https://twitter.com/settings/your_twitter_data/partner_interests
It will take you to a page which lists every inferred interest based on your browsing behavior. In my case, the majority of this data is simply not true or useful in any way – including tagging me for multiple age ranges from 25 to 65+ and older, children I don’t have, and a proximity indication for stores that literally do not exist in my country of residence.
The same experiment can be done on many other DMP platforms which can reveal what others have tagged you as – Oracle Data Cloud: https://datacloudoptout.oracle.com/registry