Is Real-Time Customer Experience (CX) Too Slow?

Recently we’ve been exploring the concept of real-time CX. We’ve examined whether real-time CX is crazy or commendable, as well as whether real-time CX is really real. We’ve also discussed the necessity of looking beyond real time to CX for the long haul. To close out our series, today we’ll consider whether real-time CX is actually too slow.

You might be thinking, “If real time means immediate, how can it possibly be too slow?” Well, the reason is because real-time CX means that you are reacting to some type of request from the customer. You may be reacting quickly — indeed, almost instantaneously — but you’re still being reactive instead of proactive.

While a fast response is important to the majority of today’s customers, speed alone is not enough. Customers want companies to go past reacting to anticipating their needs, which brings us to predictive CX.

What Is Predictive CX?

According to renowned CX expert Shep Hyken, predictive CX means “that the customer gets the support not only before they know they need it, but before the problem ever occurs — because it’s predicted that it will occur.”

Hyken points to the office copy machine as a classic example. Most modern copiers have built-in sensors to determine when they begin running low on toner or paper and then automatically notify users by way of sounds or flashing lights. This provides users the opportunity to restock necessary supplies before the toner or paper have completely run out, averting a potential crisis.

Image credit: Shep Hyken

Another common occurrence of predictive CX is maintenance alerts. The prototypical example from the brick-and-mortar world is an auto repair shop telling customers to bring in their cars for an oil change or other tune-ups when a certain period of time has passed or number of miles has been driven.

But this type of predictive CX can also be found in digital goods, especially Software as a Service (SaaS) products. For example, a file storage service might send customers an email when they start approaching their storage limit. Or, a SaaS company might display a banner at the top of their website to notify users that the site will be down for a few hours in the near future for scheduled updates.

How to Implement Predictive CX

We’ve shared a few examples of predictive CX above. But how can companies implement predictions to enhance the customer experience other than low-supply sensors and maintenance alerts?

Use Data to Find Patterns

The very best tool that companies have at their disposal for predictive CX is data. For most organizations, this is no problem. In fact, they’ve got dala galore, from biographical information about customers and reams of support tickets to logs documenting software bugs and maps of future product updates.

The challenge, then, is not in collecting data but in accessing and organizing the data that already exists in order to make use of it. Whether the work is done by man or machine — such as by using emerging technologies like artificial intelligence and machine learning — the way to make use of enterprise data is by identifying patterns within it.

For instance, companies can examine support tickets to look for trends around specific issues. If a product update was recently pushed out and now front-line agents are receiving multiple complaints about a particular feature not working, they can most likely assume there was a bug in the update. The support team can then use that information to proactively notify other customers that the company has identified a bug and is working to fix it, thereby preventing those customers from running into the same problem themselves.

Another way in which patterns can be used for predictive CX is through personalization. Today’s support teams typically have customer relationship management (CRM) tools that let agents quickly determine who a customer is, which products they’ve purchased, what support issues they’ve encountered previously, etc.

“A well-trained and experienced customer service agent will not only solve the customer’s current issue, but also anticipate their next question or problem,” writes Hyken. “All of this is based on this customer’s previous behaviors, which align with patterns established by hundreds or even thousands of other customers.”

But personalization is not solely the domain of customer support; it can also play a key role in pre-sales CX. Specifically, marketing and sales teams can use data to make predictions about online shopping behavior.

Perhaps there’s a clear pattern that shows customers who buy one type of makeup (e.g., liquid foundation) from an online retailer are likely to also buy a second type of makeup (e.g., blush). The marketing and sales departments can harness this information to make proactive recommendations to customers. For example, after a customer purchases foundation, they can send a follow-up email recommending a few different blush products.

Solicit Customer Feedback, But Take It with a Grain of Salt

Besides using data to find patterns, companies can improve their predictive CX programs by regularly soliciting customer feedback.

As Proof co-founder and CTO John Phillip Morgan explains, “Talking to customers is probably the most important thing I do all week. While there’s a level of product development that requires you to take a stance and anticipate needs — it’s impossible to do that without an understanding of your customer’s current situation.”

Proof is a startup, but it’s clear that the value of customer feedback is not limited to early-stage companies nor to those in technology roles. Case in point: SurveyMonkey CMO Leela Srinivasan said in a recent interview, “We ran some research last year and found that 63 percent of people think that marketers are selling them things that they do not need. That tells me that we as a marketing profession are not doing a good enough job of listening, of really understanding the pain points, the challenges, the opportunities for us to add value to our customers.”

Feedback can come in the form of surveys to measure common CX metrics like NPS, CES, and CSAT, in-depth questionnaires, or even customer advisory councils. But no matter which method you choose, always remember to take the customer feedback with a grain of salt.

As Srinivasan herself acknowledged, “Most brands hear from just one percent of their customers … [which] can be very, very dangerous. To get truly comprehensive feedback that represents the views of your overall user base, you want to hear old customers, new customers, free customers, VIP customers, and everyone in between.”

In addition to ensuring that you listen to a representative sample of all your customers, it’s critical to listen to the internal experience and vision that exists within your company. There’s a famous quote from Henry Ford that sums up this point: “If I had asked people what they wanted, they would have said faster horses.”

Ford is not saying to ignore the customer. Instead, he’s illustrating that customers are almost always better able to explain their problems than to prescribe solutions. In Ford’s case, customers understood and were able to clearly articulate their need to go faster. In the end, however, the best way to solve that problem was through a solution that customers could never have asked for because they could never have imagined it. That’s where Ford’s vision and his company’s ingenuity came together to predict — and then bring to life — the product that customers ultimately needed.

From Predictive to Seamless CX

In this post, we’ve explored why companies should endeavor to anticipate their customers’ needs along with some potential ways for how to do so. In our next series, we’ll move from predictive CX to the concept of seamless CX and consider how companies can solve customer issues and help them achieve their goals in a manner that’s as frictionless as possible. Stay tuned!

Page Title: Is Real-Time Customer Experience (CX) Too Slow?
Meta: Anticipating your customers’ needs is the name of the game, but is real-time CX too slow?
URL: /real-time-customer-experience-slow