Using Predictive Analytics to Boost CX for Good
With the Internet of Things, SaaS-based products, and the app-ification of our lives, data gathering has become easy and entirely non-intrusive to end users.
In today’s competitive market, customer experience (CX) is a differentiator.
PWC found that 73% of people list customer experience as an important factor in buying decisions, yet the same report found that 54% of U.S. consumers say customer experience needs improvement at most companies. This discrepancy is significant because often CX is treated too simplistically. Companies frequently perceive CX as being polite on the phone, accepting returns, or giving good discounts.
Sure, those may be elements of a good CX, but they focus only on a portion of the entire experience.
The most impactful element of customer experience is about proactively solving customer pain points, to the extent where a user may not even know a problem occurred. Predicting what customers want to do, or what they expect to happen, and then helping make that happen as smoothly as possible is a far better approach than retroactively asking them to explain their problem and attempting to solve it.
To identify these pain points, you can’t rely on one team or a single data set. You need to use analytics from all the tools across your company to form predictive analytics models of your customers.
Predictive analytics is the result of a process of combining information from historical data from a variety of sources to help you detect patterns and trends which can be used to create desirable outcomes or prevent future failure.
- Sales needs to understand what causes lost deals and the triggers for upsells
- Support should identify what causes the most tickets and work to prevent them
- Product can focus on problem areas of the offering and re-design UI
- Marketing needs to detail what mediums for ads work best to generate the most leads
Predictive analytics can help you create the smoothest possible experience from first communication to ongoing maintenance of your product. When done correctly, it’ll improve your CX, generate more loyalty and referrals, and grow your business.
Setting your Customers up for Success
What makes your customers successful? You sold a customer the value that your application provides, but how exactly does it provide it? Predictive analytics can help you determine what patterns enable your customers to be more successful.
There’s a reason why many YouTube creators publish new videos on the same days. Analytics have shown videos perform better when posted at certain times. YouTube publishes those insights so that their customers have more success, and therefore generate more revenue for themselves and YouTube.
Similar metrics exist everywhere within your customer base. If you learn from your analytics that customers have the most success with your product if it is configured in a certain way, make that the default, guide the customer through those configurations, or at a minimum publish them as best practices.
Your historical observations can help make future interactions better for your customers. Predictive analytics can also help you proactively let a customer know they may not have success.
Retailers know that if you add something to a cart, and leave the site, you are unlikely to return to purchase it.
Therefore, many businesses recognize that pattern and send you a reminder, or a discount coupon to entice you to complete your purchase. They get the sale, and (maybe) you save a little money.
These patterns are waiting to be found in data sets such as previous product usage, churn analysis, and support ticket tagging. By analyzing what made previous customers successful, or not successful, you can help new customers reduce their time to value with your products.
Set Yourself up for Success
Similarly, as with setting up your customers up for success, you also need to ensure your business is set up for success. Using data collected from your case management system, product usage metrics, and defect tracking system you can start to build models of which customers are at risk of churn.
Customer Engagement Management (CEM) tools, often used in Customer Success teams, help you to predict disengaged customers or those who are not seeing the value. If a customer is not driving the appropriate value, the risk of churn rises. The insights from these tools can alert you before renewal conversations start so that you can start to get them back on the right track before the situation escalates.
Most of these tools have rules that can automatically trigger actions. To set these rules, analyze your previously churned customers to see what they have in common.
Maybe in the months prior to them leaving your product, you see a rise in the number of support tickets, a decrease in logins or actions taken within your product, or a certain threshold of unfixed defects.
Combining those factors and feeding the data into your CEM tool you can trigger a warning to the success manager or account team to proactively assist the customer before they reach the tipping point of not renewing.
These touch points show that you value your customers, and therefore improve customer experience. Analytics about how customers use your product and services are invaluable to help you retain customers.
Preventing Support Tickets
Product manuals no longer exist. Their demise is related to products becoming more intuitive and user experience becoming a focus for companies. It also represents a shift in solving problems when they inevitably arise. As the Technical Services Industry Association found, self-help and providing that help in real time are important trends for support. Predictive analytics can help you catch where errors are and guide the user on how to resolve the problem.
Guiding users means knowing where errors are likely to occur. For example, international companies know that something as simple as a tooltip showing the expected format of a date entry field, or phone number is more desirable than throwing an error after the user has typed in a seemingly valid entry.
Another way of guiding users is to recognize product areas that result in the most tickets within support. Using tagging and other metadata about the cause of a ticket, you will find patterns within your case management system. These become predictable areas of struggle that could easily result in support tickets and customer frustration.
Create a process to feed this information into your product team or use products such as WalkMe to guide users through these problem areas in real-time inside your application.
Let customers know that something is about to happen
A more advanced way to use predictive analytics is to prepare support to solve a case.
If you know a customer will need assistance when a particular set of steps or error occurs, usually a catastrophic one, automatically log a ticket with the relevant information, or store logs/auditing, etc. in a place that support can access it when the customer inevitably calls.
Remove the customer effort to gather details by prepping your support team as best you can to solve the ticket. Another way is to analyze patterns and send customers a notice when they may run into trouble.
For example, if you see a customer approaching a threshold or limitation of your product, let them know in advance as opposed to waiting for them to hit the error and cause more frustration.
Anything you can do to proactively resolve a customer ticket, by providing self-help documents, guides, or saving a customer the time and effort of searching for an answer or opening a ticket is a sure way to build loyalty and improve the overall experience.
Predictions Versus Reality
Products often have multiple ways to accomplish the same goals. You may predict that customers follow an expected flow through the product, but analytics show they behave differently.
In the simplest case, this change could be a minor time saver.
For example, you may have a selection box that the default is set to what you predicted to be the most common selection, but in reality, your analytics show a high percentage of users change to another option. Using this data, you should change the default in your product and save the majority of users that added click.
Taking this idea one step further, you may notice that certain selections are consistently the same based on some other criteria, such as user industry, age, or location and have your product set the default choice based on that knowledge. In doing so, most of your users’ experience is smoother because you’re guiding them to the best-suited selections for them without restricting their ability to change it if required.
Powerful analytics can also come from a tool like Pendo, which can track connected actions that customers follow to achieve an outcome.
You can predict these paths, and the software will help you compare that guess to what customers actually do, or if they give up halfway through your predicted action path. These insights give you a glimpse into the reality of what your users experience as they use your software and will help you improve your user interface for a better experience.
Removing the roadblocks, extra clicks, and wasted time searching for the next action all make it easier for the user to succeed, which will help boost their productivity and the effectiveness of your software.
The New Use Case for Predictive Analytics
Understanding your customers has always been an essential part of a business for long-term growth and success. Companies have learned how to make more money in a variety of ways.
From using special glasses to predict where consumers will look on a store shelf, to making assumptions based on demographics to charge different insurance rates, these techniques are not new.
What is new, however, is companies are now using the same types of analytics to improve the customer experience, not just to sell them more product, but to make sure customers want to use the product and gain a measurable value from it. With the Internet of Things, SaaS-based products, and the app-ification of our lives, data gathering has become easy and entirely non-intrusive to end users.
Businesses which use this data to create a seamless customer experience will enjoy higher retention, more loyalty, fewer support tickets, happier customers, and can then spend more time adding features and delivering new value.