Skip to Content

You’ve Automated Your Customer Service Process – What’s Next?

Man using a laptop computer
Once you automate customer service processes, evaluate what’s working and what’s not for continuous improvement.

This is how to evaluate your customer service automation efforts to see what’s improved for both customers and employees – and where to go next.

Customer service automation is key to efficiency in the contact center. But according to research from Boston Consulting Group (BCG), 70% of digital transformation efforts fall short of their objectives. Too often not enough attention is given to the people and the work required to get the organization on board. Real customer service automation improvement ensures an automated workflow that people use, and that achieves the time, efficiency, and customer experience improvements you set out to address. But how do you know if your automation is working well? And if it’s not, what can you do about it?

Here are five steps to ensure that your customer service process automation is primed for success:

1. Test your automated workflow

Before launching company-wide, check that your automation works as planned, and that it works for the people who will use it. Test the workflow with a small cross-section of people who are typical contributors to the service process. 

For example, if your automated workflow has steps that require responses from middle- and back-office employees in addition to customer service agents, make sure that you include all these users in your test. Ask for feedback on the experience. Include employees who have been using the original process for a long time, as well as a few new to their roles. 

Finally, be sure to capture and quantify the process improvements that your test reveals, including cycle time, cost, and experience improvements. 

2. Evangelize the automation process

If your test showed positive process improvements, make your workflow automation available to all the people who will use it. People need to actually use the workflow at scale for your service organization to see the full benefits.

Make sure that users know about the workflow automation and that they understand how to engage with it. This goes for internal stakeholders — front-, middle- and back-office employees — as well as external stakeholders such as partners and customers.

Evaluation is critical. We want to have continuous learning — every day, every week — about how our automated processes are performing.

Alvin Stokes, chief customer contact officer, Smiledirectclub

Communicate the specific improvements that your test revealed. For example, if your automation reduces the process cycle time from 30 minutes to seven, tell people about it. The same goes for things like increased opportunities for cross-selling and upselling. Clarity about the benefits — and understanding what motivates your team — will help ensure you get buy-in and engagement. 

Internally, spread the word through things like learning lunches, as well as through departmental and leadership communications channels. Externally, make sure customers and partners know exactly where to find and access links bringing them into an automated process (a self-service or partner portal, for example.)  

3. Measure your automated workflow results

Once up and running at scale, gather as many metrics as possible about how the automated workflow performs. One important metric is the average time the automated process takes to complete. You’ll also want to understand the impacts of the automated process on your customer service response time, issue resolution time, costs, and customer satisfaction scores (CSAT). 

Next, look at usage stats. What percentage of your process stakeholders are engaging with the automated workflow? Also measure the number of times the automated workflow is engaged daily, weekly, and over the last 30 days. Are there instances when the automated process is engaged but not completed? Are there errors? How many?

Always remember to keep the customer experience front and center.

Finally, if a new action has been automated, like a specific self-service task, you’ll also want to measure outcomes. For example, compare how many product registrations your customers are doing now on their own versus how many are still completed by your agents. See if you can measure whether employee satisfaction and productivity have gone up as a result. 

Essentially, you want to go back to all the metrics you saw when identifying processes that were good candidates for automation (learn how to begin the automation process). Make sure you’re capturing the new metrics after automation is applied. 

4. Evaluate the data

Compare your automation metrics against the original metrics that encouraged you to select this process for automation. 

For example, if your goal was to improve the cycle time of your returns process, compare this metric before and after automation. Has the cycle time improved? By how much? If your process cycle time is not showing a lot of improvement, you’ll need to identify the issue. Is the process still too complex? Or are there data hangups and information bottlenecks (like remaining manual steps) that automation hasn’t fixed? Do this for every metric — cost, CSAT, case resolution, etc. — that you want to address with automation. 

Also evaluate your adoption numbers. If your employees aren’t using the automated workflow, you need to dig into why. The same goes for your customers. For example, are customers still calling the contact center instead of going to the self-service portal? Maybe the self-service form is not published in an easy-to-find spot or channel. Perhaps the forms you’re asking them to fill out are too long and you need to simplify. Evaluation will help you learn what’s working and what’s not.  

Let’s look at an example. Service process automation is a big priority for SmileDirectClub, an international oral care company and telehealth pioneer. “As we take advantage of automation to orchestrate the best customer experience possible, evaluation is critical,” said Alvin Stokes, chief customer contact officer. “We want to have continuous learning — every day, every week — about how our automated processes are performing.” 

Stokes pointed out that this is where technologies like artificial intelligence (AI), modeling, machine learning, and advanced analytics are really valuable. These help the company better understand every component of the customer experience and continuously identify better ways to automate.  

5. Refine and adjust as needed

Sometimes a small adjustment will improve the process further, like splitting or simplifying a self-service form that is too long or overly complicated. In the case of low user adoption, you might need to go back and do more evangelizing. 

If refining the process is not a simple fix, more investigation is needed. Is there a data integration problem that hasn’t been solved, or a missing data connection? Why are users especially resistant to adoption? It’s important to pinpoint specifically what is standing in the way. 

Also keep in mind that processes will probably need to be refined periodically over time, especially as conditions (internal and external) change. This is why it’s so important to have a flexible automation platform, with the key technical capabilities to ensure your automation efforts succeed. 

And finally, always remember to keep the customer experience front and center. Whether your initial goals were to speed the process cycle time or to free agents up for cross-selling and upselling, in the end it all comes back to the customer. Make their interactions with your company better, easier, faster, and more seamless. 

If you’ve done a good job preparing for and building your service process automation, it’s likely that your organization is realizing the benefits you set out to achieve. Close the loop through testing, evangelizing, measuring, and evaluating to help you validate and quantify the value of your automated process. Get ongoing user buy-in, pinpoint areas to refine for ongoing success, and, most importantly, deliver a better customer experience.

And here’s the exciting part: Your automation efforts aren’t going to end with one successful implementation. Bask briefly in the glow of your success. Then begin again by asking yourself the five questions to launch your next automation process. Move on to the next process challenges on your list, which, once automated, will bring even more benefits to your team, your company, and the customer experience.

Interested to learn more about how to succeed with automation?

Check out the complete guide for step-by-step, practical help getting started, creating your road map, and improving your processes.

Sophia Le-Dimitrova Sr. Manager, Service Product Marketing

Sophia Le-Dimitrova is a Director of Product Marketing at Salesforce. She focuses on helping organizations deliver exceptional customer experiences at scale through Service Cloud solutions such as AI-powered automation, Customer Service Incident Management, and Slack swarming.

More by Sophia

Get the latest articles in your inbox.