Let’s stop telling customers we can’t solve their problems

Sara Yonker

Is it common to have customer service agents who are powerless to help your customers – and can it be avoided?

Customer support agents should fix what’s broken. But sometimes, they need a little more authority – or power – to get the job done.

In our decade of research into customer experience, we’ve seen even the best companies struggle when customer service can’t help customers. We call these moments “Powerless to Help.” 

Our AI powered conversational intelligence platform spots language that indicates when an agent tries to help a customer, but can’t. Sometimes they’re blocked by company policies, internal systems, or limits. To the customer, it doesn’t matter. They just want their problems solved.

Customer support teams who can make your customers feel powerless, too  – and lead to angry customers who stop using your product or service. 

Agent empowerment: Avoiding “I can’t help”

Consider this: You call your bank to question an unusual charge on your account. The team member tries to remove the fraudulent charge, but doesn’t have the ability to dispute charges. Instead, the agent tells you they require you to fill out a form. You’re not near a computer, but the bank has a policy that the customer service team can’t take the information from you because of a policy requiring the account holder to complete the form themselves “for security reasons.” 

In this case, the agent may want to fix the issue – but can’t because of the company’s policy on disputed charges. Your customer satisfaction just took a nosedive. 

In Tethr’s system, we define “powerless to help” as a moment in a conversation when an agent is unable (but not unwilling) to assist customers due to external circumstances. This might be system issues, such as software platform settings. It can also be company policies that might prohibit agents from performing an action even if it would alleviate the problem. 

Common customer support problems 

When we analyzed 4.1 million customer support calls in the last six months, we looked at dozens of metrics, including customer service challenges like the prevalence of “Powerless to Help” language.

The worst-performing company used Powerless to Help language 10 times more than the best-performing company.

In the best case, only 1.7% of a company’s calls had any phrases or keywords that triggered our AI to notice an agent was powerless to help. The worst-performing company had it on 17.7% of calls.

What’s normal? For this analysis, we didn’t look at the average, since an outlier can sometimes skew results. The median performing company had “Powerless to Help” language on 10.6% of calls. The companies at the bottom 25 percentile had it in 14.1 of calls  and the company ranked in the top 75th percentile had it on 8.5% of calls. 

The bottom line: There will be situations at nearly every company that your agents can’t solve. But some companies struggle with limiting agents powers more than others. 

Compared to other negative agent behaviors, Powerless to Help moments weren’t the most common. We saw more agent confusion in communications with customers – Agents who were unsure what steps to take to resolve the issue. 

Want to see more benchmarking data on these and other issues? Download your copy of the ebook here. 

How to know if your agents are empowered to solve problems

Keeping track of your customer journey and what issues your agents can address shouldn’t be anecdotal or manual. Tethr uses AI to ingest every moment from your customer support calls and chats, so you can automatically see analysis – including agent performance metrics.

It also provides benchmarking – which we find can give you perspective on what your customers expect. 

If you want to know if your agents struggle with the ability to help customers (and avoid gaining a reputation as ineffective customer support)  you can test a batch of your phone support calls to see if Tethr observes any of those moments.

Here’s how:

  1. Sign up for a free trial with Tethr
  2. Upload 1,000 audio files of your calls * We don’t require 1,000 – You can analyze just one. But to get accurate ideas about the frequency of events, we recommend as many as possible. 
  3. See what insights Tethr uncovers
Share