Fixing Customer Service Will Require Better Tech

This is not good.

A new report commissioned by Replicant tells a story of customers frustrated when service fails or is way too slow. That’s not new, and Replicant, an AI-based customer service platform, is not the only vendor to ring the alarm.

For many years, Oracle has reminded vendors that one bad service encounter is enough to make customers switch — and many other vendors have followed that lead. But poor service seems to be a festering wound.

What’s new with Replicant’s report is that they put numbers on the problem. The report says that on average 74 percent of those surveyed would happily lose $15.72 to avoid dealing with poor customer service.

That’s significant because it’s often hard to get people to quantify acceptable losses in any situation. Generally, the benefit needs to be twice the cost of the loss, and frankly, fifteen bucks is just minimum wage (on a good day) so there’s room to doubt.

However, there’s no room to doubt the general mood the report delivers. Customer service is a perennial problem that companies work to improve, but it’s a bit like rolling a rock uphill every day.

Desperately Seeking Service

Gadi Shamia, CEO and co-founder of Replicant says in the press release for the report, “This data shows that customers are looking for better service, notice when it’s poor, and voluntarily switch brands as a result.”

Indeed. But what the report only implies, which is the important finding in my book, is that the problem is so universal that quitting a vendor might make you feel better momentarily, but it won’t fix anything.

The sad reality is that you can continue on a journey from one inadequate service encounter (and vendor) to another, only becoming more frustrated. But CRM and companies like Replicant are bringing meaningful change to the equation. Its survey also says the public is beginning to accept dealing with AI and conversational computers if it means shortening wait times and reaching resolutions.

That’s great, though this brings up LTV and technology.

LTV, or customer lifetime value, is not the first metric you think of when contemplating customer service, but maybe it should be. The service industry is awash in analyses of customer satisfaction and the net promoter score (NPS), Ted Mico, CEO of Thankful, told me in an unrelated briefing. Mico thinks we should be more focused on LTV for the simple reason that when a customer leaves, they take their revenue stream with them.

Sure, it’s important to have great customer sat scores, but it’s really, really, really important to grow revenue — something you can’t do easily with attrition. That’s where technology comes in. Over the last decade or so, we’ve watched as automation improved the ways we work in the front office.

Now is a good time to say that code generators didn’t wipe out software developers as a species and intelligent NLP computing won’t eliminate customer service workers, though their jobs will change. Code generators and platforms made (almost) everyone a developer, thus enabling us to have more and more well-tuned systems. What might automation usher in for customer service?

Customer Service as a Marketing Tool

Mico thinks customer service is becoming increasingly essential to marketing, an idea that goes back in time, but which has only been somewhat implemented until now. If you extrapolate just a little, customer service increasingly becomes the loss-leading lead generator that some have been advocating for.

This is also where LTV re-enters the scene. If everyone’s service is average, that’s not very good, but you might figure for every customer you lose, there’s one you gain, a typical zero-sum situation. But as soon as one member of a market adopts better service and increases LTV because it breaks the zero-sum equation, everything changes.

So, if it’s true that service is not good right now, thanks to Covid and other dislocations, you might consider, painful as it is, riding out your poor experience. However, if you are a vendor, now looks like the time to break out of a broken paradigm.

AI is still new in many applications, and although NLP has been around in various guises for a couple of decades, it’s still a good idea to approach this new world with a Plan B.

This means adopting a beginner’s mindset and studying up on what your new system can and can’t do. As usual, the new service systems will be first used in the easy situations, at least that’s what I’d advise.

Also, understand that there’s an education curve with the general public and that the older a customer is, the less interested they might be in the new tech, so give them a choice.

Lastly, study up on LTV; most companies already treat it with reverence, but there’s still room for improvement.