Reshaping CX with AI

WEBVTT
00:00:00.080 --> 00:00:03.680
CX budgets grew just 2% this year.
00:00:03.680 --> 00:00:08.320
But within those same budgets, AI spending jumped 38%.
00:00:08.320 --> 00:00:16.640
And a lot of that money has to come from somewhere, and that somewhere tends to be headcount.
00:00:16.640 --> 00:00:21.839
So with all this cutting going on, what's going to happen in the future?
00:00:21.839 --> 00:00:23.519
Let's talk about it.
00:00:23.519 --> 00:00:40.240
Gartner predicts that half of the companies that cut customer service headcount because of AI will need to rehire by 2027.
00:00:40.240 --> 00:00:47.439
And frankly, they are probably going to have to spend more than they would have if they had kept the people.
00:00:47.439 --> 00:00:49.200
So what's going on here?
00:00:49.200 --> 00:00:53.280
Should we invest in AI or should we invest in people?
00:00:53.280 --> 00:00:55.920
The simple answer is both.
00:00:55.920 --> 00:01:00.880
But we, as customer experience leaders, have to be prepared to have these conversations.
00:01:00.880 --> 00:01:10.640
Leaders currently rank live chat and phone as their most valuable channels ahead of generative AI chatbots.
00:01:10.640 --> 00:01:19.359
However, those same leaders are saying that within two years, Gen AI chatbots will be their most valuable channel.
00:01:19.359 --> 00:01:24.239
Now, with all the money flowing towards AI, this is a little bit of a disconnect.
00:01:24.239 --> 00:01:29.040
It's telling us that the money is flowing faster than the trust.
00:01:29.040 --> 00:01:40.079
So while the money is flowing in, we still don't trust that these channels are going to be able to deliver the types of experience that we want and promise to our customers.
00:01:40.079 --> 00:01:42.719
We still need humans in the mix.
00:01:42.719 --> 00:01:56.239
We still need them to approve, to test, to make sure things are accurate and to provide those better experiences when things are nuanced or complicated or emotional or complex.
00:01:56.239 --> 00:02:09.439
So if you're at one of those organizations that is quickly cutting headcount for the sake of spending more on AI, I encourage you to think about what is the vision that you really have.
00:02:09.439 --> 00:02:19.199
How are we going to get ahead of the competition if we cut the very people who can help us design for the customers they know?
00:02:19.199 --> 00:02:22.879
That headline I mentioned from Gartner, there's more to it.
00:02:22.879 --> 00:02:38.319
Gartner now predicts that across industries, 30% of employees laid off in AI-driven cuts will need to be rehired by 2029, often at a higher cost than if they had never been let go.
00:02:38.319 --> 00:02:52.000
So we may be setting ourselves up for an even bigger budget crisis when, in short-term thinking, we're actually saying we're cutting costs by investing in AI instead of people.
00:02:52.000 --> 00:02:57.039
It simply doesn't look sustainable based on what we've learned so far.
00:02:57.039 --> 00:03:01.039
And there are some headlines that have made this very, very true.
00:03:01.039 --> 00:03:12.080
Klarna announced they cut 700 customer service representatives in favor of AI and then announced within a year they were starting rehiring back.
00:03:12.080 --> 00:03:19.360
That's because they realized that it couldn't actually deliver the same way that people can.
00:03:19.360 --> 00:03:21.599
Now, can AI serve us?
00:03:21.599 --> 00:03:22.400
Absolutely.
00:03:22.400 --> 00:03:31.520
Can it speed up processes for customers and provide convenience and easier adaptation of our processes and rules?
00:03:31.520 --> 00:03:32.879
Absolutely.
00:03:32.879 --> 00:03:35.919
However, we still need humans.
00:03:35.919 --> 00:03:42.000
When we talk about serving customers, we're really talking about serving humans.
00:03:42.000 --> 00:03:49.199
And these AI bots are not going to understand every complex, complicated, and emotional scenario provided.
00:03:49.199 --> 00:03:54.319
One of the things I hear again and again is that we need more empathy in the world.
00:03:54.319 --> 00:03:57.599
We need more empathy when serving customers.
00:03:57.599 --> 00:04:08.319
The only way we're going to get to a place where these chatbots and other tools actually provide that empathy is if we have humans actually training them how to do that.
00:04:08.319 --> 00:04:17.759
Service and support organizations spent a median of about $6 million on AI investments in 2025.
00:04:17.759 --> 00:04:27.279
However, when Gartner evaluated the actual return on those use cases, about 25% delivered negative results.
00:04:27.279 --> 00:04:30.399
About 25% delivered positive results.
00:04:30.399 --> 00:04:39.439
And then there was 42% in uncharted territory, in unclear gray area, that neutral zone.
00:04:39.439 --> 00:04:41.759
We're not quite sure if it's working or not.
00:04:41.759 --> 00:04:45.680
We don't quite know if we're getting the results we want or not.
00:04:45.680 --> 00:04:50.160
That's a big chunk that we just don't know if it's working yet.
00:04:50.160 --> 00:04:56.399
So another way to say this: half of the AI investments right now in spending, they're losing money.
00:04:56.399 --> 00:05:00.639
And fewer than half of the use cases have been shown worth the investment.
00:05:00.639 --> 00:05:15.279
Now, again, I believe this will continue to change as we get better at understanding how to use these tools, where they fit in our workflow and processes, and training people to really train these tools better.
00:05:15.279 --> 00:05:20.000
But until that time, we have to think carefully about where we invest.
00:05:20.000 --> 00:05:24.959
We're also seeing some brand new roles and alternatives to this.
00:05:24.959 --> 00:05:29.920
Forrester mentioned how there are roles like bot unlockers.
00:05:29.920 --> 00:05:33.680
Those are the people who troubleshoot the AI agents.
00:05:33.680 --> 00:05:47.040
And we also continue to need the folks who handle those escalations, judges and experts to really decide what's the right thing to do here instead of leaving all that discernment to AI.
00:05:47.040 --> 00:05:48.959
So why am I sharing this with you today?
00:05:48.959 --> 00:05:52.639
You may or may not have headcount that you're responsible for.
00:05:52.639 --> 00:05:57.040
You may or may not be looking at a budget right now and deciding where to invest.
00:05:57.040 --> 00:06:01.199
But I would argue all of us have a stake in this.
00:06:01.199 --> 00:06:30.240
So if you are deciding even how to use AI or where to incorporate it into your workflows, think about the balance between trying things out, seeing if you get the results, and keeping the humans in the loop, making sure that you have enough investment in the headcount and your teams so that we can actually see the results from the AI investments that we're making.
00:06:30.240 --> 00:06:33.040
This is brave new world territory.
00:06:33.040 --> 00:06:35.040
We are all learning this as we go.
00:06:35.040 --> 00:06:38.399
We are all explorers in this phase.
00:06:38.399 --> 00:06:43.920
So I don't want to discourage you from exploring how to use AI.
00:06:43.920 --> 00:06:52.720
I just want to make sure that you're aware of the risks, the risks to your people, the risks to your customers, and the risks to your organization.
00:06:52.720 --> 00:07:00.959
One of the things that CX leaders are asked to do again and again and again is to prove the ROI, the return on investment.
00:07:00.959 --> 00:07:02.800
Why is any of this worth it?
00:07:02.800 --> 00:07:09.759
Who cares if our customers love us if we can't show that on our profit and loss statement?
00:07:09.759 --> 00:07:26.480
What we have to do right now is decide what is worth exploring and where can we still invest in the people who will actually develop, design, and deliver the customer experiences that we promise.
00:07:26.480 --> 00:07:31.120
One of the ways you can do this is be honest about your use cases.
00:07:31.120 --> 00:07:36.720
Make sure you understand what success looks like before you deploy something.
00:07:36.720 --> 00:07:41.920
And then measure and be honest about was this a success or not.
00:07:41.920 --> 00:07:47.920
All of these things that are introduced, they're not just things to try out.
00:07:47.920 --> 00:07:58.560
They're things that will help us get the outcomes that we need, but we have to understand what that vision is before we just deploy something and cut a bunch of people.
00:07:58.560 --> 00:08:02.480
Now we're living in a really interesting time right now.
00:08:02.480 --> 00:08:06.720
There are warnings about AI, there is excitement about it.
00:08:06.720 --> 00:08:09.920
We don't all know exactly where it's going.
00:08:09.920 --> 00:08:15.439
You as a leader today, your job is to deliver for your organization.
00:08:15.439 --> 00:08:19.120
So make sure you understand your organizational goals.
00:08:19.120 --> 00:08:23.920
And then as you try these things out, tie everything back to those.
00:08:23.920 --> 00:08:30.879
That will put you in a seat of power where you can then decide this use case was great.
00:08:30.879 --> 00:08:32.559
This one didn't work.
00:08:32.559 --> 00:08:41.120
That will show that you are being both financially responsible and also responsible for delivering for your customers.
00:08:41.120 --> 00:08:53.440
Of course, we don't want to deliver technology that will make things more frustrating, less convenient, or worse, provide inaccurate information to our customers.
00:08:53.440 --> 00:08:57.519
We have to constantly be testing and auditing these tools as well.
00:08:57.519 --> 00:09:03.120
Now, I hope this doesn't scare you, but I want it to scare you a little bit, honestly.
00:09:03.120 --> 00:09:09.679
We need to make sure that we understand the why before we just deploy the how.
00:09:09.679 --> 00:09:27.120
So as we embark on this new world where AI is a part of most things we do, let's make sure we're not throwing out the baby with the bathwater by getting rid of all those customer service agents or a whole bunch of people on the team because we think AI can just do it.
00:09:27.120 --> 00:09:29.039
It has to be a balance.
00:09:29.039 --> 00:09:32.240
This is a reshaping, not a replacing.
00:09:32.240 --> 00:09:38.639
So really be aware of the cost of both sides of these decisions moving forward.
00:09:38.639 --> 00:09:40.159
Thank you so much for being here.
00:09:40.159 --> 00:09:41.840
Thank you for all you do.
00:09:41.840 --> 00:09:48.000
Don't forget to review and rate and share this podcast with somebody you think should hear it.
00:09:48.000 --> 00:09:49.519
We love hearing from you.
00:09:49.519 --> 00:09:54.000
You can always leave me a question or a comment at askjeannie.vip.
00:09:54.000 --> 00:09:58.559
And until then, I'll see you and your AI chatbot next time.
00:09:58.559 --> 00:10:09.039
If you're ready to turn insights into action, join CXI Membership, our community for customer experience investigators just like you.
00:10:09.039 --> 00:10:15.679
Get the tools, support, and inspiration to move from ideas to true impact.
00:10:15.679 --> 00:10:23.039
And don't miss my book, Experience is Everything: Making Every Moment Count in the Age of Customer Expectations.
00:10:23.039 --> 00:10:25.200
A USA Today bestseller.
00:10:25.200 --> 00:10:32.000
And if you enjoyed the show, please subscribe, share with your team, and leave a quick review to help others find us.
00:10:32.000 --> 00:10:38.240
Until next time, keep asking questions, keep improving, and keep leading with experience.