TRANSCRIPT 🔗
WEBVTT
00:00:05.339 --> 00:00:06.330
Customer success.
00:00:10.500 --> 00:00:14.009
Welcome back everyone to the Customer Success Playbook podcast.
00:00:14.009 --> 00:00:15.300
I'm your host, Roman Reon.
00:00:15.599 --> 00:00:18.239
Here with me as always is my co-host Kevin Metzker.
00:00:18.600 --> 00:00:21.510
Kevin, we've made it to Friday the end of the week.
00:00:21.989 --> 00:00:28.620
I know you're excited for AI Friday as we wrap up our three part series with Jake McKee, the Community Guy.
00:00:28.980 --> 00:00:29.879
You ready to go, Kev?
00:00:30.030 --> 00:00:31.140
Yeah, let's get going.
00:00:31.140 --> 00:00:32.130
It's AI Friday.
00:00:32.520 --> 00:00:36.630
Jake's walked us through trust in AI and community driven product builds.
00:00:36.810 --> 00:00:43.259
Today we're gonna talk about how do we break bake empathy and authenticity into AI as we scale.
00:00:43.740 --> 00:00:47.850
So Jake, when companies roll AI pilots into full production.
00:00:48.075 --> 00:00:56.655
As, or when they plan to anyway, what guardrails do we need to think about to keep the experience empathetic instead of robotic?
00:00:56.655 --> 00:01:02.204
And how do we ensure that the AI applications are gonna work, work at scale?
00:01:02.295 --> 00:01:06.105
In other words, how are they gonna make sure that they stay within those guardrails?
00:01:06.105 --> 00:01:08.299
What are the things that companies should be thinking about?
00:01:09.120 --> 00:01:19.079
Well, like we talked about on Monday, this design practice that I've been working on called A IX AI AI Experience design is really focused around that sort of, you know, working principle.
00:01:19.079 --> 00:01:27.750
That, that the tools are not really the end result, the end results, the relationship we're building up with, um, our creative and critical partner, the, the AI system.
00:01:27.750 --> 00:01:27.989
Right.
00:01:28.290 --> 00:01:32.340
But I think, you know, to answer your question, I would actually step back a second and.
00:01:33.180 --> 00:01:35.069
Think about what's happening right now, right?
00:01:35.099 --> 00:01:38.489
Instead of thinking about just when we build a tool, how do we make the tool good?
00:01:38.909 --> 00:01:47.099
I think it's really important, um, that what we're seeing happen right now is, is a, is the AI business transformation, right?
00:01:47.099 --> 00:01:49.079
Very much like we saw, you know what I.
00:01:49.424 --> 00:01:56.444
10, 15, 20 years ago, the digital transformation where we did, we saw bookstores turn into e-commerce, right?
00:01:56.444 --> 00:01:58.724
And, and all what that required.
00:01:58.724 --> 00:02:15.824
That moving away from, uh, you know, it driven, uh, tech teams inside companies and moving into how do we really trans transition the way that our teams work, that we share information, that we work with partners, the way we work with customers through much, much more digital means than analog means.
00:02:16.514 --> 00:02:23.534
So I think we're, we're in this stage where we're really thinking about how does AI transform the way we do business?
00:02:24.134 --> 00:02:25.905
And we have to start there, right?
00:02:25.905 --> 00:02:30.375
Because you can't build trust and empathy if you're not really starting from what are we really doing here?
00:02:30.854 --> 00:02:41.745
And we shouldn't be anyway, just trying to create a chat bot to offload basic sales questions that we think if we put a fairly scripted user experience journey in place, that that.
00:02:42.104 --> 00:02:48.224
Customers will get on the chat bot, they'll ask the questions, and then they won't call us and we'll have call deflection and it'll be saving so much money, right?
00:02:48.794 --> 00:02:57.735
Because we're looking at AI transformation as a way to really rethink how we do basic business processes, how we automate a activities inside the the company.
00:02:58.245 --> 00:03:10.965
And that means thinking about internal processes and thinking about things like the, the data ethics and, and responsible innovation, uh, the overall operating model that we're using around, uh.
00:03:11.594 --> 00:03:14.745
How we create with ai, what are those guardrails?
00:03:14.745 --> 00:03:19.245
What are our upskilling and training and requirements for approvals and that sort of thing.
00:03:19.245 --> 00:03:26.564
All this stuff that goes into making AI more successful for the business as part of transformation.
00:03:26.805 --> 00:03:27.645
That's where we start.
00:03:27.645 --> 00:03:27.854
Right?
00:03:27.854 --> 00:03:39.435
And if we start there, I think what, what naturally starts to happen is we stop thinking about an AI tool as a point solution to replace a very specific singular process.
00:03:40.245 --> 00:03:40.485
That.
00:03:40.784 --> 00:03:51.284
Uh, and we see this a lot right now where, you know, marketing teams are saying, geez, we love, we write a lot of Facebook ad copy and this would be great to just outsource that all to JGBT and we'll be done.
00:03:51.465 --> 00:03:56.115
I think it changes the workflows, I think it changes the ability for us to create, I.
00:03:56.245 --> 00:04:04.465
Uh, in that example, you know, various AB testing, uh, copy that we can do a whole lot more experimentation, which is awesome, right?
00:04:04.914 --> 00:04:08.064
And, and, but that's a change to the workflow as much as it is anything.
00:04:08.064 --> 00:04:11.754
And so back to this idea of how do we make it more humanistic?
00:04:11.754 --> 00:04:16.944
It is like we talked about on Monday, designing that relationship with the AI system and the humans that are using it.
00:04:17.584 --> 00:04:21.425
That is happening within this general, uh, AI transformation.
00:04:21.694 --> 00:04:39.305
So if we're trying to replace internal processes or add an ability to, to connect associated, uh, data sources and make better, more interesting conclusions for our, uh, data analysis teams, these are
Load full transcript
things that are really meant to do to, to, to have outcomes that, that will help us to transfer, translate, transform.
00:04:39.305 --> 00:04:40.175
That's the word I was looking for.
00:04:40.355 --> 00:04:41.615
Transform the business.
00:04:42.144 --> 00:04:43.644
What do we then do with that?
00:04:43.644 --> 00:04:43.915
Right?
00:04:43.915 --> 00:04:51.235
So, you know, trust is, uh, uh, one of the A IX principles is, is trust happens in micro moments as much as big moments, right?
00:04:51.235 --> 00:04:54.774
That how we build trust with, with other humans, right?
00:04:54.774 --> 00:04:56.514
Back to this relationship building construct.
00:04:57.415 --> 00:04:59.095
How we build trust happens.
00:04:59.274 --> 00:04:59.725
Sure.
00:04:59.725 --> 00:05:04.795
That we're not committing fraud, that not being violent, but of course those big, huge things.
00:05:04.795 --> 00:05:08.454
But also that if I say I'm gonna take out the trash, I take out the trash.
00:05:09.285 --> 00:05:12.074
If I say, Hey, I, I keep forgetting to lock the front door.
00:05:12.074 --> 00:05:13.634
I'll work on it, that I actually work on it.
00:05:14.024 --> 00:05:19.904
Whatever it is, you know, really making sure that some of those little moments are continually built up over time.
00:05:19.904 --> 00:05:28.754
That when I hear from the IT team that this process is going, going to improve something I'm doing internally to track my hours or whatever it might be.
00:05:29.225 --> 00:05:39.935
That it really does, and that they're coming to me with not just something that they've done in a vacuum like it teams often do, and then roll it out and say, you gotta use it, and we're all like, this is terrible.
00:05:40.204 --> 00:05:46.264
My favorite example of that being the 13th screen internal process for a company that I worked with, literally 13th screens.
00:05:46.865 --> 00:05:48.875
It wasn't until the 13th screen.
00:05:49.024 --> 00:05:55.024
It took like two to three hours to do all this work in this, but it wasn't until the 13th screen that they had a save button.
00:05:55.620 --> 00:05:58.860
If your browser crashed on on screen four.
00:05:59.475 --> 00:06:00.704
You were starting over, right?
00:06:00.704 --> 00:06:05.144
Well, of course that got feedback pretty quickly, but that development team had already moved on.
00:06:05.235 --> 00:06:08.805
Gotta stop what they were doing, get back and re refresh their brains on what was going on.
00:06:09.285 --> 00:06:12.254
Trust happens in a bunch of different ways over time.
00:06:12.254 --> 00:06:13.964
That is all focused on.
00:06:14.490 --> 00:06:16.560
Building that relationship and that connection.
00:06:16.620 --> 00:06:23.430
There's faith from the user that somebody was thinking about me, somebody was trying to do something specific.
00:06:23.430 --> 00:06:28.199
They obviously understood my, my problem, not just, uh, their problem.
00:06:28.319 --> 00:06:29.550
Right Back to everybody goes home.
00:06:29.550 --> 00:06:31.290
Happy mantra from from Wednesday.
00:06:31.649 --> 00:06:43.829
So I think that's a, a, a, it's a very long-winded starting point, but that's the starting point is really thinking about this relationship building and, and how, as you've transformed your business, are you really transforming all the parts of the business that go into that experience?
00:06:44.194 --> 00:06:49.805
You really understand the experience, you're able to add value, not just replace it with some, yet another change.
00:06:50.225 --> 00:07:02.375
I think all of us that work in corporate America are overwhelmed by change, constant change, constant reorgs, constant restructures, constant new bosses, constant, everything can, that can whack our trust as well.
00:07:02.375 --> 00:07:05.704
So you know, that's a huge part of how we design the systems.
00:07:06.675 --> 00:07:10.365
Less and, and, and yeah, there was ways in the system you can design trust.
00:07:10.365 --> 00:07:20.264
But you know, I always like to talk about this from the very beginning of when trust starts and that that is telling me, Hey, I've got a new tool for you, but I understand your problem well enough that I know you're gonna be excited.
00:07:20.264 --> 00:07:24.764
Not just frustrated that it's yet another tool just because somebody got a new contract over in it.
00:07:24.764 --> 00:07:24.855
Right?
00:07:26.250 --> 00:07:36.480
Well, I, I love Jake, your examples, because I think sometimes you hear relationship and, and you know, we're the Customer Success Playbook podcast, so a lot of it's customer focused, right?
00:07:36.480 --> 00:07:41.850
Like in terms of the relationship between the, the organization and the customer and the journey mapping and all that.
00:07:41.850 --> 00:07:48.355
But you're so many times forget ai, any new tool internally, there's still relationships there too, right?
00:07:48.480 --> 00:07:52.649
Like, uh, like you gave the example, like the IT department says, Hey, we've done this and go use it.
00:07:52.649 --> 00:07:53.730
And it's like, whoa, whoa.
00:07:54.079 --> 00:07:58.790
Has anyone understood like how, how this impacts me, how I use it and Right.
00:07:58.790 --> 00:08:00.829
But as relationships, it, it's the same thing, right?
00:08:00.829 --> 00:08:04.939
Like, and keep me honest here, Jake, external internal relationships are all the same.
00:08:04.970 --> 00:08:06.860
I mean, that those relationships are just as important.
00:08:06.889 --> 00:08:06.980
Mm-hmm.
00:08:07.220 --> 00:08:07.519
Agreed.
00:08:07.519 --> 00:08:10.850
And, you know, back to this, the, to the, to the front end, the experience piece.
00:08:10.850 --> 00:08:11.060
So.
00:08:11.654 --> 00:08:17.295
I'm sure that that, um, some of the listeners, when Kevin, when you asked your question and they were like, okay, cool.
00:08:17.295 --> 00:08:20.115
Tell me all about the things I need to do to make a good AI tool.
00:08:20.535 --> 00:08:22.305
I think we
know a lot of those already.
00:08:22.305 --> 00:08:27.644
We're learning a whole lot more about those, but they're, they're really, they're still in this vein of relationship, right?
00:08:27.644 --> 00:08:30.165
That you're making things quick when I need it to be quick.
00:08:30.165 --> 00:08:31.274
You're making things slower.
00:08:31.274 --> 00:08:33.075
When I could use a moment to slow down.
00:08:33.394 --> 00:08:42.184
If you're not, and we hear about this one all the time, if you're gonna hallucinate, tell me, or at least be clear about the nomenclature for what's happening for a project I'm working on right now.
00:08:42.184 --> 00:08:49.174
I did this earlier this morning where I asked Chad CPT for something and it, and it gave me an unsourced thing, and I had to ask, is this.
00:08:49.365 --> 00:08:50.235
Based on anything.
00:08:50.235 --> 00:08:53.294
Sometimes I do and it's, it is based on something they just didn't tell me.
00:08:53.625 --> 00:08:53.924
Right.
00:08:53.985 --> 00:08:55.815
And other times it's not based on anything.
00:08:55.815 --> 00:09:08.115
It made it up, but it's able to tell me and being able to bring some of that up earlier, you know, telling your wife to be home late is, is one thing, but not showing up on time and then just getting home late and telling her, oops, I didn't tell you, is a whole different story.
00:09:08.115 --> 00:09:08.325
Right.
00:09:08.414 --> 00:09:10.034
There's a lot of those little humanistic behaviors.
00:09:10.034 --> 00:09:11.534
I think we can start to pull into that.
00:09:11.534 --> 00:09:15.764
But, but again, I, I just, I so desperately don't wanna forget that.
00:09:16.139 --> 00:09:21.149
It's the context in which we have these tools that's as important as the tool itself.
00:09:21.539 --> 00:09:21.690
Mm-hmm.
00:09:21.720 --> 00:09:32.820
You know, I could have a great, uh, AI interviewing function, uh, that collects, you know, if you're doing something with a candidate, you have an AI tool that's interviewing them as the first clearinghouse of, of people.
00:09:33.179 --> 00:09:33.870
Okay, fine.
00:09:33.870 --> 00:09:34.440
No problem.
00:09:35.190 --> 00:09:42.929
If a, I believe that that's actually gonna be paid attention to, it's not gonna get turned into an automated transcript that then goes through an AI filter and no human ever sees it.
00:09:42.929 --> 00:09:44.759
And so what's the point of me doing this?
00:09:44.759 --> 00:09:45.059
Really?
00:09:45.059 --> 00:09:45.120
I.
00:09:46.004 --> 00:09:46.455
Issue one.
00:09:46.455 --> 00:09:47.384
Issue two is.
00:09:47.820 --> 00:09:49.950
Help me warm up to be my best self.
00:09:49.950 --> 00:09:56.759
Don't just start asking me questions and now I've got a, in a robotically human voice instead of a robot or a human voice.
00:09:57.029 --> 00:09:59.399
Can't really tell who I'm talking to, and it feels uncomfortable.
00:09:59.399 --> 00:10:01.230
It takes me a minute to settle into the experience.
00:10:01.500 --> 00:10:02.669
Gimme a few minutes to warm up.
00:10:02.669 --> 00:10:06.179
Yes, it's those humanistic traits of I'm gonna have good conversation.
00:10:06.179 --> 00:10:07.470
What does that look like in real life?
00:10:07.799 --> 00:10:10.080
It's funny, I, I was doing a interview.
00:10:10.169 --> 00:10:16.080
I had an AI interview me recently on something and it conflated.
00:10:16.455 --> 00:10:19.004
Some of the numbers I gave it and got it wrong.
00:10:19.004 --> 00:10:23.024
I tried to correct it, still got it wrong and I was like, all, you've got it good enough.
00:10:24.284 --> 00:10:30.164
You get to a point where from a relationship standpoint, it's like, okay, I'm talking to an ai, you're not going to understand it.
00:10:30.164 --> 00:10:30.254
Yeah.
00:10:30.554 --> 00:10:33.794
I'm not gonna explain it 15 times because you think you understand it a certain way.
00:10:34.215 --> 00:10:36.345
It's never really understanding.
00:10:36.345 --> 00:10:36.644
Right.
00:10:36.649 --> 00:10:36.870
It.
00:10:37.215 --> 00:10:44.054
Only a predicting tool, so don't know what garbage bin that went to, but, well, and one of the reasons
00:10:44.054 --> 00:10:53.804
why, you know, when we talked on, on Wednesday about the community driven product development work that I do, where you bring in passionate customers from, from your community base into the product of design.
00:10:54.125 --> 00:10:55.024
Cycle itself.
00:10:55.355 --> 00:11:04.054
One of the values of that process is that you're not just asking people once for a thing, but you're seeing over time how they experience and learn and grow from the thing.
00:11:04.534 --> 00:11:12.274
So if I put a, you know, a new product in front of somebody, the feedback that they'll give me upfront, maybe a little colored by what they want.
00:11:13.034 --> 00:11:16.664
To say to me based on what they think I want to hear as a development team, right?
00:11:16.664 --> 00:11:17.504
As a product manager.
00:11:17.924 --> 00:11:23.384
And then that kind of fades away, but then they get excited about using it'cause it's new and they're being invited to this experience.
00:11:23.384 --> 00:11:27.434
But then two weeks later they may actually get to the point of, okay, this is the real experience.
00:11:27.434 --> 00:11:28.875
But that took a minute, right?
00:11:28.875 --> 00:11:31.184
That took a, a second to, to make that connection.
00:11:31.544 --> 00:11:37.575
And I think that that's, you know, thinking about stuff like that with, with AI experiences as well, where.
00:11:37.980 --> 00:11:41.009
The robotic human voice instead of the robot or the human voice.
00:11:41.009 --> 00:11:41.250
Right?
00:11:41.789 --> 00:11:46.320
Either I need a, a very humanistic voice to get into this dynamic.
00:11:46.320 --> 00:11:54.870
And then what we're prompting the, the interview questions, how we're going about, do we submit them in advance and let the AI read it to us, and then extrapolate from there?
00:11:55.169 --> 00:11:58.049
Or is it creating its own paths, right?
00:11:58.110 --> 00:12:01.710
And really understanding, uh, as we develop tools like that.
00:12:03.149 --> 00:12:05.250
Processes, how does this fit into the process?
00:12:05.250 --> 00:12:07.379
How does a person feel using this?
00:12:07.740 --> 00:12:10.889
How does the system come off to them from an emotional standpoint?
00:12:10.980 --> 00:12:22.470
You know, when they're nervous and doing a job interview with an ai and it doesn't, you know, if it started off saying, I'm the robot, I'm here to pre-filter a human will absolutely look at this because I can't be trusted as my own.
00:12:22.830 --> 00:12:28.019
Robot experience, you might kinda laugh for a second, then it asks a few warmup questions and you're like, oh, this is kind of funny.
00:12:28.019 --> 00:12:29.940
A, a robot's asking me how my day goes.
00:12:29.940 --> 00:12:36.029
That's a very different experience than a fake human trying to pretend like it's asking its own questions.
00:12:36.434 --> 00:12:43.544
Even though you may have fed it those questions in the beginning or it's, it's responding just to create conversation with no particular point.
00:12:43.575 --> 00:12:43.815
Right.
00:12:44.235 --> 00:12:47.654
Um, these are process questions as much as they're technology questions.
00:12:47.654 --> 00:12:48.075
I think
00:12:48.674 --> 00:12:51.345
come, I think gentlemen to the end of our three part series.
00:12:51.700 --> 00:12:53.590
Jake, awesome stuff all week.
00:12:53.769 --> 00:12:57.370
I appreciate all your, all your insights and, and for joining the show.
00:12:57.700 --> 00:13:00.399
Can you give us, uh, tell our audience where they can find you at?
00:13:00.700 --> 00:13:00.909
Very
00:13:00.909 --> 00:13:01.269
simple.
00:13:01.269 --> 00:13:02.409
So my name's Jake McKee.
00:13:02.409 --> 00:13:04.240
My website is jake mckee.com.
00:13:04.629 --> 00:13:11.860
Jake mckee.com/ix for, for more information on the, the AI experience, design work and, and practice that I'm doing.
00:13:11.860 --> 00:13:14.590
But, uh, yeah, that's it for me, jake mckee.com.
00:13:15.009 --> 00:13:15.370
Awesome.
00:13:15.370 --> 00:13:16.360
Easy enough to remember.
00:13:16.360 --> 00:13:16.720
Audience.
00:13:16.720 --> 00:13:19.629
We will have it in our show notes as well, so you can find it there as well.
00:13:20.039 --> 00:13:22.799
If you love these episodes, please subscribe.
00:13:22.799 --> 00:13:23.235
Rate'em.
00:13:23.240 --> 00:13:24.029
Share'em with your team.
00:13:24.029 --> 00:13:27.509
It helps us bring actionable insights to a broader audience.
00:13:27.720 --> 00:13:34.110
Kev, we're back next week with more strategies for our audience's, customer success playbooks.
00:13:34.470 --> 00:13:37.110
I'm excited for what we have lined up over the next couple months.
00:13:37.470 --> 00:13:40.259
Until next time, Kevin, keep on plan.