Using AI to Understand Why You’re Losing Clients

This transcript was created using AI tools and is not a verbatim, word-for-word transcript of the episode. Please forgive any errors or omissions from the finished product.

[00:00:00]

Courtney: Everybody knows it’s cheaper to keep a client than gain a new one. So how can AI ensure that you keep them?

And what if AI could also give you early warning signals that a client is at risk?

Hi, I’m Courtney Baker, and this is the AI Knowhow Podcast from Knownwell, helping you reimagine your business in the AI era. As always, I’m joined by Knownwell CEO David DeWolf, Chief Strategy Officer Pete Buer, and Chief Product Officer Mohan Rao. We also have a discussion with Noah Curran of Monkedia about why finding the right customers is one of the real keys to client retention. But first, let’s turn our attention to a brand new segment.

We’re calling AI in the Wild.

Pete Buer joins us as always to dive into some of the latest and greatest advancements in the world of AI.

Hey, Pete.

Pete: Hey Courtney, how are you?

Courtney: Doing good, Pete. This week we’re [00:01:00] introducing another new segment that we hope will become, uh, a real fan favorite here on AI Knowhow. We’re calling this segment, AI in the Wild. So many of us are just looking for examples of how AI is actually being used very practically in the real world. So, Pete, can you share with us, with me and the listeners, a real world application of AI you’ve seen or heard about recently and the takeaways for business leaders?

Pete: Sure. The example I bring along today, because I know you’re such a fan of growing vegetables, uh, comes, comes from IBM. Um, and it’s the, uh, story of how IBM is pulling together disparate data sources and feeds to create a more universal and usable view of farm operations. So, if you think about it. There is an absolute ton of data that goes into the [00:02:00] running of a farm.

Um, much of it is not connected or organized or associated or visualized, right. Um, IBM has created this, uh, Watson decision platform for agriculture, which taps into AI and data analytics. Uh, and the Internet of Things to help farmers improve the performance of their operations. So think about pulling levers like, uh, increasing profitability through greater crop yield, uh, elevating quality, more protein per pound, produced or improving sustainability, less energy consumed across time.

They take all the relevant structured and unstructured data from farm ops, uh, and associated, uh, inputs. Think weather, weather patterns, soil composition, equipment status, visual imagery from satellites and dunes and drones and, uh, farm workflow data. So how, and when you plant, you fertilize, you irrigate, you harvest, you turn over the [00:03:00] fields, whatever, um, to use it all to create an operational dashboard.

For the farm. So we got pests out in quadrant 14 B, tractor number 30, threes running low on steering fluid and the back 40 are drying out from a combination of heat and wind, right? Like. Perfect information to have at your fingertips if you need and want to react to drive farm performance to its optimal level.

And it provides all the relevant solutions to the problems that, uh, it identifies. So we always ask ourselves, what’s the lesson for leaders? Well, poetically it’s plant the AIC seeds, uh, in your organization now and watch them grow. But, but practically. Look across the business and understand what problems, big problems you wanna solve with what data, and get cracking, make hay while the sun is shining.

Courtney: This is so good, and honestly, I can’t even think about the business applications right now because I’m just thinking [00:04:00] about my own little garden and should I be using AI to build a new gardening calendar? Um, so. There you go. I, I do think this is really interesting and always fun to see how AI is actually being deployed in organizations and producing return.

Pete, thank you for this. This is a great example.

Pete: Okay, Courtney, take care.

Courtney: Losing clients is never fun. Losing them unexpectedly, even less fun. I talked with David DeWolf and Mohan Rao about the role AI will play in helping leaders understand when and why their clients are at risk and what they can do about it.

David Mohan. Today I wanna talk to both of you about how do we actually [00:05:00] leverage AI to help us understand those moments, those moments for us as humans feel like. I just got surprised. I didn’t

David: mm-Hmm mm-Hmm.

Courtney: Was a potential, felt really secure that that client was gonna renew or whatever the

David: Yeah.

Courtney: Is there an opportunity that AI can actually help take that very painful moment away? If so, tell us how.

David: Yeah. let’s frame this a little bit and talk about why is this so frame, uh, painful. I, I would start with, I think as businesses, customers are what we revolve around, right? Our clients are the life blood. Of our organization. And that’s not to diminish employees, it’s not to diminish other things, but organizations exist if you’re a business to serve your clientele.

And um, it is essential that there is some stability there. And I think the reason why [00:06:00] surprise. Is such a hard thing to grapple with is that if my entire intent, if the orientation is to serve clients, to add value to the world, to deliver this value, um, and it’s one thing for you to churn a client for that client to go away for that client to have struggles.

It’s another. If I’m not even aware that that is gonna happen and I’m caught off guard by it. Right? And so it’s this shock to the system almost a double punch, right? It’s, it’s, wow. Not only did I lose it, I didn’t even have a chance to fix it yet. It would’ve been the highest priority if I would’ve known.

Courtney: Mm.

David: And, and I think that is the root of it. And what’s interesting is you listen to that. If I would’ve known, why don’t we know? We don’t know? Because as we scale our businesses, we go from a point of being able to be involved in all of these conversations to, [00:07:00] especially in B2B services, there are these multi-threaded relationships that it just becomes impossible for anyone to sit in the middle of and pick up all of the individual discreet signals and to connect dots and find signal from noise.

Right, and, and it’s a scalability problem. It’s hard. And what has made it even more difficult in the past, probably 20 years, right? But it’s accelerated so much in the last 10, is that not only do we have more connections than ever, we have more channels than ever. So even if you wanted to sit at the hub of all of those.

You couldn’t get your hands on top of all of the different channels of communication that are happening. And so I think the, the problem has not only multiplied, it has exploded, and it has become more and more difficult.

Mohan: Yeah. Do. Do you think that’s because, with the tools that we have, uh, we sort of measure it in an analytical fashion

David: Hmm.

Mohan: [00:08:00] Really kind of what the customer is thinking, uh, uh, is more at an emotional level. They’re thinking and feeling about the problems as opposed to analytically reporting.

And, um, maybe, maybe there’s a solution there that AI

David: Yeah. I think that’s so true. Mohan. I was actually talking a couple of weeks ago, um, to a chief people officer actually who was trying to get their head around this and really understood it, uh, stand it, and, and they got it in their mind, but they really didn’t get it in their gut until I made the analogy.

I said, you know how as a chief people officer, you have this gut? Every now and then that somebody in your organization is not as engaged as they should be, maybe at risk of departure and you can’t quantify it. You can’t tell me the three things you saw that made you think that it’s just a gut. And you know, from experience to act on that.

And she was like, oh yeah, I know exactly what you’re [00:09:00] talking about. And I said, that’s it. Right there. When you are in the business of serving clients, you have the exact same thing. It’s not an analytical problem that you can compute and do math off of. I can’t create business rules for it. It is incredibly probabilistic where I notice 57 different things and it means something to me and it means something to me that I can’t.

Really, really, really process and boil down to an algorithm, and I think that’s what you’re talking about. It’s that gut instinct. It’s that feeling that we need to be able to act on that. If it’s all in a report, if it’s all in a dashboard, if it’s all just numbers, doesn’t speak to us.

Mohan: Do you think the issue is more about the incompleteness because of the very thing that we are discussing right now, the emotional masking of actual, uh, situation. um, is it more, the subjectivity of, where things are, uh, what, what do you think is the prime reason here, uh, for these constant surprises that Courtney [00:10:00] mentioned?

David: Yeah, I, I think it’s actually the intersection of a couple of things. Mohan, I think you hit on two of them. I think one is the incompleteness, right? When you simply put, you can’t find trends and connect the dots if you don’t have all of the dots. And so it’s hard, so it lacks information, but I think also because so much of that information is consumed secondhand, not firsthand. You also miss clues and context, and oftentimes you’re either getting incomplete signal from the signals themselves, right? So you may have a data point, but it’s missing all the context, right? What, what is it said that like body language makes up like 90% of communication or something like that, right?

Um, you miss all of that. And it may be interpreted through somebody else as well. So you’re not only getting an incomplete story, you’re oftentimes getting a wrong story or at least a more incomplete than you even imagined story. And going through those filters [00:11:00] of subjective interpretation, I think just exacerbates this problem. And so can you point it to one thing? It’s incomplete. Yes, that’s true. But also if you’re looking at wrong signal, it’s gonna throw you off even more. Right? And so I think it, it matters that all of these intersect together as much as anything.

Courtney: So we’ve established this is painful. We don’t want it to happen. We would all like it to end. You know, obviously we are a podcast about AI, so I’d love to talk about that. How can AI actually help solve some of this pain. You know, it’s such a unique moment in time. What does that look like?

Mohan: I think it’s about solutions to the things that, um, David mentioned. Uh, right. So, uh, there is, um, when, when an organization is scaled, no one person has all of the, uh, uh, data required. When I say data, it’s not just analytical data, but [00:12:00] also emotional. Uh, aspects of that data and, uh, and the question is how do you break those silos down, uh, into getting a complete picture, right?

Even then, you may not completely a hundred percent eliminate surprises, but hopefully it should go down a lot. Uh, it’s similar to how. Uh, when the company was smaller, you could keep it all in your head and

David: Hmm.

Mohan: a scale stage that you can’t anymore. So move more towards, uh, that sense of it. I think siloing or breaking down the silo so you have complete information of your clients is a great first step I think.

David: I think that’s true, and I think there’s just a, through all of that, there’s an institutionalizing of it. It is, right, right now it’s this nebulous thing. When I think about how we measure the health of our client relationships, especially in the services industry, what I think about is things like the NPS score, uh, csat, right?

It, it’s very, anecdotal. It is, uh, very, [00:13:00] um, uh, stale, right? You don’t have operational daily data on NPS and what the current sentiment of the client is. There’s just so many flaws to it that it’s underdeveloped and immature, and so to me it is everything that Mohan just said. Plus the institutionalizing of a benchmark.

The institutionalizing of how do we turn all of this into not just a gut feeling, but if we truly want to scale it, how do we create a standard out, out of it, a benchmark out of it, have a score that we can say, this is how it’s trending today, right? This week. Um. And then the deep understanding of what do you do about it, right?

In order to game that system and to start to improve that score, um, to really be able to act on it, I think is critical. And here’s one of the things that I really believe, keeping and maintaining your clients, right? This has nothing to do with AI, but I think the AI can help solve it. It is not a client [00:14:00] management or an account management or a client success problem, and too many organizations look at it that way.

Oh, you are. You are tasked with managing the client and retaining them, and they’re your key metric. It takes a village. You’ve got to have an entire organization aligned around serving your clients. And if you don’t, you’re broken. And so if we can institutionalize it so everybody is singing off the same sheet of music, everybody understands the state of that client and what needs to be done to improve it, and is watching that score go off up, I think we’ll be better off.

Courtney: Several years ago this was, I’m sure y’all remember this. Moment when you would go through the Chick-fil-A drive through. It’s not where y’all thought

David: It all comes back to Chick-fil-A all the time. It’s always Chick-fil-A

Mohan: What

David: No idea. That’s where we were going.

Courtney: Okay. You would go through the Chick-fil-A drive through, and I here in Nashville, uh, [00:15:00] we’re, we’re at, we’re at test market for Chick-fil-A. Um, so we get new stuff

David: Now she’s just bragging. Okay.

Courtney: We got those doors. You know, they would come out to you now like gone, was the drive through window. But at that moment you could actually look into the store

David: Yep.

Courtney: You would see this like dashboard of like everything that was happening with the cars in line.

You know, like how long they had been waiting, you know, like how many, you know, sandwiches they had made that it was like real time data on their customers. Sitting in their drive through line. And I remember, um, the CEO of the company I worked for at the time was like, that’s what we need. And I feel like in a lot of ways she was right.

I think

David: Hmm

Courtney: We didn’t have at that time the technology to really be able to do that. And think what the two of you said about AI remedying in this of breaking down silos, embracing that objectivity, really filling in the data gaps and being proactive. Basically, we were [00:16:00] just doing what Chick-fil-A was doing 10 years ago.

No, I’m

David: Yeah, though what Chick-fil-A hasn’t done yet is, I don’t think they have cameras seeing if I’m smiling or not to know what my sentiment is. Right.

But no, I’m just kidding. true, true.

They may and I just don’t know it. I don’t know.

Courtney: Good point. Let’s, let’s not put it past them. Um, but I, I do. And just, you know, breaking this down reminds me of that story, and I do, I think it’s such an exciting time. It’s such an exciting time to have technology. Okay. I feel like we say this all the time, but really helps elevate humanity.

It helps us be better engaging with, again, the most important part of our company, which is the people that we’re serving.

Mohan: Often with surprises, um, you know, you don’t know what’s missing, right? I mean, that’s the very definition of a surprise.

Courtney: Yeah.

Mohan: What, um. The technology, the AI technology can do is to look at model the good clients you have, and then be able to, uh, put a spotlight on [00:17:00] data that you don’t have for the other clients.

Uh, so you can create, uh, little cohorts of your own clients where you know their good clients, you know, they’re healthy. know, what you see in them, and then you’ll be able to see what you don’t with your Chick-fil-A analogy, maybe you can kind of look at multiple locations of Chick-fil-A and

David: Hmm.

Mohan: compare the metrics

David: Hmm.

Mohan: Able to see what you see in the best, uh, locations versus not the best locations, right?

I mean, be able to do that sort of, um, comparative analysis. It all really comes down to. Uh, if you want to eliminate surprises, you gotta think through. How do you figure out what’s missing in the data set that you’ve got both emotional and analytical, and also be able to figure out the root causes of the issues, uh, right.

If you do that, it’s gonna go a long way in a happy Chick-fil-A experience.

Courtney: I love that. Um, I think this is.

David: A happy Chick-fil-A experience [00:18:00] we’ve, really come a long way in this podcast.

Courtney: Our producer’s gonna fire us. David, Mohan. I would to say, in case you’re listening to this and you don’t know it, this is exactly the problem that our platform Knownwell solves. So if you’re interested in finding out more shameless plug, uh, sign up for our beta wait list at knownwell.com and David Mohan, uh, have a good lunch today. See you later.

David: It was fun. Thanks.

Courtney: Hey guys, Courtney here. After listening to this episode, I went to our team and said, guys, we gotta figure out a way to let our listeners get a peek at what we’re building here at Knownwell.

And so our team has opened up 10 spots for executives of professional service companies to get an inside look at what we’re building. We’re really [00:19:00] excited about it, and we’ve talked a lot about it on this episode. So if you’d like to actually see what the platform looks like, how it’s working, how it’s gonna change the game for executives, go to knownwell.com slash preview to sign up for your spot.

Courtney: Noah Curran is the CEO of Monkedia, a digital marketing agency that excels at helping their customers win new clients. PB sat down with Noah recently to talk about AI’s role in winning and keeping new customers.

Pete: No, it’s so great to have you on the AI Knowhow podcast. Thank you for being here.

Noah: Thanks for having me. Excited to get to chat.

Pete: So we can give our listeners a little bit of context for the conversation. Uh, can you give us some background on Monkedia?

Noah: Yeah, absolutely. We started about 10 years ago with the goal of, uh, trying to be the best at helping people grow their businesses as quickly as possible, but as profitably as possible at the same time. And doing that all through, uh, [00:20:00] digital advertising. And so, um, very early on we decided we were gonna do all of that through artificial intelligence and have been building AI for probably nine years now.

That gives us a big leg up within our industry.

Pete: Um, you refer to yourself as an anti agency, can you explain why that is?

Noah: Yeah, I mean, honestly, there’s a lot of, uh, a lot of different agencies out there. People were kind of done with the whole agency model. You can get a in their college dorm room, claiming to be an agency. You can get a, a, you know, long term agency that’s trying to sell you marble, Marlboro ads, um, and concept boards.

Right? So there’s not really any like definition of what agency means anymore. And we think of ourselves as an anti agency because we really are more of a partner. We’re like part of your extended team, and we’re invested in helping you grow. Even with how we charge our, our clients is based on a percentage of their growth.

So, um, everything we do is kind of the opposite of the old agency model.

Pete: If I may, I’d like to reflect [00:21:00] back on the reference you made about using AI for some time now. On the website, um, I see a citation, AI and machine learning, marketing stack. Can you tell us a little bit about how you’re using AI, uh, to create value for customers?

Noah: Yeah. Well, so that started about, nine or 10 years ago. Um, I had a background in programming, um, but had gotten passionate about consumer psychology and the creative side of things. And so, um, at the time I was helping people consult on just business growth through process and training when digital advertising came out and started to tinker with that a little bit and found myself in the middle of like a perfect triangle of, um, different aspects, the creative, the, the data and the targeting and the media buying side, as well as that growth piece. And so very quickly on, on we, I determined that you could, um, with very, very small pieces of data, be able to model out, What targets were going to be best for a particular campaign, what creatives were gonna be best for a, know, particular campaign, and then at [00:22:00] what spend levels are ideal for profitability of each single one and move money in real time. And so we started building it through more of a, a mechanism for, um, as well as insight and data.

And then as we got more and more data we run and hundreds of millions of dollars of advertising through this, we were able to apply, you know, artificial intelligence and machine learning so that the system could actually learn and say, what are the most important factors to predict success?

What levers do we need to pull? How do we pull those? When do we pull those? Um, and the technology continues to develop. And so now we’ve developed it not only into just the media buying sphere. But into actually how it develops creative and iterates onto the creative side to then also how a data, how a, how a business understands the data, um, inside of their business so that they can make more actionable decisions as well.

Pete: Could you, um, play back for us a, a case study, a customer example, where you’ve used AI to create value?

Noah: Well, every single client we use, uh, leverages AI. Yeah.

I would say [00:23:00] that, you know, a great example would be Greenlight, which is a debit card for kids. We came, they came to us with like less than 50 customers. They’re, today, they’re worth about, I think five or $6 billion, um, built off of our, our marketing, um, methodology.

And what had happened was they, um, released their, um, initial like promotional, like, Hey, here’s our new debit card to a specific consumer group. Um, within their geographic area, which probably didn’t really represent was the actual type of consumer, um, that would be buying this. And they had struggled to get any lift.

It was costing them lots of money to try and acquire a customer. We went out, the artificial intelligence built out, you know, a hundred different. Potential targets. Spent a little bit of money on each one. We’re talking dollars, and then very quickly we’re able to hone in onto what is the exact right, targeting to be able to move the product and then model after.

And then it became the snowball effect. So the more customers we were able to acquire because we were [00:24:00] targeting better, the better the targeting got. Uh, and then did the same thing in looking at each step of the purchase. Uh, process for a parent to be able to optimize. ’cause it was, at first, it was like a 16-step process from some when somebody clicked on an ad to actually signing up for it.

And we were able to move that down to seven or eight steps, um, because of how the artificial intelligence identifies, okay, where’s the drop off points? And then how can we clean those up and, and understand what that customer journey looks like.

Pete: I guess, does the old John Wannamaker. Saying about, uh, 50% of our advertising spend is wasted. I just don’t know if, what 50% is that out the window nowadays?

Noah: Uh, yeah, I mean, very much is if, if we don’t know, you know, I always compare shotgun versus sniper, right? Like Yeah. billboards, shotgun. You’re hopefully getting 1% of your audience. Um, our, our technology literally ranks every single person in the United States on their likelihood to convert, and you spend your first dollar on the most likely to convert.

I always tell this like, I, I give this people challenge. I say, [00:25:00] if I was to walk in with you and to. And two of your friends into a Walmart and said, the first person who gets somebody to buy something, I’ll give you a million dollars. What would you do? Right. Answer that in your head for a second. What would you do?

Um, and very often people don’t say, well, I would go to somebody checking out and just help them scan the item and check out and be done. Well, that’s where our first advertising dollars need to be spent. Right? Somebody adds a product to the cart but doesn’t actually purchase. That’s person number one, go get them.

And then you rank through all the way through from people that have had interaction with you to the not and model data about who they are, what their, you know, interests are, what their behaviors are. And you can literally model everybody out. And so you are just spending down that line inefficiency, um, as you go.

Pete: Your process is heavily data and science intensive. Is there much of a burden for customers to ready the data on their side before being able to engage you?

Noah: No, I mean we, we probably 30 to 40% of our, [00:26:00] our clients don’t have any of that data when they start. Um, you know, it can cost a little bit more ’cause you have to do some exploratory, things out of the gate. Right? You gotta put. $20 into a hundred different targets, start finding out what’s working.

And it depends upon, okay, well if I’m trying to sell a $5,000 widget, it may take a little bit more than $15. Or if I’m trying to sell a book, maybe it takes less. But, um, you know, there’s some, there’s a little bit more of, you know, the investigation that has to happen and testing that has to happen than if seasoned business of 20 years, it’s been doing this forever, but not, it’s not.

You think about the olden days of like how much work it looked like and how much it costs to do, you know, a group of people coming into a random room, testing your product out, giving you feedback. I mean, we can do that in seconds with pennies now.

Pete: Much of our conversation up until here has been about acquiring customers. Uh, is there a tie from customer acquisition to customer retention?

Noah: Oh, absolutely. I, I think there are, um, two that just come off top of mind. First is, however you acquire a customer is more likely gonna have a [00:27:00] direct impact on what the retention is going to be like for that customer. So if you’re, um, if they have a really good. Process out of the gate, um, you’re likely to retain them better.

And so we drip content, um, over time to people to help them understand what it is. Most people are only gonna get like one to two small ideas out of a singular ad. Well, there’s a lot of stuff you might want to communicate. Um, and so we do that over time and that allows the acquisition process to be far more engaging for people, far more of the storytelling they have.

Far greater desire for a product. And then at the same token, um, we run specific ads to people that are in that retention bucket, right? They maybe they purchased, you know, six months ago and it your time to try and get them to purchase again. And so we can bucket those with their own individual messaging and make that messaging for them versus hoping they just fall back into the bucket of every, what everybody else is getting.

And, and that’s not the most effective message to retain a customer and, and get them to continue [00:28:00] working.

Pete: And I guess same approach when it comes to upselling as retention or as to acquisition.

Noah: Absolutely same Yeah. approach.

Pete: You’ve been at it for at least with AI for nine years. Um, many of our listeners are executives in firms that are kind of trying to figure out for the first time the role that AI plays in their business and what the big opportunities and threats are. What advice would you give to those executives for how to get started?

Noah: I would say really clearly identifying the problem you wanna solve. When we iterated our artificial intelligence over the years based upon a really lengthy discussion of what is our greatest priority. And the reality of it is, is AI can solve just about any problem you want, but unless you really set out and say, this is the exact thing that I want to go after, you know, you’re, you’re not gonna be much success and you’re, it’s gonna be innocuous and it’s not gonna [00:29:00] train the models, right? And all those kind of things. So set out one exact problem that you wanna do, just, you know, stair step it, say, this is the problem I wanna solve. And then just take everyday iterative, execution against that and start seeing how it will impact your business.

And that will snowball. Um, and you’ll see, you’ll start to say, you know, like when we, early on we made one change, uh, in direction on our program and we were. Able to save our, our salaries, uh, and overhead costs by 65% because it eliminated the need for a lot of manual work. Um, and so, you know, you start taking that savings, you reinvest into the next thing, and then it just keeps going.

So, um, and it will create that snowball effect in your business, but you gotta be really honed in on what is that first thing you wanna try and solve.

Pete: No, it’s been a pleasure. Thank you so much for joining us.

Noah: I appreciate it. It was great. Uh, great to get to talk

Courtney: Thanks as always, for listening and watching. Don’t forget to [00:30:00] give us a review or to press the subscribe button if you’re over here on YouTube. It really helps grow our show. At the end of each episode, we’d like to get one of our AI friends to weigh in on the topic at hand.

So, hey Claude. Welcome back to the show. This episode we’re talking about using AI to understand why you’re losing clients. What do you think?

Using AI to analyze customer data and feedback could shine a light on pain points or issues that are driving clients away. With the right models, you might uncover some surprising insights into what’s making customers jump ship.

Courtney: and now you’re in the know. Thanks as always for listening and watching. We’ll be back next week with more headlines, round table discussions and interviews with AI experts. Hey, if you made it this far in the episode, first of all, thank you. And second, we know that AI is changing all the time, and so we’ve got you covered. Here is another great episode for you to watch, to stay in the loop on what’s happening in AI [00:31:00] and business, and second of all, we’d love for you to subscribe.

You can do that right here. I’ve always wanted to do that, by the way, so do that now. That would be great.