What is coaching like at your bank? More importantly, what should it be like at your bank?
Banks that are tapping into the true potential of coaching are doing much more than giving pep talks, or one-on-one mentoring sessions.
PrecisionLender CEO Carl Ryden explains what transformational coaching at banks looks like in this week's podcast.
Helpful Links
Carl Ryden on LinkedIn and Twitter.
How Coaching Networks Will Create the First Facebook Scale Enterprise
What Can Andi Do for Your Bank?
Applied Banking Insights: Gather. Analyze. Act. Improve
Accenture Workforce Banking Survey: Realizing the Full Value of AI.
History of the Jack Russell Terrier
Podcast Transcription
Jim Young: Hi, and welcome to The Purposeful Banker. The podcast brought to you by Precision Lender, where we discuss the big topics on the minds of today's best bankers. I'm your host Jim Young, Director of Communications at Precision Lender.
I'm joined today by Precision Lender CEO, Carl Ryden. Most of you know that when Carl comes on the podcast, he's not here to talk about CECIL or Libor, but rather some of the big picture ways of thinking about banking.
And that's the case again today as we're going to talk about coaching at banking. What that means, where it can provide an impact, and what role does tech play in it?
So, Carl, let's start with what we mean when we say coaching at banks. Back when we wrote Earn It, we included a speech that Al Pacino gave during Any Given Sunday. Is that what we're talking about here? Motivational pep talks, that thing?
Carl Ryden: Well, there's some of that. Of course, it all starts with folks being motivated and energized and having a sense of purpose of what they're trying to achieve. We of course, talk about that a lot with The Purposeful Banker Podcast with BankOnPurpose our conference.
But once you have those things in place, it really boils down to what are we trying to achieve against that purpose, how are we measuring ourselves? And a lot of banks focus on reporting, looking at arrears, where we've been versus our goals, those things, but I think the key part of coaching is not just seeing where you've been and what you've done, also, seeing where you're going. But then once you see that report, it boils down to one simple question for me is what are you going to do differently Monday morning once you know that, right? What are we going to do differently?
So we can look at all the stats. Going back to the football metaphor, we can look at all the stats of the last game and we didn't get enough yards per rush. We fumbled the ball, and ultimately, you can't just say, score more points, give up fewer points and don't turn the ball over. Right? That's not coaching. That's just obvious, right? That is a slightly different restatement of the goals.
What you need to say is, "What are we going to do different? How are we going to prepare differently this week? What are we going to change in our behaviors and how are we going to remind ourselves that we're going to do things differently?" And provide that coaching back into the moment to help change behaviors and change what you're doing.
One of my favorite quotes, I think it's true, I think it is largely true, is, "Statistically speaking, the world's most accurate weather forecast is tomorrow will be just like today." Habits occur. Habits have to be built. Habits have to be changed. If you look at a report and you're not getting what you want, you've got to say, what behaviors need to change? What do we need to change about what we do? That's where think coaching comes in.
I think tech actually is playing a larger and larger role in this as it evolves and we can talk a little about the history of how tech, particularly information technology has been used within institutions and banking as well.
Jim Young: You talk about this type of coaching and that not just the here's what happened and go out and close more deals, but much more of a here's our approach and what we need to change, but then the question becomes, and we've seen this in banks, there are some sales managers who are really great personal coaches. They know how to intuitively do this, but how do you take that at a bank and scale that thing? And that's, I would say, is where is tech can come into play. Can you talk a little bit about that and I guess also why the phrase network gets used rather than just maybe tool?
Carl Ryden: Well, I'll start with what do we see happening within banks? And this is not just true of banks. Lots of companies, but we deal with banks. We see pockets of improvement, pockets of change, pockets of agility where a sales manager sees that his market is changing. Salesmen, regional manager, regional vice president sees that their market is changing, understands that their strategy is a little bit unique or a little bit different for the market that they serve. They're in a growth market versus a market where they have dominance or other things. And then they would actually coach their RMs to make sure that they're aligned with not generic goals, but around where they're trying to take the portfolio, what are the strategic objectives of the bank, and they would coach them on behaviors they would coach them on, here's what we ought to do differently, here's what we need to change.
We would see that and we would see the difference in performance between one group and another based on that within PrecisionLender. And what we would do is try to disseminate that. And at PrecisionLender, a lot of stuff we do, we do people, then process, then product. So we started out with great people who interact with our clients and help them get better. Then we actually started seeing stuff like this and we develop a process of saying, "Okay, last time we saw a bank like you with this strategic needs, here's what we saw really work. Tell your folks to do this."
And then once we see that has an impact and we do that through our client success reviews, once we see that has an impact, then we'll go to the product stage. Okay, how do we scale this wider? How do we make it so that people are infinitely flexible and really malleable and smart, but they don't scale really well? Process scales across people really well and the product ultimately scales really immensely.
That's what we introduced with Andi and as we started going down that route with Andi and the coaching and I think I've told the story behind Andi, and it might be worth repeating here. You tell me if is. But it came from that where somebody at a bank, Andy Max of First National Bank of Omaha was saying, "Hey Carl, could you put a dot on the screen an nudge them to do this when they see that? When you see a lender like business pricing a deal like that to a customer like that, nudge them this way, remind them this is what we want to do differently."
And those little nudges add up and there's actually a lot of behavioral economics and things around, there's a book called Nudge, which was about the cumulative effect of these little nudges and they started adding up to real impact, so we try to take that and build it into the product delivered at scale.
Jim Young: So then I'm going to go down the two letters here that are both powerful and sometimes misleading and go ahead and say that these are our AI related questions and I guess that's the first one.
Is this concept of a coaching network and that thing, is this fundamentally a artificial intelligence concept and then is this about automation taking this and doing this so that humans and replacing humans or is this about augmentation?
Carl Ryden: No, it's about augmentation. I mean, we've covered this before on reference other podcasts. We've done this thing. But it really is about intelligence augmentation and it's not necessarily about some artificial. I mean, I always say nobody really wants artificial intelligence. They just want intelligence. And if I have to go artificial and that's the most effective and efficient way of doing it, I'll do that because now there's a wide range of tools and the data that allows me to do that. There's some problems that suit itself to AI and more machine learning type answers, estimating the usage on a line of credit or mathematical problems, but there are other things that actually simple heuristics, the coaching that the best sales managers provide their folks given strategy. It just putting a heuristics, the rules based coaching that you put in there.
The key part about that is when you put it in the rules based coaching is, by the way, is no different whether you do it with tech or you do it with humans, right? Is to know that we're putting in this rule because this is our strategy for where we are today and here's the outcomes we're hoping to achieve. And then you can actually measure that the outcomes are what you thought they were and know when to take it out because that strategy is no longer in effect, right? Because the world does change.
So being able to automate the delivery of coaching and customize it is a huge trend. And that's where Gordon Ritter, who's a friend, who runs a company called Emergence Capital Partners comes in, I've talked about him in some talks I've given recently, and we talk a lot, and he comes up with this concept called coaching networks, so it's worth given the history of Gordon in his view on the world.
Jim Young: Gordon was next up on my question. Go for it.
Carl Ryden: And then, I think really like to come back to and we'll see if it makes the final cut here is the history of IT within banking and how that's evolved within orgs. I think there's two parallel paths here that actually line up and point to the same place. I mean, they run quite a bit.
So back to Gordon. Gordon runs a venture capital firm out in San Francisco called Emergence Capital Partners, and it's a smaller firm for Silicon Valley standards and it's very thesis driven. So they actually forced themselves to have the discipline of what's the thesis, what do we believe is going to drive the next wave of innovation, the next wave of impact and value in the world?
So 20 years ago, which seems like forever ago, his thesis was enterprise SaaS software, software as a service, cloud based solutions, was going to take over the enterprise. And that was 20 years ago. That was not as obvious as it is today, and his first investment he put a million dollars into a little company called Salesforce. He was the first million dollars into that, and that turned out really, really well. It turned into quite the juggernaut now that most banks and others have found their way to adopting.
His next thesis 10 years later was that the first wave of software as a service of SaaS software, was horizontal in nature. Horizontal meaning Salesforce is a CRM software, but it serves a wide array of industries. It's a horizontal function that is applied to many different vertical industries. And you saw tools like Marketo, Netsuite, other things. Same thing. Horizontal functionality that was in customized and applied to a vertical.
His next thesis was that the next wave of SaaS software was going to be vertical and it was gonna be bigger than the first, in that the specific domain knowledge about a vertical allows you to deliver a far better solution to that vertical and the adoption of cloud solutions has got to the point now where it starts to verticalize. The first investment you made there was a company called Veeva. He put $6,000,000 in to own about a third. That company never raised any more money and went public about 8 years later and now they're worth $15,000,000,000. It's a vertical SaaS solution fo, the life science industry and has been a very successful company. And Peter Gassner, the CEO, we've talked to him quite a bit as well.
So his first thesis was enterprise SaaS going to take over. The first wave was horizontal. Second wave is going to be vertical. His most recent thesis that he's come out with is what you're seeing now is that these forms based applications, the fill out of form, hit next, hit next, hit next, hit next, put the data in the database, then get a report back from that and then try to adjust course based on that. He thinks those are going to go away and he says now where we're getting to with AI machine learning and cloud and large data, is that the machine, as a human, you do what you do. The machine sees what you do, gathers that information, then compares it to what people like you do and who achieve better outcomes, and then coaches you on how you could do it better.
He takes it even one step further, he says the machine sees what you do and it goes and fills out that system of record for you. And so one of the examples, which is one of his portfolio companies, which I'm actually talking to the CEO of it next week, I think, or maybe even today, if a schedule separate, but the company called Chorus.ai, which I think is when I was with a group of 200 commercial bankers at their sales kickoff meeting a couple weeks ago. I mentioned this and I told them the story. It was funny when I told them the story, how they just all lit up and they were like, "God, that's what I need."
And I told them the story about this idea of the flywheel. Start with the bankers. You build them a system that allows them to do what they do, have empathy and judgment and creativity and trust and the things you really need a human to do. The machine then sees what they do, gathers what they do, fills out the systems of record for them. Does what the machine should do, then compares what they did versus other folks grounded to the outcomes that they achieved and coach them on what they could have done better.
I gave them the story about as an example of that, one is Precision Lender was they were all familiar with. The second one is a company called Chorus.ai. And chorus.ai is one of Gordon's companies and what it does is on sales calls, the salesperson has the call with the customer they're going to have it with, chorus.ai listens in on the sales call, records it, uses natural language understanding machine learning to extract out what will said. It takes that, fills out Salesforce.com for the user, captures the next steps, what was discussed, those things, the notes for the call schedules, the follow ups in the to do's into Salesforce for them. Then compares what they said versus what other at Salesforce said, and coaches them on what they could have done differently.
This is Gordon's idea of like if you think about this, the machine's doing everything, then the machine should do well really well so that the humans can get better at what they do well. And in this case, Gordon who runs ultra marathons, he says, "I don't run a mile and then go fill out a form based application, hit submit and get a report on how I did the last mile and then run the next mile. I just run and I have a Fit Bit in my ear. It knows my pace. It knows my cadence. It knows my goals. It knows what I can achieve. It knows the terrain I'm running on and it says, here's a place where you need to pick up your pace, where you can gain some time. Here's a place where you need to slow down, you're going to burn out. You're not gonna make it to the end. I do what I do to achieve what I'm trying to achieve. It knows my goals, it knows what I'm doing, and then it provides that feedback."
And this is what he calls the coaching network. He thinks the world is going to be dominated by these coaching network companies. So back to his original thesis, the first wave of SaaS software in the enterprise was horizontal. The next day wave vertical and it was bigger than the first. The next wave we're in now is, this is one he and I talk about, is personal. First it was about the function, the horizontal function. You're doing sales. You get to CRM. The next one's about the industry. We understand you're in the life sciences or you're the banking industry, so we can bring that domain knowledge to bear. The next one is just like in Precision Lender, for a customer like this to when a deal like this with a banker like this, at a bank like this, with a strategy like this, here's what you ought to do differently. Here's what you ought to talk to them about. Here's what the best bankers face with this situation, the ones that produce the best outcome, here's what they did.
And that coaching network feedback loop is incredibly powerful because it really allows you to learn as an organization at scale and get better and better and better.
Jim Young: He mentioned a phrase that I thought was interesting and we will link to an article he has on Emergence Capital's site. It's called How Coaching Networks Will Create the First Facebook Scale Enterprise Solution. He talks about humans are the mutation engine and the evolving process of the coaching network. What does that mean exactly to you?
Carl Ryden: I know we have bankers, so I'm a math guy, so I might get a little bit math wonky, so you may edit this out.
Jim Young: You're saying bankers wouldn't like this or would?
Carl Ryden: We'll see. I'm just kidding.
In math and in machine learning and algorithms and in math in general, there's this idea of exploration versus exploitation. The exploitation is not the moral term of exploiting, mathematical exploitation. So for example, you have to explore what works and then once you figured out what works, you then exploit it, right?
Jim Young: Right.
Carl Ryden: And you see this just so example, one of the algorithms for this is called a banded algorithm, which is basically like the setup is, there's a bunch of slot machines and some of them are better than others, so you have to go pull all the arms on the machines to see which ones are paying better and then as you have a certain bankroll size and your goal is to maximize your winnings, you have to explore to find the best ones, but then you also have to exploit the ones that you know are paying. And that's a really interesting mathematical problem and a really difficult one.
But this algorithm occurs almost everywhere. So when you go and you search on Google or you go to ESPN and they decide what articles to put above the fold, they want to maximize their engagement with you at ESPN.
Jim Young: Right.
Carl Ryden: Right.
And so they want to give you articles they know you're going to read, however they want and sometimes inject one that they're not so sure about to explore, and this is called the exploration exploitation trade-off and there's a mathematical algorithm for that. And so we come back to this in terms of the Facebook-scale company in Gordon's, I don't even know what got use started on this actually.
Jim Young: I was asking you about humans as a mutation engine.
Carl Ryden: Oh yes. So humans are the exploration piece. So what we use is, is really hard for what computers are good at and what humans are good at are really two separate things and in fact, we're almost exactly complimentary. Everything that machines are really good at, we're bad at, like doing massive amounts of mathematical calculations and huge amounts of data and huge amounts of long-term and short term memory, we don't have that.
Jim Young: And doing process number 1000 the exact same as you did process number one.
Carl Ryden: It's very difficult for a person. Machines are really good at that. However, the judgment, the empathy, the creativity, the intuition, all the things that humans are really good at, that a three year old is really good at and that's one of the best ways to see what's good at, the things that a three year old really good at catching a ball, a computer machine is really bad at. Intuition, other things, empathy, they have that. The things that humans have to work 20 years to be good at, the things we actually teach you in school, statistics, it takes a lot to get a human to do that and there's only a few who can exert the discipline to be really good statisticians. But computers, they're all good at statistics. I mean, they can calculate that stuff constantly.
The idea of the coaching network is the lower half of that loop where it's the machine gather, compare, and coach, right?
Jim Young: Right.
Carl Ryden: That's the exploitation piece. That's the mechanical mathematical piece, but the top of that, the exploration in a uncertain world with other humans, or it's a human interaction role as is it can be in commercial lending, you need the humans to be that, he goes to the flicker in the flame there, the exploration engine, they're the piece that introduces that, finds these new ways to do it. They explore. Then the machine figures out, "Okay, Jim just found a really good way of approaching this and it really created a great outcome. I want to make sure we exploit that and every body like Jim who's doing a deal like Jim just did sees that."
So I think it's that exploration, exploitation and it's this hybrid of using the humans for the exploration and using the machine to determine the best path or exploitation, and we do this. This is not unusual. Amazon does this, Google does this, this is not crazy in the world.
Jim Young: Yeah, that's interesting. That does play off that part of, obviously with these networks you are crunching numbers and all that stuff, but you're also taking what the best performers are and in our world it's relationship managers and the best ones they're doing and taking that and applying that as additional coaching. It's not always a number, number, but it's also Bob over here is great at this thing and this is how he does it, so we should.
Carl Ryden: But by the way, that's what doing it with the relationship managers and what the behaviors they exhibit. Remember, the goal is what are the behaviors that drive better actions? So if I can see the behaviors and outcomes they create, then I can actually, when I see a similar person in a similar situation, I can coach them on the behaviors that tend to lead to better outcomes. Now, that piece right there is really not any different from Netflix. Netflix, they know people like you who've watched what you've watched, who've liked what you've liked, who've gone and binge watched all 47 episodes of Seinfeld, which you probably have or however many. There's 400 episodes of those. They're going to say people like you tend to watch this. Right?
Jim Young: Yep.
Carl Ryden: And what they're doing is they're looking at the behaviors you have and the outcomes that generated, which for them was engagement with those shows, engagement with that content. And then they say, "Well, people like you tend to engage with this stuff" and they feed that back. Amazon does the same thing when you go shopping on there and you look for a 52" television and then you look for a 60" Sony and a LG, and it says, "You're probably gonna buy the 54" Samsung, and when you do, you're going to buy these three cables with it to hook it up to your other thing." They can front and run that and say, "People like you who have exhibited these behaviors tend to reach this place, tend to go here?"
And you just front and run that. And the key thing is, in every one of those cases, the training data came from human behaviors, right? These AI systems that it's all machine, what good is that? Right?
Jim Young: Right.
Carl Ryden: I mean, the most important interaction for us is that human to human interaction. So how do we enable that to be better?
Jim Young: And I guess the one final thing I would ask about this though, is that these are suggestions and not necessarily commands, right? At the end of the day, it's still up to a human to decide whether they're going to take that information and apply it or whether they might have some judgment that tells them maybe not.
Carl Ryden: Oh, absolutely. And that also gives us two bites at the learning apple, right? The first one is, well, okay, what have other folks done, right? For other bankers at the bank who are looking at a deal like this with a customer like this, and an industry like this, with a relationship like this, here's what they did. But then I offer you suggestions, hey, you might want to think about this. Either you take that suggestion or don't. Either you click on it or you don't, either you act on it or you don't.
That gives us another bite in the learning apple, which gives us, for a banker like you, and by the way, this is no different than the Amazon thing. Amazon takes the area data and says,
"Okay, here's some suggestions for you Jim." And then you either click on them or you don't. If you click on they're like "Yeah, plus one, we got that one right. We're starting to understand Jim a little better and how to move Jim forward."
Google does the same thing. Now there's social implications of this that we end up in a filter bubble where all those things, which is a whole different story, but on the enterprise side, within a bank, that's less of a concern.
Jim Young: Right, and like you said also, that's also where the explore element comes into it, right? Where you can get out of that danger to filter bubble by occasionally throwing in something that's different from them.
Carl Ryden: Correct. You have to add some diversity into the system.
This is a side note. I've have to Jack Russel terriers and I love Jack Russell terriers. And what's interesting about Jack Russell terriers, for the longest time, they weren't part of the AKC, the American Kennel Club. The AKC was built by Victorians where they would inbreed dogs, to get a particular look and you'd end up with a dog with a really particular look. But they were absolutely genetically a mess, right?
Jim Young: Oh yeah.
Carl Ryden: Like prone to disease, don't get me started on the Cocker Spaniels, I have friends like this, but it's like the dumbest dog in the world. But the Jack Russells for the longest time, they weren't bred for look, they were bred for function. They were meant to hunt foxes and rats and rodents. So they would intentionally introduce beagle blood, other things to drive behaviors. So they were bred for function. This is why Jack Russells, they all look a little bit different, but all of them have the same attitude and the same tenacity, and the same persistence.
I think this is the thing is if you end up with that feedback loop without introducing outside genetic material, you end up with a really dumb algorithm that's fraught with peril, it's very fragile. Introducing that diversity in there, that exploration phase, brings in that diversity that makes the algorithm not only more valuable but more robust.
Jim Young: Yeah, and I guess also, if you don't introduce that stuff, you end up really, really good at bank at selling a product that nobody actually wants anymore. Right? Because if you haven't looked at what's going on and how people are doing.
Carl Ryden: Well, there's that and then the other thing is you can end up with a really, I'll draw the metaphor, the Cocker Spaniel is a pretty dog, but not really smart. And because it's inbred is prone to disease and other problems and you'll see these really hyper bred dogs where they are really prone to ailments. An algorithm, like I said, not necessarily machine learning algorithm, it can be a simple heuristic algorithm, and those algorithms can be deployed via software, they can be deployed via humans. And so I would say the eight is great algorithm, which was the cross sale algorithm at Wells Fargo.
That algorithm, I would say the feedback loop from that algorithm, it wasn't inherently bad at the start, but the feedback loop amplified itself and it became inbred, which made it not only fragile but prone to disease. So that resiliency you want to put in here of is it still our strategy, is it still achieving the customer's objectives, is still fit for our purpose as a bank? Whether you do it with technology or whether you do it with a sales slogan, those are algorithms you're deploying into your organization. Being able to do that with tech gives you the advantage of being, one is it's more scalable. Two, it's more agile. Three, it's more flexible to change. And four is you can better connect it and two metrics to know and look at early warning signs. Is this still fit for purpose?
Jim Young: Yeah, yeah. This has to be the only commercial banking podcasts in which Jack Russells and Cocker Spaniel breeding plays into it. But it's a fascinating topic of coaching networks and definitely recommend you read the and we'll link to it again in the show notes, the article from Gordon Ritter and Emergence Capital. And this also, you probably heard us talking about it. There are concepts, and Carl mentioned it, applied banking insights and this flywheel. We'll link to those pieces of content here as well.
We've seen some interesting information from Accenture recently that a lot of frontline bankers are very much excited about this, they found a discordance where the sales managers assumed that frontline people would be freaked out about this stuff, the automation aspects of it and losing jobs, and rather the frontline people were saying, I really, really am excited about what these types of technology can do. And so I think the coaching network concept is ripe for really a lot of success in the commercial banking space.
Carl, unless you have any other dog breeding related insights here, I think we'll probably call it a wrap on this podcast. Thanks for coming on again.
Carl Ryden: Sure.
Jim Young: All right. That'll do it for this week's show.
Now for just a few friendly reminders, if you want to listen to more podcasts or check out more of our content, you can visit our resource page precisionlender.com, or you can head over to our home page to learn more about the company behind this content.
Finally, if you like what you've been hearing, make sure to subscribe to the feed in iTunes, SoundCloud, Google Play, or Stitcher. We love to get ratings and feedback on any of those platforms.
Until next time, this has been Jim Young for Carl Ryden, and you've been listening to The Purposeful Banker.
About the Author
Follow on Linkedin More Content by Jim Young
