How do you whittle the murky business of creating a data-driven culture down to a proven process? Today we talk to a guest who has done this time and time again, helping companies transform their operations. He points out the small nuances and details about the process, like questions to ask to start on the right foot, critical feedback loops to put in place along the way, and how to overcome some of the most common problems that make people give up.
Ginette: I’m Ginette.
Curtis: And I’m Curtis.
Ginette: And you are listening to Data Crunch.
Curtis: A podcast about how applied data science, machine learning, and artificial intelligence are changing the world.
Ginette: Data Crunch is produced by the Data Crunch Corporation, an analytics training and consulting company.
Now, let’s jump into our interview with Ryan Deeds, VP of technology and data management at Assurex Global.
Ginette Methot: How do you whittle the murky business of creating a data driven culture down to proven process? Today we talk to a guest who has done this time and time again helping companies transform their operations. He points out the small nuances and details about the process, like questions to ask to start on the right foot, critical feedback loops to put in place along the way and how to overcome some of the most common problems that make people give up. I’m Ginette and I’m Curtis and you are listening to data crunch, a podcast about how applied data science, machine learning and artificial intelligence are changing the world, a vault analytics production. Let’s jump into our interview with Ryan deeds that VP of technology and data management at Assurex global.
Ryan Deeds: Uh, I think it’s an interesting time in the whole a data experience because I think so many people failed. You know, in the last like decade that this next couple of years everybody’s now trying to look at root cause. And so culture actually is becoming important now, you know? And so that’s kind of a cool thing.
Curtis Seare: What do you mean by that? In terms of a lot of people have failed.
Ryan Deeds: I think when you look at bi projects from 2003 to 2013, they were just, companies went through litany of failures and trying to get data to a place that what made sense was easily accessible, had had a good quality. Um, but they didn’t address that. They just put the visualizations on top of kind of crappy data and they did that over and over and over again. Um, and then finally it seems like, you know, in the last year or two years, we start really having a conversation about what has to happen inside an organization to make data usable. I mean, it’s just like water, right? You can’t just take water from a stream and start drinking it. You got to process it and clean it and make it and make it valuable and make it worthy of consumption. And that’s exactly the thing we got to do with data.
Curtis Seare: Sure. Maybe we can dive into that as well, because you’ve had this experience taking a lot of companies through those steps, right? So what do you see as the major roadblocks? How do you start this process of helping people get their hands around? How do I get value from my data?
Ryan Deeds: So it’s interesting. I kind of have, uh, you know, I’ve done this a lot and so I have, uh, organizations that come to me and they say, hey, you know, we want to, we were ready to start leveraging data. Um, and the, the typical thing is there’s just a lack of expectation of the time it takes. Um, and so I threw together like a timeline to try to help, uh, educate individuals on that, you know, and kind of like the steps that it would take to get to usable data, um, in, and the first is really a recognition that today we don’t, you know, the organization that we’re in is not effectively using data, um, as a, as a strategic advantage. And generally the way that I help companies understand that is how long does it take to, to prepare your reporting? Um, and that, you know, if they’re like, hey, it takes three days, well that’s three days of a human taking data out of a system in massaging it and manipulating it, using their intuition and their learned the learned bad data practices in the organization and correcting that and that one human and then distributing out a report.
Ryan Deeds: And so oftentimes that’s, that’s kind of the first thing I look at, um, to help educate them that they’re not effectively using data today as a, as a strategic advantage. You know, cause I think data is very different than reporting data as a communication mechanism that allows, or top level we’re going to top level leaders in an organization define strategy and then each individual in the organization leverages their component of the work to push that forward. And you use data to educate those individuals on what they should be doing to push that needle for them and then incentivize or deincentivize behaviors based on that. I mean, that’s what I like to see. Um, from a data strategy standpoint, unfortunately it’s typically not the way it is. I mean there’s a, it’s a long leap from I, we’re not using data to what I just described and it takes a lot of time and energy in the right people. And um, so often they just give up on it because the, to get the culture right is, is very challenging. It’s hard.
Curtis Seare: That’s interesting. Are there things that you do to help people get through the dip and soldier on to getting to where data can help them? Or does this have to come internally?
Ryan Deeds: No, I mean, I think most, most companies today, especially small to medium sized businesses, they understand that they are at a disadvantage. Um, from an operational standpoint. They don’t have the staffing, um, to leverage data in the past effectively. And now they’re like, oh my God, we’re behind the eight ball. We really need to do this. The challenge is they don’t know exactly how to go about it. And in my, in my industry, there’s some specific things that I look for in organizations to determine if, if data’s going to work for them. Um, and a lot of times that’s how many young owners do you have in an organization? You know, how many agents of change are, are there? What’s the, is the culture one of empowerment from the bottom up or is it more I dictate rules from the top down. Um, and trying to educate them on where they are today with their culture and where they need to be to get to a place to get good data.
Ryan Deeds: Because in my opinion, data projects typically fail because they don’t give feedback to the people that enter them instantly on what data elements are needed to be correct. Um, and so I like to see dashboards with a feedback loop to the users that are entering the data. So once they entered date, so say that you’ve identified seven fields, these are the seven critical fields that we need to fill while you bounce those off of normal ranges. And if they come outside the normal range, you send something back to the user on a dashboard or email or however you’re going to do it and say, hey, this was incorrect. You have three days to to correct this or it gets escalated to your manager. Um, and generally when you give them that kind of guidance to get the stuff correct, it works pretty effectively. I mean, the employees want to do the right thing. They just oftentimes don’t know how to,
Curtis Seare: you’re talking about operational data or maybe something you would type into salesforce, something like that. This kind of information, right?
Ryan Deeds: That’s right. Whatever system you’re using today to, to quantify whatever you’re trying to metric out, that is, that’s kind of what you’re looking at. I mean, right now I would think that a lot of people have CRMs and maybe they don’t use and maybe the company doesn’t think they use the CRM effectively. Well, my question to them would be how many people, how many times don’t they pay on an account that’s not in the CRM? Um, because that will, that will absolutely drive behaviors. I mean, we have to talk about the behaviors of individuals that are in there. I mean, that’s kind of what we have to continue to look at and help drive the right behaviors using analytics. But it’s, it’s oftentimes difficult cause I mean, you’ve got to go into an organization, you’ve got to figure out what metrics matter, then you have to figure out how off those metrics are.
Ryan Deeds: You’ve got to put in a framework to get those metrics correct. Uh, you know, and it, it takes two years. I mean, when I come in with a data management strategy, that’s generally what I tell people is like, look, this is going to be a two year slog. As we identify the main data elements, what they mean, make sure that everybody’s on board with that because you can’t judge somebody on a metric that they don’t believe in. Um, and so a lot of the stuff that I do is trying to build consensus around metrics that actually matter to the, the kind of the end user that’s being judged with those metrics. Be that a salesperson or a CSR or you know, an operations manager. What are the things that are driving their success and what, what should they be watching to determine if they’re successful? And those are, and that’s the hard stuff. I mean, people don’t want to do that stuff, you know?
Curtis Seare: So it sounds like when you’re choosing these metrics, it’s really focused on what’s making the people successful. And that’s what you try to measure. Is that right
Ryan Deeds: for this kind of internal dash boarding? For sure. I mean, if you’re doing client facing dashboards, you’re going to be leveraging a different analytic. But I always liked to get operational analytics done first because it gives the organization, um, the ability to start kind of thinking about how they’re using it and then they can apply that to their client’s problems. And a lot of cases, you know, w conceptually they can take how they’re leveraging analytics and kind of put that into a use case for their clients and then we can deliver that to help push the business advantage forward. I mean, um, and I always like to get my scale in place before I get my growth in place, you know, so I use operational analytics to ensure that we all have an understanding of the, I spend two years, you know, doing data management typically.
Ryan Deeds: And then we do two years of internal efficiency building and the metrics that we’ve chosen in the first two years on our data management, we should see some kind of corresponding row reaction based on technological efficiencies or outsource deficiencies that we bring in. You know, as an example, say that you had a metric that measured how many things got attached to a document management system. So you’re in the financial services industry, you’re a lawyer. You know when paperwork comes in, you have to attach that to some system. Well, if you bring in an email archiver and you say, okay, now only these four types of documents have to be attached. You should be able to see a correlating metric reduce in the amount of stuff that’s being attached to the system. Um, because the archivers in place, if you don’t see that over time, then there’s a noncompliance issue happening. Or the individuals aren’t netting the efficient. They’re not seeing the value of the change that was made. And so I always like to make sure that I have, if I’m bringing technological solution in, what’s the result of that and for a success look like what’s a result of that for failure look like? And what’s the timeline?
Curtis Seare: What do you typically see? I mean, so this is a big investment, right? Two years to get stuff up and running. Two more years to see what the result is. When you put this in place, is this typically transformational for the business? I mean, are you seeing like a 10 x 100 x increase in output?
Ryan Deeds: I mean you, you, it is absolutely transformational because it’s, the larger organizations have, have already started to um, gathered data in mass in relational ways that they can start to leverage. We’re still at the place. A lot of the small and medium sized businesses are still at the place where they’re recognizing how they are not using data. So once they can start going that direction and be data driven, it totally changes. I saw my organization, this is a very small example, but we were 14 million at 100 employees and then we were 23 million at a 100 employees and a lot. And that was over a five year period of time. That was, you know, two years of analytics, two years of efficiency building. Because that two years of efficiency building is really your innovation engine. You’re actually taking problems from the staff, you’re building solution, you’re applying those to the thing and you’re watching the metrics that you, that you’ve assigned to that, um, uh, increase or decrease.
Ryan Deeds: When you come out of that two years, now you’ve got a forward thinking, engaged staff that understands how the things that they bring to the fruition change the needle, pull folks want to be. They want to work smarter. You know, they know the tools that they have at their disposal. They understand how it’s supposed to shift stuff. So I believe it’s a total cultural transformation from one to the other. Um, and while it does take a lot of time, when you look at the next, you know, the next four or five years and your journey, it’s so much more effective because a lot of companies get that, um, they get changed weary and because we throw a change, we expect some result. We don’t give time for implementation or adoption. We don’t really have a metric associated with it. A year later we don’t, you know, we churn the software and we go by something else that that’s supposed to do a better job because we’re not looking at the root cause of the issue, which is our inability to associate that with driving goals forward.
Ryan Deeds: Um, and so no, I think that again, it kind of has to be a perfect storm in each organization as they make that transformation. Because like you said, it’s a lot of work and you lose a lot of staff. I mean, once you bring, because analytics brings a tremendous amount of transparency if they’re done correctly. So not only can, you know, operations leader, look at top line stuff like retention or call volume, they can also, you know, really drill down into the details and very rapidly to out in the slugs, you know, that are in the organization. And so I think we, we saw about 15 individuals leave over the three or four years that we had brought the analytic stack in once we were, once we had delivered the dashboards. And that’s how the managers are running stuff.
Curtis Seare: That’s really interesting. And you’ve talked about this concept of an analytics maturity model, right? And I’m assuming during this two to four years here, you’re taking people through the steps of this maturity model to get to where they need to be. Can you define that for us and help us understand the steps?
Ryan Deeds: Sure. I mean, you know, and it comes from questions, you know, a lot of, a lot of understanding where they are analytically cause I get brought in to do analytics a lot. But, um, I always start with a top to bottom interview session with the firm. You know, I, I want to talk to the, uh, executives. I want to talk to the managers. I want to talk to the line staff. I needed determine, you know, what the alignment is in the problems that each places seeing. Um, and oftentimes, you know, in a, in a non analytically mature organization, what the leaders see as the root problem and what the staff sees as the root problem are extremely different. And, and many times not even on the same plane, you know, leaders think they have unmotivated staff that are, are, are not buying into their messaging.
Ryan Deeds: Staff believes that they have leaders that they’re, they’re just coming in and they turn on a fire hose, they sit down and they shut it off and they’re, they’re not able to communicate on any normal level. You know, they’re so disconnected and in that case, nothing, they’ve got to at least get some kind of common ground before you bring technology in. And a lot of times data can bring them to that as they discuss data elements that matter to them. You know, the, the, if it’s a CSR then it may be call volume. Um, if it’s, uh, you know, an insurance agent and maybe book of book of whatever the metric is that’s driving those business units. You have to have these discussions around the table with all parties. And so that’s, that’s kind of the first step is have we done, what are the things in the organization that actually matter?
Ryan Deeds: How many different ways do you call revenue? If I come into an organization, the CFO speaks in a very different language than the account managers and the CEO. It’s a big problem, you know. So if they’re talking about, uh, collected revenue, estimated revenue booked rev, if they’ve got all these different words for money, then everybody kind of sees things differently. They haven’t had these hard conversations to create business definition behind these individual analytics, which then, you know, portends to a lack of analytical ability. Each person is using their own reports for their own devices and those reports may not be interlinked. Um, and in the industry that I’m in, there’s a couple of specific analytical metrics that I use all the time to determine, you know, how, where they are, even if they recognize those metrics exist and a lot of times they don’t. And so, you know, but as we move along that journey of analytical maturity, that means that now we’ve, you know, if you’re, so as I say, it’s a kind of a five step thing, you know, and Mico what kind of one of my mentors she, she has this Mico Yuk Bi Brainz great lady.
Ryan Deeds: She kinda has this, you know, information, data knowledge and wisdom, kind of look at it. Um, I think of it more from just from the insurance agencies first perspective cause I, that’s where I’ve been stuck forever. It’s, you know, I’m not using data. I recognize that, that we’re going to start using data. We start normalizing the, the metrics within, and then we start building small proofs of concepts, um, in a data model. And I always push people to a centralized data model. Typically I leverage a specific Microsoft technology and that’s what I like to see my agency’s use because they’re so, it’s such a flat thing, but as they start to understand, all right, we have this one metric, this is how we’re going to monitor if it’s in compliance or not, this is how we’re going to get a back to the account manager.
Ryan Deeds: We understand why we need to do that. It trains. That’s the kind of stuff that just takes forever and that’s the, that’s that maturity model that they go along and I mean anybody that’s sitting on a ton of data today that consistently plays with it, you constantly find new ways to leverage that data for value. And so that’s why I think that the larger organizations have a pretty heavy advantage because they have more data and they have more people that are just in there playing with it, trying to figure out what might be cool to use.
Curtis Seare: Sure. That’s interesting. And so we’ve talked a lot about operational metrics, right? Getting your organization on the same page, getting these efficiencies down. Once you’re there, you’ve talked about, okay, now you have the ability to play with your data and find value in it. Things like that. What else are you seeing? Companies that are already analytically mature, what, what additional things can they do with data? Are they then do a client facing analytics are spinning off new products? What kind of things do you see?
Ryan Deeds: Yeah, I mean, so when I came into the last shop that I was at permanently, I came in with a five year plan, two years of data management, two years of efficiency building. And, and after that we would turn the client facing value. Um, and I think that that’s correct. I mean the internal stuff doesn’t stop but it’s almost got it’s own like motor now. So it keeps kind of going, you know, and there there’s an area of diminished return that you can just, you know, try to get hyper efficient and so, but once you’re there you kind of know, you’re like, Hey, we’ve done a lot of this. Are Our net profits looking good? Our revenue per employee is looking good. I mean key metrics that you’ve set up to determine efficiency are now within where they should be, um, in consistently increasing and you have your process down for that.
Ryan Deeds: Yeah. You turn towards how do I now make the business more successful? Um, and if that’s client value, then it’s obviously trying to determine, all right, you know, from an analytics perspective, what can we deliver that differentiates us from the person, the competitor down the street. You know, if I’m a, if I’m a call salesperson, if I, if I’m a, uh, like a call center person, um, and we’re outsourced our call center to this group, can I provide a dashboard back to, uh, our clients to show them, you know, the rate of calls and all this different stuff that may not be there. I mean, it’s really having the same conversation that you had internally, externally, it’s what’s value to that class of client? Is there an analytic play there? Can there be something that we differentiate? And it just depends on what segment of business you’re in.
Ryan Deeds: I think that nowadays though, everybody wants data. So it’s really you just being creative with the skills that your organization, cause I also think that a critical component to this is to do this internally. Don’t go by contractors, don’t go get outside people. Cause you lose the institutional knowledge of that. And that’s again, in a lot of cases why we are where we are. It’s because so many times we brought in a Deloitte come in for a data management program in place or some other data consultants and they do all the pain and then they leave and now we have the result of the work. But we don’t have the lessons learned through the work. And that drives me insane because I mean, we really need to understand the pain of data management so we can consistently do it after that, after that initial push goes. Um, but yeah, I mean I think that it could be, hey, we’re going to, you know, we really want to do a wellness campaign for our own employees. So we’re going to provide some kind of analytics to them about wellness and create competence. It’s the world’s kind of your oyster after that because now you know the process that you have to go through to deliver clean analytics to anyone else and you understand how challenging that can be.
Curtis Seare: And what is your thought on, you know, nowadays machine learning, artificial intelligence, huge buzzwords, slots, people talking about it. What’s your opinion on how people should approach that space?
Ryan Deeds: Man, I think that everybody that’s, that listens to this podcast hopefully understands what an inflection point is and um, and, and I hope they understand the goal and Alphago scenario that we saw a couple of years back because I think every industry that leverages AI is going to go through specific inflection points. It was like with the radiology one when the radiologist AI was able to identify cancer more effectively than humans. That’s an inflection point, right? And I think that every one of the technologies, because right now we’re really talking about this narrow AI. I hate the words that we use in tech because we do not do a good job. Like I think cloud set us back five years. Just the word. So Ai, AI, when you say AI to somebody who doesn’t know anything, who’s not very educated on it, they group that as as the same as everything.
Ryan Deeds: And I always think of it as this is a very narrow focused algorithm that can handle, that can use historical data to handle specific problems. Um, but I think each one of those algorithms get smarter with time and we just need to be looking for inflection points. And some of those are going to be legislative inflection points where we get more comfortable as a society with some of this. And some of it will be technology, technological inflection points that we just cannot, like, we’re like, oh my God, yeah, I’m not, I don’t want to go to a cancer per somebody who’s looking at me with cancer. Without that, I mean that I’m much prefer them to have that AI sitting behind them as a second set of eyes or a primary set eyes.
Curtis Seare: Yeah, I think that’s right on inflection points is a good way to look at it. Maybe for the last piece here, I’ll just open it up to you. Is there anything else that you’ve seen in your experience that you think would be valuable to share?
Ryan Deeds: I think when you look at where we’re headed, the next big technology that we’re going to see is augmented reality. I don’t know how it’s going to be delivered, but when you look at the success of HUD’s in cars and how it’s effective at delivering a multitude of information to a consumer through, uh, through ocular input, it’s so powerful. And so I’ve been playing a lot with, um, uh, magic leaps developer edition because I think that when you think about how much you can combine from with data to people today, that that’s, there’s just going to be a gigantic space there. So I think companies that are really trying to do cool stuff should be looking at how they’re, how they’re leveraging kind of a far looking, how are they going to be engaging employees, engaging clients, using augmented reality, leveraging their data. Uh, it, it’s just cool stuff. That’s our show for today and we, it’s been helpful and understanding
Ginette Methot: what you can do to drive change in your organization with data. As always. Our show notes are available@datacrunchpodcast.com and if you enjoyed the show, leaving us a review on iTunes or wherever you listened to, your podcast would mean a lot to our small team and for new listeners, if you’d like more in depth information about how to get value from your data, you can sign up for our newsletter on the website as well. If you’re starting a data visualization project or you’re stuck on one, had over to our website at data crunch Corp Com and book an appointment with us, we’d love to help your project be a success.
Ginette: That’s our show for today, and we hope it’s been helpful in understanding what you can do to drive change in your organization with data. As always, our show notes are available at datacrunchpodcast.com, and if you enjoyed the show, leaving us a review on iTunes or wherever you listen to your podcasts would mean a lot to our small team. And for new listeners, if you would like more in depth information about how to get value from your data, you can sign up for our newsletter on the website as well. If you are starting a data visualization project, or you are stuck on one, head over to our company website at datacrunchcorp.com and book an appointment with us, we’d love to help your project be a success.
Attributions
Music
“Loopster” Kevin MacLeod (incompetech.com)
Licensed under Creative Commons: By Attribution 3.0 License
http://creativecommons.org/licenses/by/3.0/