Doug Laney: A framework to turn data into $$$ + Data Juice book giveaway

**Loris Marini:** Alrighty. So while we are here, we're here because we wanna turn data into business impact that means. Many things, brand differentiation, bigger margins, less exposure, and the resilience that the business needs to scale.

Often the data is already there and all we have to do is to come up with the business cases that are strong enough to justify spending time, money, and resources on all the things we need, modeling, infrastructure and talent. So today I wanna talk about the process of developing new ideas of monetizing.

Data assets. Is there a framework that we can use to come up with new ideas, to turn data into outcomes, get the business buy in, and measure the impact of our work? And I wanna explore these questions with Doug Laney, thought leader in data strategy expert. Doug has been on the show before. That was episode 20.

The Discovering Data. If you haven't listened to that, I strongly recommend going back into our archive. They're gonna be links and the show notes as usual. And you may know Doug from his best selling book Infonomics: how to monetize, manage, and measure information for a competitive advantage. Doug is a consultant advisor, author, speaker, mentor, and instructor.

Is all of those in one person. He's a three time recipient of Gartner's annual Thought Leadership Award, and he regularly writes for the major data and information management magazines. Doug just published a new book, The Data Juice, which we'll talk about today. And just a heads up that we're giving a few copies away the end of this episode.

So you wanna participate. You can head over to discoveringdata.com juice and put your details there. Doug, welcome to the podcast. It's a pleasure to have you back on the show.

**Doug Laney:** Thanks, Loris. Great. Great to be with you again. How are you doing?

**Loris Marini:** I'm doing well, enjoying the countryside. We moved from Sydney. We're a couple of hours away and yeah, when I look out of the window, I see horses and blue skies, so can't complain

**Doug Laney:** Very nice. Closer the, Blue Mountains or not?

**Loris Marini:** Yeah, just past the blues.

**Doug Laney:** Lovely, lovely.

**Loris Marini:** Alright, so started reading your book and it's fascinating how you manage to really do walk the talk. We talk a lot about information science and managing knowledge. And what you've done with this book is really not only collect the wisdom and the and insights from many experts a hundred to be exact. What you've done is create a almost a glossary. So you can walk, you can read the book. In many ways, there are hashtags, they're categorized by type of business impact that just love that type of work. So really

Thank

**Doug Laney:** Thank

**Loris Marini:** guess

**Doug Laney:** you. Thank you. Yeah, there really wasn't anything like it in the market and, kind of based on Infonomics, people kept asking me for more and more examples of what, other organizations are doing. And I've been compiling this Kind of library of stories for, for years. I rejoined Gartner as the kind of the big data analyst cuz I'm known as the guy who came up with the three Vs of defining big data, volume, velocity, and variety and.

And so they said we want you to cover big data. And I was you know, advising our clients on that and writing about it. But the questions evolved from, what is big data to, how do we do big data to what do we do with all this big data? So that kind of compelled me to start collecting stories on how organizations were generating value from.

data and analytics, and so that library's grown to, we're 500 stories today and decided to put a hundred of them into the book.

**Loris Marini:** Wow.

**Doug Laney:** like you say, it's more of a reference book than a book that you sit down and read. Although I keep meeting people who said, they started reading one story and they couldn't stop and next thing they knew they were, finished reading the book.

but yeah, one of the things about the book is we indexed it also. So the book is indexed by the type of use case, the type of industry, things like that. So you can flip to stories that are relevant to. maybe the type of data that you're interested in or the type of analytics that you wanna deploy, or again, the industry or whether you're interested in cost savings or revenue generation or improved, customer satisfaction, you can flip right to those stories

**Loris Marini:** That's right. You almost say that taxonomy work

the book, book

Yeah.

**Doug Laney:** Yeah. And the book is meant to be, inspire, people to inspire organizations to do more with their data, or in some cases shame them, into doing more with their data. Yeah.

**Loris Marini:** So I wonder like after all these coversations in this many years talking about this topic, what are some of the common patterns that you see of in organizations that are successful in turning data into business impact?

**Doug Laney:** Yeah, there are a number of commonalities that I actually wrote about in the introduction to the book and, lemme see if I can think of some of them. You, one is that a lot of the stories are very functionally specific, and. Quite, vocational. They're not enterprise class, blah, blah, blah kinds of solutions.

They were like, we noticed this opportunity. We think we can use data to solve this problem or exploit this opportunity. And and very often they generated revenue value streams well beyond the, investment. yeah, I'm sure there's a lot of, infrastructure. That has to be dealt with. But, most of these stories did not have to build a, big enterprise data warehouse or data lake in order to leverage data in, innovative and high value ways. a lot of the stories also involve analytics well beyond just, pretty pie charts and. Dashing dashboards, they involve more diagnostic or predictive or prescriptive kinds of analytics. In fact, when I did a meta-analysis of all the use cases, I found that very few, maybe a couple percent had anything to do with a dashboard or pie charts or anything like that.

They had to do with, deploying analytics to diagnose or predict a prescriber, automate or digitalize something. So there's a good lesson in, in that, as. a lot of the stories also involve the use of multiple data sources, not just within the organization, but outside as well. So a lot of the high value stories will involve integration of public data or social media data or data that's been harvested from, websites or that's been exchanged from partners or suppliers.

So those are some of the the kind of the key lessons.

**Loris Marini:** Yeah. And I'd love to hear your favorite story or maybe your favorite top two stories. We could dive into that or, but there's a section of the book that really caught my eye and is the myth versus reality table. I think it's brilliant because it's really, it condenses a lot of the issues that we find when we talk when we try to have conversations with business stakeholders.

**Doug Laney:** If you wanna talk about that for a bit, that I think that's really important for folks who are thinking about monetizing data and what is, we should probably talk about what data monetization actually means, but Yeah, there are a number of, myths data has to be sold to be monetized, right?

So you don't necessarily have to sell it or sell it for cash. very often we can exchange data for goods and services or commercial. Favorable commercial terms. we we can do that. Also, as I mentioned, a moment ago, you can monetize data that isn't necessarily your own. You can go out and harvest data from the web or grab. Data from public data sources or social media, package it up, analyze it in certain way, report on it and, sell that. In fact, there was a company here in Chicago called Food Genius that a number of years ago started harvesting restaurant menu data from from the interwebs and they packaged all that data up and we're able to sell insights into which ingredients and what kinds of menu items and pricing data and all of that and sell that to restaurant tours and.

And, grocers as well. They ended up getting acquired by US Foods, a major food distributor, but, they, they built a entire business, har. With data that wasn't even their own. Think well beyond your own four walls when it comes to data that you can generate value from very often integrating it with your own data.

For sure. You, the other things that are often myths are that you have to be in the data business to monetize data. Absolutely not. Most organizations today are generating value from data. It's just, if you don't measure that value, then you're an, it's difficult to claim you're monetizing. There's another one in here that that I often mention, and that is due to privacy regulations. We can't monetize our customer data well, that just shows a, an abject lack of creativity. Plenty of companies today monetizing their customer data without exposing who their customers are.

They either create synthetic data sets or they aggregate data or they anonymize it or. They do what I call inverted data monetization, which is I can't sell you my customer data, but I can sell your stuff to my customers. So without ever exposing who they are. So for example, we're working with a hospital who knows who its diabetes patients are, and it can't sell that data to anyone due to privacy regulations, but it can introduce healthy meal plans and gym memberships and.

At home glucose monitoring, test kits to others and, take a referral for that. It's very much kinda like the Facebook model where, they don't share or sell customer data anymore ostensibly, but what they do is they allow you as a, commercial organization to target certain types of customers via on their platform. that's the kind of thing that we're setting up for, for organizations at West. yeah, there's a number of myths and I think, it's important to to talk about what is the definition of what is, what, how do we define data monetization, right? And really we define it as a, as the process of generating measurable value streams from available data assets. Like a little bit to unpack there, right? So it's a process, so it shouldn't be a one time thing. It should be set up as very much like any other. Should take a product approach just like any other product or service offering. Measurable value. Again, if you're not measuring, the, value that's being generated, it's hard to claim you're monetizing it and therefore it's hard to get, investment dollars

**Loris Marini:** do and

**Doug Laney:** then,

Value streams, so different ways of measuring value and generating value. It doesn't have to just be revenue. It could be cost savings, it could be reduced risk, it could be improved relationships, it could be. Improves safety or satisfaction if you're a non-commercial organization. and then from available data assets, And by that we mean, again, look beyond your own four walls for data that is potentially potentially.

**Loris Marini:** Yeah. And obviously the critical part here is measuring and to me is also getting the teams to work together, right? Cause when we talk about business outcomes, is there outcomes for the business, not outcomes necessarily just for marketing or sales or

**Doug Laney:** Yeah.

one of the challenges there getting teams to work together is some of the language that we use. And I've been railing against this a lot lately, and that's the concept of, data ownership. I think data ownership is one of the worst ideas we've ever come up with as a profession.

The data management profession, we're always very quick to assign owners to data and. Manifests in is a lot of data hoarding and, architectural data silos that's no good for anybody. So when I started to look around at other kinds of asset management concepts I quickly gravitated toward the concept of a trustee or a fiduciary, someone who's legally and ethically responsible for the wellbeing and, safe keeping of a particular asset.

But that isn't an. Outright owner and I think that's the kinda language and concepts that organizations should

adopt.

**Loris Marini:** That's right. Cause ownership implies control of the asset. And I think there's a bit of an unfortunate collision there as well because in, software development we use the word own, the verb to own something to express, I think in general, in project management to express the.

The fact that someone is accountable for a specific part of a larger system. So you own data modeling. It means the engineering function is responsible and accountable for the models that are delivered and they need to keep systems up and running.

**Doug Laney:** Yeah. If you've ever created like a racy diagram, you notice there's no notion of ownership in there. It's, who's responsible, accountable, what's the C for

**Loris Marini:** It could come from the agile world, the word ownership which really, I, I think the way that I interpret it at least is to say there is one as Bill Schmale called a one throw to choke. There's one team, or one,

**Doug Laney:** Yeah,

**Loris Marini:** unit

**Doug Laney:** I'm a little less violent than him when

**Loris Marini:** Yeah, I know it's a bit's, a bit of a aggressive , so walking the line between business and technical data leaders, what I see a lot is gaps on both sides, right? So the business talks about outcomes and I actually, that's what they want, right? And the technical folks, the professionals go all of this talking is great.

We want them to, but how do we do it? And not just tactically, but strategically, what do you actually do first? Have you seen from your experience anything that is a lead indicator to success in terms of being able to turn data into our outcomes from, a people perspective?

**Doug Laney:** A few things. one is, and I've looked at this a couple different ways, just having a chief data officer. we ran a study last year, late last year, and found that organizations with a chief data officer are significantly more likely to be generating economic value from their data. They're more likely to be doing advanced analytics.

They're more likely to be democratizing data throughout the organization, sharing data freely. They're more. To be dealing with data quality issues at the core, lot of real benefits that come from having a chief data officer. And then I also ran a story, look, a study looking at How do data savvy organizations differ than those that aren't?

And so we looked at certain qualities like do they have a chief data officer? Do they have a data governance function? do they have a data science organization? Those are indicators to that, that an organization is more serious about managing and monetizing and measuring their data as an actual asset.

And what we found is that, Companies demonstrating those kinds of characteristics have a market to book value ratio that's nearly two times higher than the market average. Oh, this is not 2% or 20%, It's a 200% difference. So while accountants don't value data, when we talk about that as well, the craziness around, around that investors do and investors definitely favor organizations that are data or digital organizations, Or data savvy. But we also found that, companies that make, a living by, selling data, licensing data or, creating data derivatives or digital products, have a three x market to book value ratio. So there's, something significant that investors recognize in those kinds of organizations.

And this has led to some companies thinking about how to carve out their data and digital, Assets as a separate entity because they get value differently than the core business. And also because again, accountants don't allow you to value your data assets, but you can value a company that owns your data assets as a subsidiary.

So through a bit of accounting shenanigans, we're helping some of our clients, actually quantify the value of their data by including it as a subsidiary on their balance sheets. So we call that appetizing, Their data.

**Loris Marini:** izing the data.

**Doug Laney:** Izing. Yeah,

**Loris Marini:** That's interesting. Cool. So let's go back to the process. Cause I'm thinking, I'm imagining a room with a bunch of people. There's the head data, VP and even the chief. You are actually sharing the same room for an hour. They have a bit of a problem.

They've done a bunch of things, measure a bunch of value in a bunch of ways. Now they're running out of ideas. How do you go from, we don't know what to do next? How to use the data that we already have to do, Pick the one or two activities that are most likely to impact the business.

**Doug Laney:** It's a great question and something that I, I just don't think a lot of organizations have dealt with formally. So we at West Monroe, we developed a kind of a formal approach to this, and it starts with workshops to generate and refine ideas for innovating with data assets. And in those workshops we look at a few different things.

We look at what data's available both inside and outside the organization. and what it could tell us if we, tortured it enough. Oh, there I go. Being violent, Bill sch, Marzo,

**Loris Marini:** Yeah.

**Doug Laney:** and then look at stakeholders. What, is the range of stakeholders that could possibly benefit from the data, both internally and externally.

And so we're looking at the extended business ecosystem of, suppliers and partners, suppliers and suppliers partners and partners, customers, and so forth. Hedge funds, industry organizations, ad firms who might find some of your data or insights valuable and pay for it. then we generate various hypotheses.

Again, we're trying to get beyond just hindsight oriented reporting to look at what could we do with this data to diagnose or predict or prescribe something, or automate or digitalize something. we also look at, various use cases. So we'll look at what. Other organizations inside and outside the organization are doing with data as a source of inspiration and, so you ask, why out, why organizations outside the industry?

I think, you know that my kind of, my flippant response to that question is, why do you want to be in second place or third place? Why not gain ideas from outside your industry and be the first to apply them, within your industry? Classic example is, the police department of the city of Los Angeles.

Somebody came up with the idea that hey, it looks like crimes follow a pattern that are similar to seismic aftershocks. So they apply a seismic aftershock prediction algorithm when a major crime occurs. And lo and behold, they're able to identify the increased likelihood of where a future crime is going to occur, deploy resources there, and reduce violent crimes in the city by, 30%. So pretty significant. So again, taking an idea from another industry is I think, really important for organizations to do anyway, at the end of these workshops, we'll end up generating. 30, 40, 50, 70, ideas. and then them refine them, maybe combine and then the next thing we wanna do is assess them.

We wanna assess and prioritize them based on a spectrum of feasibility characteristics. are they technically feasible? Is the data of good enough quality? Are they managerially operationally feasible? Are they scalable? Are they legal? Are they ethical? And so when we assess them on these characteristics, certain ideas will float to the top as being high impact, low complexity.

**Loris Marini:** So there's low

**Doug Laney:** those are the

**Loris Marini:** engineering, there, is

**Doug Laney:** Yeah.

**Loris Marini:** There is. There's architecture and infrastructure. There

Talent available to do it.

**Doug Laney:** Exactly. Exactly. And so you can gauge all that fairly easily and come up with a score for these ideas. and then seek what kind of floats to the top. And then the next phase, the second phase is really designing the, data product, if you will. And that is where we go out and test the markets, identify potential buyers and users of the data products, assess how we're going to package and license the data or, Derivative then we're, how we going to identify and curate and prepare the internal and external data assets, specify the technical requirement, governance, organizational requirements. And then because we're treating this as a, we're really taking a product management approach. We also wanted define the support, maintenance, and reporting requirements for that.

And then the third phase is then we'll start to architect and engineer the infrastructure requirements, build the data products, implement them, introduce the support maintenance, and maybe. If they're externalized, train up or, set up a, sales and marketing and support function for them as well.

So really, it's, very similar to any other kind of product management, product marketing approach. Yeah, I don't, I'm not really one for reinventing a wheel that works already.

**Loris Marini:** Yeah. You published your post recently on LinkedIn. What is the data product here? The definition, like the most I'll give a book out. What did you

**Doug Laney:** had.

**Loris Marini:** picking and what is your

**Doug Laney:** I ended up picking from a certain guy. I'd have to look it up right now, but I don't have it at my fingertips. but yeah, it was a really good we wanted to define the difference between a data source, a data asset, and a data product. and I had a particular interest because we were working with a large manufacturing client of agricultural products.

Machinery. And they are interested in understanding the value of their data throughout the organization. but not just the data sources, but also the data assets, the a the aggregates of, similar data. And then the data products, where data's actually used. How are we valuing those use cases?

As well. and they're using that to target their investments in acquiring data, generating data, managing data, building data assets, and then building data products. and, yeah, it's a really initiative. We're doing to build that framework for them and we're now into actually, valuing, the data assets and data products.

**Loris Marini:** So how do you see the difference between data assets, data, products?

**Doug Laney:** So a data source would be, an individual kind of data set, right? Probably raw data of some kind. A data asset is a more of a logical grouping of data that's. Typically for a single purpose or somehow related. a, data product is more of a use case, how that data is used, which may include applications or algorithms those kinds of things on top of the data. I'm not sure I got that exactly right, but the, guy who, who, who came up with the definition, I like the best. There were dozens of definitions submitted on linked. I really liked his definition the best. I thought it was pretty, pretty good. And, anyway, he ended up winning a copy

of the

**Loris Marini:** And I almost got the, I was tempted to create small side project for fun really to apply what we discussed with Jessica Talisman on episode 40 on taxonomies and ontologies to that post, Cause there are more than

**Doug Laney:** Yeah.

**Loris Marini:** answers. They're all different

**Doug Laney:** So you may know, I really don't like splitting hairs on definitions or coming up with new definitions for existing, terms or words like I rail against folks who come up with new definitions of data versus information. I just think it's ridiculous. it's driven a wedge between us as data professionals and the business when we.

Splitting hairs over data versus information, I, it turns them off. It, just doesn't help with data literacy at all. And there's really no, no real reason for it. There are things we do with data to make it that much more consumable. We improve its quality, we improve its availability, we integrate it, we analyze it.

All that is a spectrum of stuff that we do with data to make it that much more consumable and usable. To generate real realized value, but I don't think there's a state change between what is data and what is information. I, dunno, On

**Loris Marini:** I've I so recently a post of someone that made me think, it took me a while to actually understand how I felt about it, but it was a, on difference between data products and data as a product. And the point of this person was that data products are things that you sell. Which, we just discussed the selling and money exchange is only one way you can monetize it.

And the other, the data as a product applies to the thinking, the methodology of treating data as a product. So that doesn't really mean necessarily the dashboards, but it could be a view on a data in a database, right? That

could be a product.

**Doug Laney:** Listen, whatever helps us communicate better with the business and help have them appreciate and understand what we do help us, help them I'm all for it. So

**Loris Marini:** Yeah, absolutely. I know that most of your clients are large companies, but there's a big need in the startup space as well to do education, and help a lot of these, often engineers that end up leading the data function.

And when I say leading, Quote unquote, because a lot of times the culture and the processes are actually come from the tools they end up using. So there's not a lot of leading. There's more pick your tool and join a tribe and you'll now be, in the camp of DBT or of Snowflake

or whatever.

Right.

**Doug Laney:** I think it's always been that way. Yeah. Back in the days when it was IBM versus VAX versus hp, people pick their tribes right.

**Loris Marini:** Yeah.

and it's a bit unfortunate because what ends up happening in the end is that people they still feel the pressure from the business to see tangible outcomes from this investment. And it's not fun to be working in or constantly feeling that you're not delivering.

It's okay to have pressure because without pressure, nothing happens. But it's not fun when you're struggling to go, like you see the value of the team. We'll be working hard for months to fix all our issues, and now the data warehouse is usable. We've got those systems in place, now we can do stuff with it.

But until, as Brent Dyke says like the value chain is long, and until you make a decision or you do something that is, that physically, has

an impact

**Doug Laney:** And hopefully measure that impact, And attribute it back to the data assets that are used,

**Loris Marini:** Yeah. And so that's, to me, that's really the challenge, right? That until you make a decision that impacts the business, it's a, it's an investment with an unclear roi.

Then the ROI comes when you make the right decision, hopefully based on

**Doug Laney:** right? So when we help, clients value their data, we're looking often not at. Not just at current use cases, how that data's currently being used, but at projected planned probable use cases as well so that we can understand the probable future economic benefit of a data asset.

**Loris Marini:** Interesting. And so with probable, you mean likely to happen or

probable

**Doug Laney:** Likely we include the probability that it's going to occur times the benefits that are likely to be, that are expected to be accrued less the cost of deploying it and operat.

**Loris Marini:** So that, I suppose it's a massive conversational exercise of talking to a whole bunch of people understanding what, how they

**Doug Laney:** It is. It is. But we find it's not that much different than valuing any other kind of asset. It's just applying the similar, kind of approach.

**Loris Marini:** Yeah. Yeah.

**Doug Laney:** financial organizations are doing this with their other assets, all the time, or their accountants are, but because accounting standards don't allow data on the balance sheet, it's, something that's unfamiliar to a lot of organizations.

**Loris Marini:** Yeah, that's actually a question that I wanted to ask you for a long time. At least a year since last time we talked and these the process of valuing an asset, in, in data we say a lot that value comes when you use. The thing being a model or a dashboard, whatever

**Doug Laney:** interestingly, that's inconsistent with the way that other assets are valued.

**Loris Marini:** True. Because if I buy a desk or a laptop that has value, I can exchange it for cash even

**Doug Laney:** Even before you use

it

**Loris Marini:** even before I

receive it.

**Doug Laney:** on a store shelf has value, even though nobody's eating.

**Loris Marini:** So how do you see that? Is there intrinsic value to the asset even if you don't use it or you have to use it or

have some sort of value?

**Doug Laney:** Yeah. Their assets can be valued based on their cost. Okay. What did it cost you to, acquire it or generate it? And typically you use that methodology before an asset is, deployed or being consumed. Then there's the income approach, which is, what's the revenue generated or the cost saved by using that asset.

And then there's the market approach, which is what is this assets value on a open marketplace? What could, what would people buy for, buy it for now with data that gets very interesting because typically when you sell data, you're not exchanging ownership of it, you're exchanging rights to use it.

there's an optimization curve there, which is, what can we sell this? What can we sell this newspaper for that we can make multiple copies of, to optimize our revenue, right? If we sell it for a thousand dollars, maybe only a couple people will buy it. If we sell it for, 1 cent, then maybe we're not making enough from it to to make up for a cost.

But there's the kind of optimization curve in there that mentioned that in the infonomics.

**Loris Marini:** And that to access the data, does it apply to insights as well? So any form of intangible sort of, of information, if you wanna call it that way? Yeah.

**Doug Laney:** Yeah. So for some companies, we look at comparables like what? We're working with a client that wants to understand actually a. American football team wants to understand the value of its data assets. and so we're looking at, similar kinds of maybe entertainment companies and what their customer lists are valued at externally and using those as comparables.

**Loris Marini:** Do you find that companies tend know, going back to what you said that really resonated, right? You talked about the approach at Westboro that you guys look at the. Ecosystem that surrounds the company, the organization. So hedge funds industry bodies, suppliers, partners, every, everyone that interacts with the org.

Do you find that organization tend to ha be aware of that or

**Doug Laney:** It's an exercise that we go through. They, generally have pretty good awareness of who these potential stakeholders are. Yeah. but they've never really laid them out before and, looked at them and thought about their personas and profiles and needs What's driving those businesses?

and so we'll explore that as well. not just identify the, players but understand the personas and, needs.

**Loris Marini:** Yeah, that's what a lot of people talk about and refer to as design, right? Design thinking is or

**Doug Laney:** I guess so. Yeah.

**Loris Marini:** Yeah. Asking those questions, who exactly are we talking to and what's their pain at the moment? How can we alleviate that pain somehow? Yeah. Super interesting though. Just to close this off, what is your favorite story?

Out of the a hundred in the book,

**Doug Laney:** I don't know. There's some great stories. There's one really cute story. and this just goes to show you that anybody can monetize data, right? So there's a story of a 16 year old girl in the uk. Whose father who's doing business in China often. And so she would travel, the family would travel with them, and she realized that a lot of Chinese people were naming giving their babies also Western names.

So that would make them easier when they get older to to go to school or do business in, in, in the US or Australia, where, wherever else. and what she, noticed is that a lot of these names were funny names. Like they were naming their, kids after things they'd seen in movies like Cinderella or Snow White Gandalf or, like that.

And that was gonna, she knew that wasn't gonna go over very well when they traveled to, to Western countries. So she's why don't I create an. to help these families name their babies. And so she created an app and integrated it with kind of the tendencies for names and how they relate to different kinds of characteristics in individuals.

And she, I know she's got some data on that. And then asks the families what kind of characteristics they want in their child, and then it recommends a bunch of names and then the family can vote on the names. And she's, her app within the first year had named a quarter million, Chinese babies. And I don't know how much money she generated, but. I think she was doing it for 50 bucks each or something like that. And real, really cool, example of using data and analytics and app and an app to name babies. there are other great stories like. On our Walmart I love this story with Walmart.

and so we'll go to the other end the spectrum, from really small business to like the massive business, right? So Walmart had a great search engine and helping people find what they were looking for was aggregating tens of millions of searches per month and combining that, matching that to product descriptions.

And one week, however, They realized that a certain search term was resulting in a very high level of shopping cart abandonment, and the search term was the word house, and it was taking people to housing goods and housewares and doll houses and dog houses. It wasn't at all what folks were looking for.

They were, and they investigated this and they learned that, realized that this search term coincided with the week that a particular television show season premiered. It was the show house,

drama Hugh Laurie. And so people were looking for obviously was the box DVD set or the ability to stream previous seasons.

And so Walmart realized what their search engine wasn't paying attention to was what was trending. It was staring at its own navel. It wasn't paying attention to social media or anything that was going on outside of Walmart. And so when they upgraded the search engine to. Take into consideration social trends.

They reduced shopping cart abandonment by 10 to 15% across the board, which in Walmart terms is I dunno, billion dollars a year of

additional revenue.

Like I said, a lot of these ideas are very vocational, very functionally specific. They're not. They're not enterprise class or recognizing there's an opportunity or a problem to be solved and that you have the data at hand or can get the data and, apply some, higher order analytics than just, building pie charts to to solve a problem or, capitalize on an opportunity.

**Loris Marini:** Yeah, if I think the take home here for the listeners that belong to the smaller organization, startups in particular, especially the engineers that are leading this functions, is get out of the engineering camp every once in a while. Talk to the folks that have problems within the business. Doug Laney Data Juice 101. Stories of how organizations are squeezing value from available data asset to Doug. Thank you very much for being on the show again, and we'll talk soon.