What is the Role of Artificial Intelligence in Marketing?
Digital marketing gathers and generates more data today than ever before, and every single bit of it can help you improve how you build and reach your audience. The problem is, how on Earth can anyone make sense of all that data? Who has the time to sift through it all and pull out the most important information, or make connections between things and draw new conclusions? No one! That’s where artificial intelligence comes in. But before we can rely upon machines to tell us what’s important in the data we gather, or what to do with it, we need to establish better systems for gathering and managing all this information.
In this episode of Expert Marketing Matters, Chris Butler and Lindsey Barlow discuss how the CRM has evolved to become a master data repository and analytics engine capable of telling you more about your business reality than you ever thought to ask.
You can listen to the episode using the player embedded above, or you can read a full transcript below.
Chris Butler: Welcome to Expert Marketing Matters. I’m Chris Butler.
Lindsey Barlow: I’m Lindsey Barlow.
Chris Butler: It’s the two of us for this episode. Lindsey, you were on an episode, I don’t know, a few months ago, is that right?
Lindsey Barlow: Yeah. After Dream Force.
Chris Butler: That’s right, after Dream Force. So for those of you listening who didn’t listen to that, Lindsey is basically at the center of everything that we do here at Newfangled, especially in regards to what we’re doing with Sales Force. We are resident Sales Force experts. I’m being kind of cagey about that because right now we’re trying to work out what your title actually ought to be.
Lindsey Barlow: Yeah. Although we decided today.
Chris Butler: Oh good.
Lindsey Barlow: Yeah.
Chris Butler: All right. And it is?
Lindsey Barlow: Director of digital operations.
Chris Butler: Yes. That’s great.
Lindsey Barlow: It’s official on the website.
Chris Butler: Yeah. It’s interesting. The reason I wanted to bring you on the podcast today is because your role has really evolved as our business has evolved. I think our business has evolved as not only CRM as a concept has evolved because it’s so much more than just customer relationship management. We all know that. But the landscape of what the digital agency does or the sort of expertise driven firm, the firm that offers some kind of service to another business, how it’s managed through something like CRM. So you began here as somebody who was … We called you a resourcer. What that really meant is that you were in charge of scheduling, pricing, forecasting, right? Making sure that what we do gets through the system efficiently and profitably. So there’s a little bit of project management in there and some other technical things, but that’s really naturally evolved to what you just described. But it’s interesting because what you described is both an internal and an external discipline.
Lindsey Barlow: Right.
Chris Butler: So I thought we could talk a little bit about that because I think that a lot of people listening are probably experiencing a similar evolution how they manage their firm, and we’ve talked about this numerous times on the podcast about how the data that plays a role in sales and marketing tends to play a role or have a lot of overlap with business metrics that people want to look at. But I thought maybe we could start off by just thinking … Maybe you could tell a little bit of a story. Talk about what your role was and why its become what it is.
Lindsey Barlow: Sure. Well, my title when I first started was traffic and marketing coordinator. I think the marketing coordinator was a little bit of misnomer but the traffic wasn’t. So back when we were building websites, I mostly sat between our project managers who were more like account managers and our developers and sort of routed things and just kind of oversaw that stuff big picture to make sure that we were hitting our ultimate deadlines. Then I just kind of grew the role when I saw a need. So for instance, implementing a sort of resource forecasting solution because we were selling things and then it was turning out I don’t even know if we have a resource who can do this project in two months down to getting involved in billing to make sure that we were billing our clients correctly and putting in the right invoices. Then on the other side of that, Mark had just said, “Hey, we have Sales Force. You see like you have time to check it out. Why don’t you just like see if you can figure out how to make us use it?”
First getting Mark to use it for sales and sort of figuring out the best way of configuring it for that need and sort of working with him to stick with that for a while. After that I think because Mark is how he is, it was like he was really forward thinking about it. Like, “Okay. Maybe we can help our clients with this.” I ended up developing what’s now become our sort of Sales Force consulting module because a local non-profit out of the blue got in touch with me and asked if I could help them. So I did and then I just sort of turned that into this offering that I delivered to some clients and it grew from there.
Chris Butler: If I could just interrupt for a second, so for those of you listening who haven’t been in our audience for a very long time, Mark, who Lindsey’s referring to, is our CEO and is principally involved in sales. So this idea of adopting a system like this, it needs to be built around both what the system offers and what it demands of you in terms of the discipline or a logic the way that certain information’s managed and I want to talk about that in a few minutes. But also, it needs to be customized to how that person operates, both in terms of what’s required for sales for that kind of context like selling an expertise driven service that intels a long term engagement, but also how does that person do it. So I think you’re right. A lot of it is down to the person.
I also wanted to mention that there’s so much there in terms of committing to it for the long term. Because you are making an investment. The idea of using a system like this for sales as a procedure, that will make perfect sense to anyone who’s sales team is large. Right? Because you need consistency, you need a standard way of it being done so that all future clients and customers experience the same thing. But when it’s one or two people principally doing it, the difference is that you need to invest in it so that in the future, you have information that is of value to you.
Lindsey Barlow: Yeah, exactly. I mean, you can get by by using a spreadsheet. But then if you wanted …
Chris Butler: Or white board or something incredibly low fi. Yeah.
Lindsey Barlow: But if you want to answer an actual question about something like if you say, “Well, I wonder …” Just as far as like how much business did we win versus how much did we lose? What’s this ratio and you don’t have to go build a bunch of Excel formulas. It’s all just sort of there. Yeah, I think a lot of what we’ve done with Sales Force has come out of asking those questions, realizing we don’t have the answers to them, and then thinking about what’s the best tool to get us those answers. A lot of ways it’s Sales Force because it’s this hugely customizable relational database that we don’t need … Our old project management system or business management system, whatever it was it fulfilled some needs but you had to get a developer to do really anything. It wasn’t anything that I could customize. But if there’s a question that needs to be answered, I can basically go in and look at all of our business data and get you the answer in like 10 minutes.
Chris Butler: Right. Because the thing that’s interesting about Sales Force, as you said, is it is an API in a sense, right? You can use it, as we’ve talked about many times on this podcast, as a database, and you can build all kinds of things on top of it.
Lindsey Barlow: Mm-hmm (affirmative).
Chris Butler: But it’s also its own platform that is very much usable by somebody who isn’t an engineer. So if, for instance, someone says to you, Lindsey, in a meeting, “Hey, we’ve got this question that we haven’t asked before or answered before.” Your ability to go and configure a completely customer report and interface, it’s actually fairly straight forward.
Lindsey Barlow: Yeah. But to your point I think we never would’ve gotten here if we hadn’t just stuck with it, even when it didn’t seem like it was really fulfilling that big of a need. Because, as you said, with one or two people doing sales, it’s very easy to drop off using the CRM because maybe it’s slightly quicker to just use your sort of spreadsheet that you have open in a tab or not think about putting data in the CRM because you’re not necessarily really going back to that specific note or …
Chris Butler: Or you close the deal in two weeks, you were traveling, and it was all over email.
Lindsey Barlow: Right. Exactly. So what’s the point of putting it in there? We’re getting the cash. It’s already sort of in the flow of other processes and systems. But, yeah, once you’re able to look at a couple of years worth of data and see comparably what happened at this time, what happened at this time, and maybe you don’t even know what questions you’re going to ask yet, but you have to sort of assume that at some point, you’re going to be asking a question and this data’s going to pay off for you in that way.
Chris Butler: Right. I like that. I like what you’re saying because we’ve talked so much about how a digital organization, a digital expertise driven organization. We’re not talking about product development. We’re not talking about business to consumer. We’re talking about b to b, specialized expertise services, knowledge based. All these structures always initially feel like overkill. Because well, it’s like, “Look, we’ve got less than 50 people, and we’re looking to close 12 engagements a year,” or something like that. Although there’s a huge amount of variance there among our clientele, and the idea of jumping through all these hoops and building these new habits, it takes months if not years to start to see what you talked about, which is the investment is there because something like Sales Force offers you thousands and thousands of correlative affordances, right?
I mean, you’re talking about questions that you can’t anticipate that you’re going to ask. As soon as you ask them, you can answer them assuming it’s in your field of view to see what you need. But also the kinds of reporting that we are emersed in all the time, there’s always something that comes up that you don’t … It’s not that you asked a question, you just see connections.
Lindsey Barlow: Mm-hmm (affirmative). Yeah.
Chris Butler: For instance, there might be a connection between the time to close over a certain season with a certain kind of engagement and something that happens five years down the road in the relationship with that client. If you have this idea or if you have something like wall to wall where you start building the data history of that relationship from the first touch point all the way until the end, that’s the pay off is some interesting connection that will come up but that you could never anticipate.
Lindsey Barlow: Right. Exactly.
Yeah, I mean, I think it solved a lot of problems that we didn’t even really know like necessarily existed or yeah. It’s been a very, very interesting journey to where we are. What we basically built on Sales Force was a way to run our entire business. It’s like a system of tracking our time and our tasks and projects and our clients and then our sales, so that portion of it is what we already had for a couple of years before we then decided to build everything we were using, I think like three, maybe three or four other systems to manage.
Chris Butler: This speaks directly to not only your roles evolution but the evolution in the way that we see this took, this CRM concept. Most people that are listening probably operate with some kind of centralized resourcing component in their business. Meaning, you have an account manager but they’re not necessarily going to promise resources to a client without coming back and seeing how that could work out with somebody whose primary role it is to administer that. Because we have shared resources. Maybe if you’re like us and you have engineers or consultants, they’re not fully allocated to one account. So you need somebody at the center to manage traffic and manage promises. Traffic is kind of like it’s almost too simplified because you are also talking about pricing. You’re thinking about the relationship value over time. You’re thinking about performance. You’re collecting a lot of data. But if you have that model, it immediately starts to become clear if you start thinking about it long enough that all of the standard metrics that you would expect to go through something like Sales Force, which is typically marketing and sales data, start to blend very, very quickly into operational performance data. That’s why this idea of digital operations is really apropos to your role.
So I want to take a break in a minute but when we come back, I’d like to talk specifically about what is a standard object structure that you tend to … That’s like the backbone of what you consult and how do those things tend to evolve? Like what are the most common adjustments or customizations to that, and how does that start to bleed into this idea of the business operation?
You’re listening to Expert Marketing Matters, a podcast about generating ideal new business opportunities by creating and nurturing digital marketing systems and habits that have a measurable impact on your bottom line. This podcast is brought to you by Newfangled, a digital marketing consultancy focused on empowering experts to do better digital marketing. You can learn more about Newfangled’s digital marketing method at Newfangled.com.
Welcome back. We’re here talking to Lindsey Barlow, our digital operations … Director of digital operations. New title. So it’s not rolling off the tongue quite yet. Actually it’s interesting where they came from. You and I have been talking about this for the better part of a year. Trying to figure out as your role has evolved, what’s the best way to describe that role, and as many of you listening have probably have experienced for yourselves, coming up with a title is tricky because you need the title to speak the language of the people on the outside, but you also need it to accurately represent what you are doing. If you have a role that’s both internally facing and externally facing, that becomes all the more complicated because you need to position it around what your future clients might experience of you, and you need them to understand what that is. But if you’re doing something really new, that also become complicated. So we grappled with that for a while.
One other complicating factor is that my title for a long time has been chief operations officer. That made a lot of sense because of the way Mark and I have worked our dynamic for managing this place. But as everything we’re talking about today has evolved, the word operations in so far that it applies to what I do and what this business does has completely changed. It didn’t really make sense for anyone. It was funny, we were sitting at this coffee shop and I said, “Well, what if that word wasn’t my title, we’d have yours solved,” right? We would use it in yours because that’s what you’re spending all your time on. Whether it’s client facing or internal.
So that made a lot of sense also for me because my title now is chief design officer. What that means to Newfangled and to me is that … Well, for me specifically it means that I’m generally the person that notices how all things are integrated. Actually we use that term that’s attraction term. So within the firm I’m the integrated. So I’m always looking at everything that we are doing now and trying to do in the future and figuring out how does this cohere, right? That is fundamentally a design discipline. So I see that as an active design but then in so far as I work directly with clients. I’m also consulting on design related issues. So it’s made a lot of sense. It’s thankfully allowed you to have the right label for the outside world to understand what is that you’re doing for them and for us.
Anyway, where we left this was to think about what is the standard object structure for most of our clients and how does that evolve? Because I know there’s a backbone that we start with but then everyone’s a little bit different.
Lindsey Barlow: Yeah. So I mean, Sales Force word comes out of the box with this certain kind of database structure, object structure. On the one hand you have your leads and that’s kind of info about a person and the company they work for then all of their different activities. So if they’re interacting with your website, you’re sending them emails maybe or even talking to them. Then on the other side of things you have accounts and contacts and opportunities. So accounts would be businesses that either you currently work with, either in a partnership or as a customer. Opportunities would be your current deals. So an account would also be a prospect you’re in the sales stage with. Then contacts are the people who work at those accounts. So we make a delineation between leads and opportunities. Leading being kind of maybe not yet sales qualified. So they’re in your kind of marketing ecosystem being nurtured, but they’re not really ready to be sold to yet. They’re not primed yet or they haven’t gotten in touch.
One of the biggest things that we try to figure out with clients is where is the point where a lead crosses over into an opportunity? That looks really different depending on the way that you work at your business. So for us we convert people basically as soon as they raise their hand. If they’re offering up that they’re ready to talk to us about a project, then we go ahead and convert them and say, “Okay. This is something that’s actually in our pipeline.” Also, it means that a lot of things go into the opportunity stage and then immediately are disqualified. It’s like they can’t afford us or it’s not a good fit or they don’t understand what we actually do and they got in touch and they don’t actually need our services.
So we were talking about sort of the end point being the data that you get. So our choice to convert so early often leads to a kind of like lower win rate percentage because we’re disqualifying so many opportunities so quickly that it can kind of skew the numbers. But you have to sort of look at that number and then think, “Well, how many of those were just things that we disqualified within five days of them coming in?”
Chris Butler: Right, which you could represent in a report that you customize.
Lindsey Barlow: Of course. Yeah.
Chris Butler: You could filter out.
Lindsey Barlow: Yeah. So you can choose whether or not you want to necessarily count those immediately disqualified opportunities or what we even call them. So that’s the basic structure. That’s what we work with clients on. I mean, the biggest thing I think that we focus on is a, that conversion point and I think a lot of the times for clients who don’t have a CRM, we recommend to just go ahead with the sort of when they raise their hand, conversion approach. Because if you start there, you can sort of do that for a while and then think about whether or not that’s working for you or not or if there’s a point at which they’re not really an opportunity yet when I’ve been converting them. It’s at this later stage.
The other things that we really try to think about are, especially with regards to forms on the website, thinking about what are the fields that we need in the database so that we can be doing the reporting that we need to do in our CRM. All of the data, so whether or not somebody selects what industry they’re in or anything else that’s going to help you qualify them in the lead stage and then once they’re an opportunity. Do something that you sort of have to think about and make sure that you’re getting the right data into the CRM.
Chris Butler: Do those things become discrete objects in their structure or are they something … Are they individual data points associated with an object? How does that work technically?
Lindsey Barlow: So an object is basically like table. If you’re looking at a Google sheet, one sheet would be like leads. Then all the rows would be … All the columns would be your fields. Name, address, maybe industry or number of employees. Then every row is like an individual person. So for accounts and contacts and opportunities, you’ve got the same thing. If I were to say you’re a lead and I convert you and I have all this information that I’ve collected from the website over months or years or whatever, all of that is going to stay. It’s just going to be sort of transformed into these new objects. So I got all your company data in your account. I got all your personal data in your contact. So I’m not losing anything. The whole system kind of seamlessly continues to work in the way that it has.
Chris Butler: So you have a lot of control over scaling out this object structure in the sense that from what I’ve noticed, and this is also somewhat old information because it’s been a long time since I’ve had to correct any of our clients on what an object is. But I think when they … You mentioned that out of the box structure. I think a lot of people see that and say, “Okay. Well, that’s what has to be.” Whereas what you’re really talking about as objects is somewhat of a categorization method for business data or any kind of data really. I mean, in theory one could harness Sales Force to be a database for something that has nothing to do with sales, marketing or business.
Lindsey Barlow: Exactly.
Chris Butler: If you choose to. It’s fairly agnostic in that regard.
Lindsey Barlow: It is. Yeah. What we’ve built is obviously we still use those five standard objects to manage the sales process. But then we have I think 10 custom objects that we’ve built to mirror the way that we talk about our projects. So we have related to the account, we’ve also been able to add a custom object called project. So that has a lot of our internal information about specific projects like whose the billing contact and who owns this project in our world.
Chris Butler: What are its stages.
Lindsey Barlow: Exactly. Yeah. So you can sort of like build on top of objects and relate them to each other very easily so that then you can build sort of a broader picture of what you’re actually working with. They’re even beyond those five there are many more standard objects that come with Sales Force. So the thing that I try to tell people is I see often that people have … They come up against what they think is a wall with Sales Force where they can’t use it how they want to use it. So they’re like, “Well, screw it. I’m done with this. I’m going to go use my spreadsheet,” or whatever it is. Often if they just had somebody else who could see … Who knew enough about Sales Force to see what their problem was and say, “Well, what if you do it this way?” They would break through that wall. It’s so customizable and honestly, it comes with so much already built in that you just kind of have to know about and configure that there’s often not a problem that you can’t solve.
Chris Butler: It’s so interesting that you describe it that way because that dynamic that you just described is sort of assuming limitations or edges or assuming inflexibilities based on a cursory knowledge of something that can be dangerous because you know enough you think these limitations exist because you know something about it. If you knew nothing, you might be naive enough to think it could do anything. So it’s sort of a parabolic state. But it’s very similar to the developer, the relationship that developers and project planners and kind of where maybe someone planning the project, whether they’re an account manager or project manager or a user experience designer, knows enough about programing methodology or approaches because they’ve worked developers long enough to assume certain edges, certain limitations. A lot of times developers have to unpack that and that’s difficult because they’re like, “Well, what are you telling me is the need?” A lot of times people think that they’re saving the developer time by saying, “We’ll do it this way.” But then the developer is able to make those connections and say, “Actually, based on the system, anything can be done. Let’s figure out what’s the best solution.”
Lindsey Barlow: Right. Exactly. When you’re posing the problem, you’re including all your assumptions about what possible solutions exist instead of just saying, “This is what I want to solve. How would we do that? If it were possible, how would we do it?” That’s also the scary thing about Sales Force is you can sort of say, “Well, there’s so much there and I only right now need these five things. Is it really worth it?”
Chris Butler: I think it’s interesting once you got past the core objects that pertain to sales and marketing, starting with something like projects is interesting because that was our experience that the first thing … When we started to push out the application of Sales Force or Sales Force’s per view in our business, the first stage was how much project management can we get done with the system? You and I, specifically, were trying to figure out, “All right. The system we’re using, it’s really starting to buckle under pressure. The complexity of our projects, the speed of them.” We were using an open source tool that was more about deep software development projects as opposed to a wide array of accounts. We weren’t able to find a good solution. So it became evident we need to build something. It eventually became evident that we could build that on top of Sales Force because yes, you could build this project object and it could be a very extensive table of all kinds of data. But that’s the first step to putting your whole business in there. But it would probably be very difficult, right? For someone to hear this story and say, “You know what, I’m all in. We’re going to re-architect everything tomorrow.” Right?
Lindsey Barlow: Right.
Chris Butler: It’s better to sort of start in the middle and push out. I think.
Lindsey Barlow: We had a pretty good system down for sales and we had good reporting around it. I had a good sort of idea of what Mark wanted. We had a system in place for making sure that he had the dashboard that he wanted. But we kept pushing up against the fact that when something went from sales to accounts, it was like a black hole. It was like we had no idea how those things might be related or even for a client who comes back to us for two or three projects. What overall does that look like?
Chris Butler: Right. And what makes it even more complicated is what happens when you have an engagement that is based on a recurring logic where actually this engagement consists of consulting that goes on into infinity assuming it’s working out well based on these relationships, but we also might be doing segregated bursts of project type work. So having to figure out how to account for that complexity because it’s totally different types of management. But that brings in, which is like the ultimate data point, which is time. Right?
Lindsey Barlow: Mm-hmm (affirmative).
Chris Butler: Time is sort of the bedrock of any organization like ours. How much time goes in. You’re thinking about the big picture business metrics like utilization. Utilization requires that you know where you’re spending your time. If you’re not tracking your time, you don’t really know what your utilization is. You could make an assumption about it based on how much revenue you get and how many employees you have and how much you assume they’re working. Maybe you assume it’s 40 hours a week. But if that’s not actually true, then you don’t really know what your utilization is.
Lindsey Barlow: Right.
Chris Butler: Therefore, if you don’t know the utilization of your firm at large at any given point, you don’t really know the health of any individual account. So I mean, when we took our first step into accounting for project management or our own digital operations, time was the first piece that we needed to get in there. That was a pretty interesting thing because Sales Force wasn’t really ever built to be a tool that captured time.
Lindsey Barlow: Right. Yeah.
Chris Butler: I always think, I’m a visual thinking, so I always think of this in my head when I think of this process … [inaudible 00:28:10] [Commercial break scene overriding speaker].
This is Expert Marketing Matters, a podcast about generating ideal new business opportunities and creating your future.
It’s an ecosystem that all these things depend on one another. I think that in the end, you’re never going to survive as a business if you do not let this sales experience, the most external, have a feedback loop to what you actually doing. You hear from the customers what it’s like when they’re buying, what they think they’re buying, how you’re articulating it. That needs to touch marketing. That needs to touch R&D. It needs to touch operations. All those things need to be talking to one another. So that could be doable even if all you ever have in your CRM are marketing and sales data. But I don’t think it’s that feasible to sustain. It would require a lot of discipline on the behalf of people. Whereas what you’re talking about is a system that a lot of people allow themselves to be accountable to.
Lindsey Barlow: Yeah. I think what you’re describing that I think for a long time it was like I was kind of the person who needed to be that feedback loop because I had eyes on everything. But I think that’s … There’s only so much that one person can really do there. One thing that I think is great about what we built is that it allows people in our organization who maybe weren’t ever thinking about that kind of data to really pay attention to it. It was so easy to be disconnected from that because it was very siloed and only certain people knew how to access it or needed to.
Chris Butler: Well, it’s kind of the historical disconnect between sales and marketing is that the sales person is like, “Well, tell me what we’re selling and I might be able to tell you what the customers want,” and the marketing people are saying, “No, here’s what we sell.” They need to be immersed.
Lindsey Barlow: Right.
Chris Butler: In the same way, a developer in addition to being given a spec, needs to contribute to the spec.
Lindsey Barlow: Mm-hmm (affirmative).
Chris Butler: In the same way that a consultant needs to have their sort of base line perspective on the world, but they also need to ask a lot of questions in order to figure out, “Well, how does this thing adapt to those circumstances?”
Lindsey Barlow: Yeah.
Chris Butler: And again what you’re describing is a system that makes that dynamic happen without necessarily depending on the integrity of each individual, right? It’s a thing that we can depend on to draw that out from us at the right time.
Lindsey Barlow: Right. I think it’s made us a little bit more in tune with each other’s worlds.
Chris Butler: Right.
Lindsey Barlow: Kind of understanding, “Well, how do we get to this point,” or even thinking about … Even asking that question like, “What do we talk about in sales or how are we articulating this thing that we’re doing?” Instead of just, “Somebody told me to do this so I’m going to do it, and I’m going to log my time and I don’t even know how much this cost. If we’re billing the client correctly. I just worry about my one piece.”
Chris Butler: Right.
Lindsey Barlow: I think the enhanced visibility of Sales Force has really helped that. If it’s in your face, you can’t help but notice it, which is a good thing.
Chris Butler: That’s a great segway to something I wanted to make sure we talked about before we stopped, which is what the future holds. I was thinking that for most people listening, what the future holds in regard to what we’re talking about is something that I think is very much a part of our present, which is the scorecard that we look at every week. The scorecard is this idea that all the core business metrics, the things that we need to be looking at, and they represent every corner of this business. There needs to be a central place where all that stuff lives and it needs to be reviewed every week by the core team. We have a scorecard, but it’s not a spreadsheet. It’s a interface in Sales Force. I think for many people listening, the idea of having a some source of truth that describes every single measurable point that’s critical that if you look at it, you can tell if the business is working and it’s healthy. Is your mission being accomplished in one place? That’s remarkable. Most places don’t have that. But the idea that it would be sort of living, breathing interface that can be adjusted based on the data that passed through it, that new reports can be added to it and it can be shared and centralized and shared by all, that’s remarkable.
So I think that is the most immediate future for many people is the thing that we’re very much enjoying in the present, this scorecard. But what’s beyond that?
Lindsey Barlow: We talked about this on the other podcast I was on. But I think beyond just having the data as like having AI. Instead of having to ask the question, have a machine telling you, “Hey. This is an interesting thing that I’ve noticed based on all your other data because I can sit here and make a thousand pivot tables and see that this is significant. So why don’t you do this?”
Chris Butler: Basically what we’re talking about there is an AI having the processing power to run every possible corelation observable and dismiss the ones that it perceives as irrelevant.
Lindsey Barlow: Exactly.
Chris Butler: But in theory, a good AI wouldn’t just make that judgment. It would say, “Hey, we’ve noticed this. Is this a thing?”
Lindsey Barlow: Right. Exactly.
Chris Butler: As extensive as we’ve been extolling the virtues of Sales Force and how centered it is. But there’s always a story beyond what the data capture. So it’s likely that a good AI, initially anyway, would catch some things that are legitimate corelation but you can dismiss because you hold knowledge it doesn’t have.
Lindsey Barlow: Right. Exactly.
Chris Butler: Perhaps you do that enough times, then it becomes even more sophisticated.
Lindsey Barlow: Yeah. One thing that’s been interesting about our experience so far with Einstein, which is Sales Force’s sort of AI. Framework is .. For sales insight, we don’t have nearly a large enough a data set for it to matter because we only make say 24 sales in a year. That’s not a ton of opportunity to be sifting through. But because we run the rest of our business on Sales Force. We have thousands and thousands of time data entries or task logs or things like that. I think as this tool grows, that data set will be able to be minded in a way that our sales data won’t necessarily be able to.
One thing we have done, which I don’t think I’ve showed you yet but I should, we bought the … I’m not going to be able to remember what it’s called. Sales Cloud Einstein, which is pretty much the most basic AI tool that I think they kind of sale and basically it’s just this little widget that sits on … You can put it on say like an account record. It’ll say, “Hey, you sent an email to this person and this person, do you want to add them to the database?” Or, “Should I do this?” Or “Hey, I noticed that …” it just sort of gives you these little insights and it sort of calculates stuff.
Chris Butler: It’s Clippy for Sales Force.
Lindsey Barlow: It’s essentially Clippy.
Chris Butler: But smart. It’s not an idiot.
Lindsey Barlow: But actually useful.
Chris Butler: Yeah. I’m really fascinated by the idea that as the product itself is being nurtured and maintained, the product being Sales Force, as a database, as an engine, that you can start to build on additional layers of things that can sort of dip in and out of the product and be somewhat of an advocate in that regard. Someone else who’s observing. If you think about what that really is, it’s like that’s an extension of your brain that can work beyond your own boredom, right? Fundamentally I think I have all kinds of philosophical objections to AI in our industry and in the world in general. I also have objections to it because I don’t think it’s good for society but I also think that most of the time when we’re talking about AI, what we’re really talking about is just really fast processing. There’s a difference. A lot of people hear AI and they’re like, “Oh, it’s a conscious machine. It feels pain and now I have to think about my ethics.” No, we’re talking about a piece of software that is sophisticated enough to extend it’s per view beyond what we would normally assume is a limitation. But also can just process faster.
Lindsey Barlow: Right. It’s doing all the … You could do these calculations, but you just don’t have nearly enough time.
Chris Butler: Or the interest. I mean, fundamentally you keep talking about the extensive nature of our database especially when you get down to time, right? There’s a long list of data points and objects in between accounts and time. There’s hundreds of them if not more. The ability for human being who’s interesting to sort all that out is pretty much zero. You would get bored.
Lindsey Barlow: Well, in the computer is kind of agnostic. It’ll be like, “Well, let’s see if time is correlated with bananas.” If it find one, it might say, “Hey, these are related.” You’re like, “No. They’re not.” That’s what’s nice about the demos that I saw is that Sales Force always assumes that you’re smarter than the machine when it comes to like looking at an insight and seeing if whether or not that’s actually insightful.
Chris Butler: Mm-hmm (affirmative).
Lindsey Barlow: Like one of the examples they used was it found that women … It was like women get pregnant was basically the insight that it … You know what I mean. Like a high correlation between women and pregnancy. It’s like, “Well, yes. Of course.”
Chris Butler: We can praise the marketing AI for the fact that it was able to deduce that because there’s no reason to train a marketing AI in basic human biology.
Lindsey Barlow: Right.
Chris Butler: So there’s no way it would know that.
Lindsey Barlow: Yeah.
Chris Butler: I mean, we think that’s hilarious because it’s like, “Of course.” I mean, has a man ever gotten pregnant? No. But why would an AI built specifically for marketing purposes or basic data analytics purposes know that?
Lindsey Barlow: Yeah. Then the flip side of that is if you’re going to get to the answer, you have to ask the question. So if you didn’t even think to ask the question that you need the answer to, you’re never going to get that data. So it’s just asking a bunch of different questions and giving you the answers. You get to decide whether or not they’re relevant.
Chris Butler: That’s why I like that, and we’ve mentioned this numerous times on the podcast, is we continue to work internally on building what we call the inside engine. Fundamentally the idea is we’re not trying to build an uber AI for our clients, but what we are trying to do is help them answer a simple question, which is what should I do next, right? For me, that’s the governing principle of anything that we’re going to invest in on R&D level is that can we advance the line a little bit further in terms of what we can do for our clients?
A human could do that. A human could spend a bunch of time looking at a client’s situation and say, “Well, based on what I’ve seen here, this is what I think you should do next.” But if we go about it that way, that means that we can answer that question very infrequently. What we want to be able to do is increase the frequency of answering that question and that’s where you require machine learning, which I think is a much better phrase to invoke than AI. Because I think AI’s a bit over blown. It’s catchier from a marketing perspective. It sounds like, “Oh, wow. There’s a little brain in to that somewhere working for me.” But the machine learning is basically building an algorithmic backbone to how data is being processed so that you can get that question answered more frequently. What we’re trying to do is answer it more frequently than once a month because that’s our general pulse with our consulting relationships is monthly interfacing with our clients and looking at their situation and helping to steer it. But if we can be doing that in an automated fashion as often as the client needs it, then that question, what should I do next? That’s huge. That’s powerful. I mean, that’s the governing principle of where all this is going then that’s fine.
Lindsey Barlow: Mm-hmm (affirmative).
Chris Butler: This has been cool. I really appreciate your insight into all this stuff. We should probably have you on the podcast more often because this is really is where everything’s headed.
If you’re listening and this has answered questions you have or maybe even giving you questions you didn’t know to ask, you should get in touch, especially get in touch with Lindsey. You can reach Lindsey on the website I believe.
Lindsey Barlow: Mm-hmm (affirmative).
Chris Butler: You should read things that Lindsey’s written. Normally we conclude by recommending content on the site. We used to do that. It was part of our …
Lindsey Barlow: Oh yeah. I remember.
Chris Butler: … show structure. I would just say go to our insights page and filter by Lindsey if you’re interested in these subjects because really nobody … I mean, we’ve all written things about Sales Force, but no one’s really written anything about Sales Force other than Lindsey. We’ve mentioned Sales Force here and there, but Lindsey knows how it works.
SO thanks again for this. It’s been a lot of fun.
Lindsey Barlow: Yeah. It was fun.
Chris Butler: Thanks for listening. You can find us at Newfangled.com. Find us on iTunes and give us a rating or a review. That would definitely help in spreading the word about the podcast, and we’ll see you next time.