Speaker 1: 0:23

Hello and welcome to another episode of OpsCast brought to you by MarketingOpscom, powered by all those MoPros. I am your host, Michael Hartman, joined today by my one co-host, Naomi Liu, up in the hot country of Vancouver.

Speaker 2: 0:37

We are having a bit of a heat wave, although I don't know how hot it is for Mike where he's at in lovely Vegas. The desert, yeah.

Speaker 1: 0:44

I don't know that I call it lovely. I'm not a fan of Vegas. They don't have me on their ads so but yeah, he's probably suffering as well, but that's all right, we'll. We'll make it through, all right. So for those of you who are longtime listeners, you probably know that I believe that there's a gap between marketing ops professionals out there who have analytics chops, if you will, especially in the marketing domain. So I'm excited to talk to our next guest. So our guest today is Shivani Bhatt.

Speaker 1: 1:17

Shivani is currently Director of MarTech Solutions and Analytics at CloudFlare.

Speaker 1: 1:23

Shivani has built and scaled marketing teams at Fortune 500s as well as high growth SMBs. Prior to her MBA, she built S&P's first marketing ops team and growing that team from three to six in one year so doubling we might want to talk about that even and managing the firm's Marketo instance, arquetto directly by executive leadership to develop intellectual property for the company around how to use and productize marketing data and analytics and impact the firm's product like motion. As it aimed for an exit there, she hired and scaled a team to do so. So shortly after the acquisition by Adobe, she pivoted her team and led Adobe's flagship data governance project. After that, she was at Zuora, she ran the strategic planning and analytics function, including the modeling and forecasting of pipeline targets, reporting on attainment and down to BDR compensation. At Cloudflare, where she is now, she directs data solutions within the entire MarTech stack, which includes 26 plus platforms I think they're slacking there and analytics that flow from it, so I need to take a breath. Shivani, thanks for joining us today.

Speaker 3: 2:34

Thank you so much for having me.

Speaker 1: 2:36

Yeah, this is going to be fun. So we already talked before we started recording that we're all suffering through the heat as we're recording in late June.

Speaker 1: 2:45

So we're all fighting through it All right. So I think there's a lot of different ways we could start this. You've got a really interesting background, um, but I I'd love for you to share maybe if there's any other like little bits of your career path that we didn't touch on when my with my little introduction there. Um, maybe, like, if there are certain people that were pivotal in some of your your sort of steps along the way, or if there are certain people that were pivotal in some of your sort of steps along the way, or if there were just decisions that you made along the way that you felt like had sort of an outsized impact on your career. I think our audience would love to hear about that.

Speaker 3: 3:19

Yeah, absolutely. I think I can't take all the credit. I've had the privilege of reporting to some really great folks in the past. Alex Fleming would be one of them. I currently report to Jess Cao and gosh just wow, it's been just such a huge learning experience to work with one of the industry's top thought leaders. So I have to obviously call that out. And then you know, I would also like to mention I've had some really great professors, um, whether in undergrad or um also in my MBA. So I would, I would probably say folks like George Durizo, um Minka Kumar. They have just had a very outsized impact on the way that I think and the way that I approach problems and, um, I still keep in touch with all, all of them today to talk shop every now and then.

Speaker 1: 4:15

We're going to have to start bringing all of Jess's team on. We already had her on. She comes up regularly on these discussions. I was talking to a mentee of mine earlier this week. I was like that would be a good person for you to talk to. So, yeah, that's great, all right. So, as I kind of alluded to in kind of the intro, I really in fact, I was talking to somebody earlier this week, just a few days ago, about what I see as a gap in understanding of analytics in the marketing context in particular but we call it marketing and sales. Really, yeah, I see it as a pretty big gap where I think there's some people who know how to pull a lot of reporting but not really get insights out of them and not kind of see patterns. What's your take on that? Am I way off? Do you see something different? What's your view of that space right now?

Speaker 3: 5:34

Like folks are getting really good at understanding metrics and there's a lot of kind of homegrown knowledge that comes on analytics, just from being in this space. But I do think the statistics one that you mentioned is kind of an interesting one. That's part of what I focused on in my MBA. Um, I think my take is that there's so many great statistical tools and um to try to be able to understand how they, how they interact with with marketing data is is absolutely huge.

Speaker 3: 5:56

And I would say if there's a gap to me, it's kind of around that Like we've already, we all know about kind of statistical significance but um, but you know, like I think the concept of linear regression is is really well suited for marketing data. It's got tons of different applications. Um, I've built tools for my and worked with my teams on providing those tools to different um, you know, bdr or sales audiences, and there's just so many different ways you can skin a cat essentially with some of these statistical tools. So I think that that's an area where I would kind of see a gap. But I think there's there's lots of ways we can we can try to fill that gap.

Speaker 1: 6:42

I think it's interesting, cause I so I actually don't see a lot of people who know things like statistical significance Like I know enough, like I was doing. I was actually was talking about an analysis I did on a proof of concept project a couple of years ago and I had 12, this is for an event thing, so 12 events that I was looking at, comparing three years worth of progression to see if we could identify any kind of trends from a change in how we were doing communications, and like 12 just didn't seem like enough, right. But I freely admitted that and I just like, hey, even if the change looks to be this, even if it's half of that or, you know, 25% of that, I still think it's still worth the investment in sort of doing it anyway. But think it's still worth the investment in sort of doing it anyway. But I don't know that a whole lot of people that.

Speaker 1: 7:35

But I like, literally this week I was watching someone we were kind of talking about a project and she was doing something in an excel spreadsheet and she put median, mean, you know, and I was like, oh, I'm so glad somebody else is using median, right, I'm like I don't think most like I, I still see it because I've got kids who are in like high school age, going through, you know whatever. That would be Algebra two, I think Right, but they don't have statistics per se and I feel like that's missing in education right now. I don't know.

Speaker 2: 8:03

I'm curious, michael, if you think the gap exists with people who know how to create the reports or interpret the results, or do you think it's a mixture of the two? One person should be able to do both, and the reason I ask this is because I've worked for organizations where oftentimes the analytics function will roll up under the IT organization. So folks in finance, folks in HR, folks in marketing will go to the IT team and request certain reports to be pulled out of things like Power BI or Tableau or whatever you know data modeling tool that they may have, and then they give the parameters and then they're presented with the results, but there's no actionable insights from that. The person who's creating the report is not then saying, ooh, that's not so great, never do that again, or you should really do more of this, right. And so I'm curious, like when you say that you see that there's a gap there. Maybe I just kind of want to blow that out a bit.

Speaker 1: 9:15

No, I think you've. That's a great question. So I do not see it as an issue with people being able to like run reports and pull data from these systems. I don't think that's really the issue. I think it's. I think there's a little bit of like, how do you put data together from sort of different systems so that, like that, makes sense, right? I also think there's a lack of recognition sometimes about the movement in data that happens so quickly in sales and marketing. That is like less so in, say, finance, where there's lots of controls. So I think you know like sometimes, like you pull data and it's like it's a point in time, it's not a trend, right? So to get the trend, you have to be looking at it at multiple points in time. That's one sort of little thing. The other, I would say, is really that last bit you were talking about, like getting insights.

Speaker 1: 10:08

But I think part of that is because people don't actually know how to assess. Shivani, you brought up linear regression, which I know enough to be dangerous about like what it's trying to do. I'm not sure that I could actually do one of those kinds of analyses, but you're trying to like hold everything else constant to see, is there some variable that has an outsized influence on output than others? Right, I mean, I think that's the way to think about it and I don't think people really would understand that. And so they misinterpret data because they don't understand sort of fundamentals of statistics, so the difference between median and mean and um, things like that, and they also don't understand about, like, the difference between correlation and causation.

Speaker 1: 10:54

And very little, I think very little, we have for marketing and sales data that could be causal.

Speaker 1: 10:58

I mean there's probably some, I'm sure there's scenarios we could think of that we go, yeah, this one thing changed the outcome, change at the same time, like in a correspondent way, and you can say, yeah, that was probably causal, but in most cases I think it's correlation because it's it's multiple things that are happening at once.

Speaker 1: 11:16

The other thing is like just a recognition that marketing and sales data is just generally not going to be quote right. If I I always I it really like rubs me the wrong way, people will look at marketing and sales data reports or things that go. That's just not right, because I think like, well, it may not be what you expected, but also recognize that there's a lot of ways in which people don't have discipline or whatever that cause problems with that data. So I think there's a number of things layers to it from my perspective, naomi, but I don't think it's the like, actual just like can we get the data? It's like once we get the data, how to do something with it that's meaningful on the other side, where you can take action.

Speaker 2: 12:09

That makes sense. No, it was just. It was just a point of clarification that I just wanted to kind of discuss, because I do think that that's something that a lot of people struggle with, myself included. Right, it's that I will pull a lot of the reporting. I see the results. I make, as I best can educated assumptions on what I think has been happening and what needs to happen, because a lot of the tools that we use they're either going to ingest data or they produce it, and so some of them do both.

Speaker 2: 12:41

And when you have all of these moving pieces, sometimes it can be hard to pinpoint, okay, what exactly is moving the needle forward or back. And then the challenge that I have sometimes is okay, there's a lot of stakeholders and business partners that we work with senior leadership. How do we then effectively communicate in very succinct, short form, what is happening and what we need to do? And sometimes that involves, like monetary investment, sometimes it doesn't right. So those are all kind of things that I feel like marketing ops. People are constantly juggling what the answer is. I don't know. If anyone knows, I would love to know.

Speaker 1: 13:21

I think that's true. And so, shivani, given my sort of monologue there of more detail, do you see the same thing? I mean, I know you're probably not in your organization because you're there, right, but in other places that you've worked with, maybe when you were at Marketo with clients, curious, were you seeing that same kind of gap in terms of the ability to really find insights and value out of the data?

Speaker 3: 13:48

Yeah, and you know I have what is probably a controversial opinion on this topic, but you know, as, as someone who's built dashboards, as someone who's run teams that build dashboards, as someone who's run teams that have statisticians or statisticians or data scientists on them, like I've seen a lot of the data, I've seen a lot of you know under the hood, and so while you both were going mentioning your thoughts, I think two things occurred to me. Two things occurred to me. The first is that I think what you said, michael, about marketing and sales data and like kind of the cleanliness, like that was something that kind of picked up for me too, where it's we are never going to have a perfect, clean data set. There's a lot of, you know, cleaning, transformation of data that we want to do in order to get the appropriate statistical results, and if it's not statistical, then it might be. You know, we just want a clean data set. Like that's a completely valid requirement and wish, and so I think it kind of split things into two categories for me, which is one like sometimes we don't actually have to over-engineer, like sometimes, if we're looking at, you know, movement through the funnel, for example, I'm a big proponent of let's look at the conversion rate, like that kind of tells you everything you need to know. You know whether it's conversion from one stage of the funnel to another, whether it's an MQL into an opportunity, whether it's a MQA or a marketing qualified account into an opportunity. You know there's lots of metrics that we can use, whether it's a MQA or a marketing qualified account into an opportunity. You know there's there's lots of metrics that we can we can use and we can understand that the conversion rate is kind of the North star in a lot of them.

Speaker 3: 15:29

The second part of my perhaps controversial opinion is that I think oftentimes and again, this is this is like, this is home for me. So I'm I'm being very, I guess, vulnerable here is um, I think that there is this desire where everybody wants a dashboard and nobody knows what to do with one. Like you get the dashboard and you're like great, this is awesome, you did a really great job. This doesn't answer my question and um, and I'm I guess I'm laughing Cause if I, if I didn't, I would cry, and it's one of those things that I think it's just really hard to get right.

Speaker 3: 16:10

I want to, I want to leave room for that, like we're currently dealing with volumes data sets, like we've never really had that type of data before, even in the last five years it's, it's grown exponentially. You're talking about, you know, if you're talking about an enterprise MarTech stack, that's 20 platforms, maybe more, you know maybe 30. And so it's a lot of of data to manage and understand, like what's happening in this point of the data architecture. And then where does it, what happens? How does it transform, how does it flow through, how does it change? And plop into, um, you know, a data warehouse, or then, you know, into a dashboard, whether it's power, bi, tableau, looker, however you want to, whatever tool you're using there. And so my, my thought really is that I think, beyond the, the, the standard, you know, spiel of let's define requirements properly, so on and so forth, um, I think my, my thought process is like, sometimes you don't actually need a dashboard, like, sometimes what you need is a data set, and that's completely valid and fair, you know.

Speaker 3: 17:19

And I think it's about getting really creative. You know, we all have had experiences in enterprise, or maybe we haven't, and that's okay too, but I try to keep a very non judgmental approach of, like, just because it's not in a dashboard. Does that mean that this isn't authoritative information? Like, does that mean that it's you know if the data sets been cleaned, if it's been reviewed, if it's kind of followed a peer review process, if it comes from? You know the Salesforce extension, like you know, sometimes it's like we know where the source is, we know the cadence, we know how it was cleaned, we know when it will be updated and we know whether it's, whether or not it's a snapshot.

Speaker 3: 18:03

I think if you can kind of answer those questions, you're well on your way to having some data with which to make a decision, with which to draw insights, and I think it's sort of where a lot of marketers are, are craving that information, and it's also where I kind of see a lot of us kind of headed where it's like yes, this isn't a dashboard and that's where we get the big, meaty questions, the executive questions, answered. You know, there's that's where we're building our run, the business decks from, and also, like we can't forget about the practitioners, and sometimes the best way to really understand the insight is kind of digging into it. You know yourself and and I think marketers are craving that we're really smart. So that's my, that's my thought so far on that.

Speaker 1: 18:50

I mean. So what our audience can't see is that I've been vigorously nodding my head in agreement about like dash, like I'm not convinced dashboards always and cause I've been in places where they request, like we need a dashboard, and it takes forever just to define, like all the different spots in the dashboard, what format there'll be, and I'm like let's do one report and get it to where we want it. Then let's do the next report and if eventually we get to enough where we say, well, we can actually and people understand, like then we want to put them all together in one place, fine, do a dashboard. But my experience has been that like if you go start trying to build a dashboard, you almost never get it done to the point where people actually want to use it or it's useful to your point. But getting individual reports right that answer some questions specifically has been a lot more effective. And I would say the other thing is like it's just like when you, when somebody asks a question and they usually ask for data they don't actually ask, like they don't actually tell you sometimes what question they're trying to answer. Right, so if you know what question you're going to answer, you can hopefully be, not just a hopefully be not just a data dump. But I also think what I've experienced is you do the first analysis of a report or data analysis and you find another question.

Speaker 1: 20:11

So there's this iterative process that happens of going deeper and deeper and a lot of times this is the other thing I would say about a gap. I actually have a coaching client who recently we were talking about this and I was like one of the things that I've learned to to do. When I see data reports or whatever is, there's almost always something that's like it. You know, like it's not in the pattern, right, there's an outlier of some sort and everybody gravitates to that.

Speaker 1: 20:41

And I said if you're going to be presenting something like that to set of stakeholders who aren't familiar, they're going to immediately see that because we're all sort of trained to be pattern recognition animals, right, as humans. And if you don't already know that, they're going to ask like you should be ready to answer the question Like why is that one like that? Cause, usually it's like there's usually a reason like oh well, that we didn't code stuff on that campaign right. Or like there's a calculation wrong in the, in the spreadsheet or the whatever it is, but if you don't have the answer, people will start to question the quality of the all the rest of it, absolutely, absolutely, no-transcript.

Speaker 3: 21:54

But yeah, I think you're right. I think we're moving into an era, perhaps in the industry, where you know it kind of feels appropriate several, you know, in several types of questions of, as long as we can get the guardrails right and we know what this is being used for and we define what the limits are like, don't share this to an. I think it's really important to be aware, mindful, um again, nonjudgmental, about that type of ask and just kind of curious, open, honest, of okay, yeah, I see what you're trying to do here's, here's, here's an option, here's a couple options, and sometimes those are data sets. They're not quite and actually maybe I should clarify, like, when I think of a data set, I think of like it could be a Salesforce report, it could be maybe even a marketer report of some kind or a marketing automation platform report, bite-sized, like something that you know a marketer can can dig into and kind of they can create a pivot table.

Speaker 3: 23:01

They can, they can look at some you know quick trends, um, and and it's not necessarily being used at a broad scale or or for a you know highly scoped um, broad scope question, but it is helping them understand what's going on in their campaign or it's helping them understand like great. So my next decision is and I think that's where that's where it's kind of again, I've, I've built dashboards, I've run teams that build dashboards and like I think it is a fall down Like I would tend to agree, you know, in some cases with my stakeholders of like this isn't quite what I needed and I think we have to be kind of open and sensitive to that. This isn't quite what I needed and I think we have to be kind of open and sensitive to that.

Speaker 1: 23:45

Yeah, yeah, well, and I think you hit on something that I'm a big believer into. In fact it was a big part. I wrote a white paper for marketing opscom recently about marketing measurement and and sort of five different ways of looking at marketing from a measurement standpoint, Although one I would say people would probably go like is that really measurement? But anyway. But I think the key part of the key is like what is it? What is it? What's the who's the data and who's that report intended for, right? So if you were to take that low level data and try to put it in front of an executive, right, probably the wrong audience, you know.

Speaker 1: 24:26

So I think that's the other thing Like, think about who this is for and how, like, make it easy for them to consume it. If it's not going to be easy for them to consume it, then you either need to do you've got the wrong kind of report or you or you need to think more about like how do you, how do you take that and abstract it to a way that somebody who's not as familiar with the way the sausage is made right To to consume it and understand what the implications are. Yep, so that's my thought. Okay, so, if so, if my assertion is right and it sounds like there might be some question about it, maybe, but that there's a gap, like how can people start to build up on some of those skills you know, to start to build those muscles, that skill, the experience, the knowledge for people to take data and then turning it into actionable stuff? Not that everything has to be that way, but what's your thought on that?

Speaker 3: 25:26

Yeah, I mean, I think especially around like closing knowledge gaps, um, I think is is kind of where my my head is and I would encourage people to be scrappy, which, you know, not everyone loves scrappy, and that's okay. But I think, from my perspective, like you don't necessarily need to go, and like you know, I've gotten questions from friends in the industry or even coworkers of like, does this mean I need to go get my MBA now, because you know that's a pretty intense thing on, or, you know, a master's in analytics or statistics or data science? Like there's so many routes we can go now, um, you know. But also, if you're you know you've got a busy life. You're juggling work and family and, um, all other types of commitments.

Speaker 3: 26:28

I think there's lots of things you can do to kind of close gaps and try to view it as like a bingo card, essentially of like, what are? What's the bingo card of things you want to be able to do? Um, and so I think one of the things there is just understanding like really basic functionality, um of Excel, and that feels like maybe like a 1990 answer, but, um, I think from my perspective it's like it's actually really helpful, like that has been helpful, um for me to just like visualize, like I. I sometimes need to stop and pause and outline and kind of sort out like what am I trying to do with this data before I just plunge into it? Um, which is helpful too and great sometimes, but a lot of times we need to organize our thoughts and kind of outline the query or the process of what we're trying to get to. And understanding pivot tables, understanding a VLOOKUP, those kinds of things. And if you can kind of run in your head like this is what's happening, it's running down these columns, it's running down these rows or it's it's formulating the data in these ways, I think is actually really foundational and very, very helpful to just understanding, like what's happening in a massive you know in, like you know if you're, if you're looking at something, um, in a data warehouse, whether it's snowflake or data bricks or a larger format, much, much, much larger format. But I think it's important, like it's easy. It's easy to get overwhelmed by the here's, like the the big, massive thing, but we need to start small like, start small, review, kind of understand like what's, what are like some things that you normally have to do, analyses, even that you do with marketing data.

Speaker 3: 28:18

The other thing, um, and then that I mentioned on a podcast several years ago that I was on, is marketing data is, um, I think, foundationally it's got lots of different types of data categories in it. It's got categorical variables, so they're not easily transferred into like a numerical value. It's got numerical variables and, um, you know, sales data is a little bit cleaner in that respect, like a lot of times it's just deals, acv, um, average contract size, you know we're looking, or average deal size, like we're looking at numbers and things that can be very easily compartmentalized. But marketing data isn't always that way. We've got, we've got words, we've got letters, we've got things that we can't, um, you know you can't easily convert and sometimes you can convert them, but you have to be mindful of that.

Speaker 3: 29:07

And so I think it's like understanding those bits of and and kind of broader, um, not requirements, but like broader strokes of marketing data, and then arming yourself with like, okay, how does how does it work if I want to do this type of analysis, or X or Y, and then those are the building blocks that you can then use when you're looking at something like a linear regression, which, um, if anyone doesn't know, I'll just mention it now. It's like the best way I've used to describe this is that it's kind of like the slope equation in high school, um, but like even fancier and better and and it kind of it's. It's kind of a nice way of entry point of um understanding what it's doing. But, um, I think that it's like very easy to get overwhelmed and I see marketers and people go down this rabbit hole of like I got to do this big thing. And it's like if you're going to just start to close some gaps, start with like start small and then you can kind of build to the bigger, bigger pieces.

Speaker 1: 30:11

Yeah, I mean it's. It's what's running in my head is like there's sort of two, two categories of how to do it, like one is like learning how to use the tools that we probably all have. Like probably everybody has excel or, if they don't, they have google sheets right, and there's a lot of overlap in there. There's a lot of function, a lot of functionality that both have that most people probably don't take advantage of and it takes time to figure that out right, but there's tons of online. I was actually looking because my oldest was going to go into his freshman year in college and knew that he was going to have to start doing spreadsheets. I was starting to look for like stuff to help him train. Like Microsoft actually has a ton of really good like building blocks on how to use excel, and I know there's other third-party stuff the other one that so there's that, I think.

Speaker 1: 31:00

Then the other category to me is, um, the more I guess general kinds of knowledge, so statistics, linear regression, like more topical, and I found like I I actually use this a lot with my kids as they like helping them. It's like like con academy is great, right, it has a ton of it's free and it's actually really good at some stuff. Now they don't have everything right. They may not have linear regression. I'm pretty sure they have basic statistics, because if you don't know the basics of statistics, like go there, it's like you'll probably learn something. Yeah, and are you to?

Speaker 3: 31:37

me or YouTube.

Speaker 1: 31:40

Yeah, I haven't, I haven't used you to me, but YouTube for sure, YouTube university. That someone said to me the other day like, have you like? Do you? Can you think of any other? Like, have you done this with any of your team or yourself? Can you think of any other? Have you done this?

Speaker 2: 31:56

with any of your team or yourself?

Speaker 1: 32:01

What specifically? How have you learned or helped your team learn some of these things, or have you? I don't even know.

Speaker 2: 32:09

We kind of just learn as we go.

Speaker 2: 32:10

A lot of times we will get questions where we kind of just let us get back to you and then we convene as a team and we try to figure it out, whether it be asking our peers, whether it be, you know, googling for information, watching YouTube videos, finding resources online that people have done the same or similar things.

Speaker 2: 32:30

It really depends on what is being asked. I feel like a lot of you know on the topic of reporting and analytics. A lot of the time it's kind of a hybrid for us. I would say right, it's in the sense that sometimes we'll be asked to produce reports and there's other times where we are creating reports based on things that we feel like people should know. Right, where there may be that gap on the business side, where they don't know what they don't know, and we're like, hey, we're actually seeing X, y, z, and here it is shown in a way that you may not have originally thought of, but I think it's important that you ingest this information and here's why, right, and then those become standard sets of reports that the business uses.

Speaker 1: 33:17

That's great. I like that. I mean then you're sort of sharing. We could all send emails to Shivani right? You know when we get stuck on something.

Speaker 2: 33:28

How do you do this? No, but I mean but seriously I mean the marketing the community is great for this. I just asked chat GPPs yes or perplexity?

Speaker 3: 33:44

I've been really on that right now. Well, okay.

Speaker 1: 33:52

So you brought up the topic of of AI, cause I I so I actually do my. My prediction with AI is that this like current sort of big bump in seeing AI as a, as sort of a both a threat and a help to content for marketing, I think that's going to wane because of some of the like it's all almost all centered around like chat, gpt type things. Like it's all almost all centered around like chat, gpt type things, and I think the the risks of like um IP being lost by using like putting it in there and try to generate content is going to make that slow down a little bit. I don't know if I'm right or not. Well, where I do see where it has could have value fairly quickly. I mean, shivani mentioned it.

Speaker 1: 34:30

Right, we're all generating just volumes and volumes of data. Every patterns that it's finding that individual humans would have to take time to pull the right set of data. Do the analysis, test the hypothesis, not right? There's just not enough time Now. To me, that also implies that you still need people who can interpret the output of that, and so I still think there's the need for that. So I wasn't just rolling my eyes at AI. I think I was, but it was just. I mean, I'm like I think I just don't think it's going to eliminate the need to have people who can understand it and interpret what comes out of this.

Speaker 2: 35:35

I think it's more supplementation rather than full direct replacement or automation. Right, I feel like it's going to become a really important, integral part of their day to day in the business that you know will. Just, it'll be like a tool.

Speaker 1: 35:54

What's your thought on that, Shivani?

Speaker 3: 35:56

yeah, I mean I, I think I'm quite aligned. I think you know. If it's a question of replace, I don't think you know I mean yeah, call me crazy, but I don't.

Speaker 3: 36:08

I don't think we would get into like a replacement territory, but I think there will be a little bit of displacement while we're getting used to the new normal. I think I think one place where where um AI will be very, very helpful is in that, like we need an initial data set, like it doesn't have to be um perfect, but path of operationalizing that pretty quickly. Um, especially with you know, with all of us who, like we it was funny my husband the other day we were at this like happy hour and with some work, uh, friends, and you know we were saying, like do your spouses know marketing terminology? And I looked at him and I was like do you know any? And he goes M, q, l and I was like, okay, great. So you hear me say that. Enough Um, but I think you know that's like a great example. Like we, we all know our metrics pretty well. We were developing new ones Um, but I think that the um.

Speaker 3: 37:24

Sorry if there's some background noise there but I think that the the kind of piece there is, like once we need, once we get to, okay, this is kind of the operational piece. Like we know that we need something to iterate on. Um, we all know, and a lot of our stakeholders know, our terminology and I think AI can be really helpful in determining, like here's your, here's your starting point, here's the standard starting point. You know that you want to, you want to move from, and then the rest of it is kind of a Candyland map of like I think we is still very critical and helpful, and I think that that's where I'm going to, in my opinion is we would see folks leaning in, which is nice, because I think that's what we all kind of want to do anyways, Like it's nice to be able to focus some more creative energy than it is to maybe be slogging through a data set that you know could maybe be generated quickly.

Speaker 1: 38:28

No, I think that's right. It doesn't necessarily eliminate the need. It sort of shifts what you need to work on to stuff that I think, in general, most people would find more interesting and challenging. But you know, look, we're all pretty terrible about making predictions, like so who knows what it's going to be? But, like I said, I'm kind of hopeful that that's going to be the case, right, because I think, with that like sort of the challenge of not having that, the skills to really interpret it, kind of pull together, interpret, you know, analyze data that I see plus just like the amount of time it takes to do it well, so even if you have that skill set, you don't have the time to do it right, it's going to be a challenge to get any value out of that.

Speaker 1: 39:21

So if there's things that can help that, right, we help drive people who are going to be like be able to interpret the data, and then we have the things that can help that, right, we help drive people who are going to be able to interpret the data, and then we have the tools that make that happen in the background, so that you're able to make progress on other things that are, I don't even want to say, more of a priority but time sensitive, then I think there's some value there. That's my hope at least. And then the Terminator will come, and while we did, I just watched.

Speaker 2: 39:50

I just watched uh, I can't remember which.

Speaker 1: 39:53

It was one of the the newer ones last night, so oh yeah yeah, like sometime in the last year, I watched terminator 2 with all my kids. I think it's like it was the one that was on netflix or whatever. I couldn't get the first one, so I think we went back and watched it, wow the CGI is so bad.

Speaker 2: 40:13

That was an explosion. What was that?

Speaker 3: 40:17

I think that's what our kids are going to tell us about our data sets that AI created for us right now right, it's going to be crazy, yeah, 20 years from now. Oh, mom, that's a silly, that's a silly data set.

Speaker 1: 40:32

I can't believe that was an explosion I mean, I still remember y2k being a big thing and that was only 24 years ago 25 years ago, yeah, so yeah, that was a big deal and I thought you're only 23 much ado about nothing something like that some some, some multiple of that.

Speaker 1: 40:56

I missed it. No, I didn't. Um, yeah, oh, I get it. No, I get it. It now I get it. All right. Joke's on me. All fair, all right. So is there any other? Like are you seeing anything or hearing about anything, shivani? Uh, that's going on with data ai platforms. I'm I'm hearing bits and pieces and there's a couple of people I talked to who are closer to it. But, like I said, I'm hopeful, but I haven't heard of any specific tools.

Speaker 3: 41:29

I have heard about a tool that someone's working on that does involve doing something like attribution, but using a regression analysis type of approach, as opposed to traditional attribution, which I think is interesting yeah, yeah, I mean, I think the one that comes to mind that I think probably for you all too is like stack, moxie, like they're doing some really really great work and um, you know, I think, I think, there there's, they're probably the, the main one I would call out in terms of um, just how well thought out the, the marketing use cases, and like how they're just so much great documentation and like I think, if you want to, if that's quite honestly where I go, like if I'm trying to think through here's my whole architecture and like here's, here's the thing I'm trying to solve, like here's the use case, um, I think they just have a really nice framework marketing um specific. You know I've seen a couple of really interesting case studies OpenAI, I think you know there's some really great use cases and like thought and even like kind of strawmans being built out from that. You know Workato is another one. You know Workato is another one. We just had a Marketo user group a few weeks ago at the Workato office in Denver where Michael Van walked us through kind of how he's using AI in workflows and it's really really fantastic.

Speaker 3: 43:03

I did do a bit of research before this to try to sort out if I could find, like I was like is there a vendor that is doing this thing that I think would be really helpful? And so it's not quite AI or maybe not yet I don't know the word near and dear. Maybe like also could bring up some like oh God, like some anxiety, I don't know, but it's slide decks, like I think that's another area where it's so important to have visuals and like we want to be able to communicate clearly and I think oftentimes for executives that's kind of the format that they're looking for, and so I could see there being um, especially with a tool like roles tech, where it's integrated with your, you know, power BI, tableau, um, and it's just kind of automatically creating you um, some slides. I don't know if you all ever watched um Dave Chang's Netflix show ugly delicious, but if not, I'd recommend it and he has a point in there.

Speaker 3: 44:11

He makes about like he loves food that's like DIY, homemade, it's food that's like you're not going to find in a fancy restaurant. It's not like picture perfect, it's ugly, but it's delicious, it's got good nutritional value, it's delicious, it's got good nutritional value and like that's kind of my thought process too on a lot of how I think AI, um, platforms, tools will start to, to to develop in this space, because I think, again, it's like we, we all need a place to start. Sometimes starting is the hardest part. Um, it's easy to kind of get distracted when you've got chats and things and emails and texts and calls from daycare whatever coming in, and so I think that that's going to be a value add of like here.

Speaker 3: 44:55

We started it, we started the slide deck. Is it perfect? No, is it ugly? Delicious, probably, and we can start to iterate from there. You can take it and run and I think that's like I said, I didn't actually see if I didn't see AI mentioned on Rolstech's website, so I don't necessarily want to call them out as such, but but you know, it's like tools like that that I see as like okay, this could be, this could be a really great potential here.

Speaker 1: 45:23

Interesting, ugly delicious. I'm going to go add that to my Netflix queue now Go check that out. I love that. Wow, my head is spinning now. So, Shivani, this has been great. We covered a lot of ground and all of a sudden, it's like we're almost at a point where we have to wrap up. But is there anything that you wanted to make sure our audience heard today that we didn't talk about?

Speaker 3: 45:51

um, I don't, I think, for maybe, yeah, you know, one, one item I would say that I kind of mentioned in a lot of different formats is, um, I think as like marketing, ops and analytics people um, it's really important to just stay grounded and like empathetic, like I think for me, one of the the I've had people come to me before and kind of be like I know you're really smart and you get this, but I don't, and if someone's saying that to me, I'm like oh God, I've failed, like I don't want that to be the case. I don't want people feeling like I have this information and they don't, and it's just like I think that that's kind of my takeaway, even with with all of this, is like I think there's lots of great ways to um, to partner with stakeholders, to partner with different functions, and not that it's always like perfect and I don't know kumbaya, but um, but you know, I think it's. I think like having that empathy and having that kind of curiosity and openness is really really important, especially in this space.

Speaker 1: 46:53

Moving forward, yeah, I think that's a great call out. All right. So, shivani, we are about to wrap up. If folks want to keep up with what you're doing or connect with you, what's the best way for them to do that?

Speaker 3: 47:08

You can find me on LinkedIn.

Speaker 1: 47:10

All right, linkedin it is. I think, like 99% of our guests say that that's okay. Um, awesome. Thank you so much, shivani. This has been great. Love this conversation. I hope I hope our audience uh great love this conversation. I hope I hope our audience uh takes it to heart. And those who feel like they've been intimidated from all this stuff uh found some nuggets here to go build their their knowledge. All right, naomi, as always, enjoy being with you here fun times.

Speaker 1: 47:40

I wish it was cooler here all the time obviously right always, always in my 23 years in your 23 years, that's right, yeah yes all right. Well, thanks also to our audience. Thanks for supporting us uh providing your feedback and any suggestions on topics or guests, or if you want to be a guest uh have a topic, please let us know. We are available on linkedin, on the marketingopscom Slack or the marketingopscom website. Until next time, everyone, bye, bye.

Speaker 3: 48:12

Bye, thank you.