Customer Success Playbook Podcast Season 2 Episode 21 - DataPlant The Game Has Changed - Samuel Cummings

In this episode of the Customer Success Playbook podcast, we're excited to host Samuel Cummings, CEO of DataPlant, who is at the forefront of revolutionizing customer success through AI and data science. Samuel shares his fascinating journey from being a data scientist at Gainsight and LinkedIn to founding DataPlant, a company dedicated to tackling the challenge of scaling personalized customer engagement.

Join us as Samuel delves into:

• The evolution of customer success from reactive support to proactive, data-driven strategies.
• How DataPlant synthesizes complex data to create hyper-personalized messaging.
• The critical importance of digital empathy in managing a growing number of customer accounts.
• Real-world examples of DataPlant's impact, including their work with Restaurant 365.
• Predictions for the future of customer success and the transformative role of AI.

Samuel's deep expertise and innovative perspective offer valuable insights and practical strategies for anyone involved in customer success. Don't miss this engaging episode that explores how DataPlant is changing the game and what the future holds for customer success.

HTTP://www.dataplant.com
https://www.ai4diversity.org/podcast/episode/1a1b4367/ai4diversity-meets-aura-or-ai4diversity-podcast-32

Please Like, Comment, Share and Subscribe.

You can also find the CS Playbook Podcast:
YouTube - @CustomerSuccessPlaybookPodcast
Twitter - @CS_Playbook

You can find Kevin at:
Metzgerbusiness.com - Kevin's person web site
Kevin Metzger on Linked In.

You can find Roman at:
Roman Trebon on Linked In.

Transcript

Roman Trebon: 0:30

all right. Welcome to the customer success playbook podcast. I'm Roman Trebon and with me always is my cohost, Kevin Metzger. Kev, how are you doing today?

Kevin Metzger: 0:41

And I'm doing great. I'm excited about today's guests and excited about plan hat and learn more about it. So what do you think?

Roman Trebon: 0:48

wild. Without, well, I'm also excited about these awesome new Polish shirts we got, which, you know, again, if you're, if you're on our LinkedIn page or our YouTube channel, not only do we got Sam coming on, we got the new gear. We're repping as well. So. Yeah, this will be a great episode. Kev, why don't you introduce our guest and our topic for the day?

Kevin Metzger: 1:05

Yeah. So we've got Sam with us. He's a data scientist who's been in customer success at Gainsight and LinkedIn and now started a company focused on AI and customer success called DataPlant. I've got I've got to listen to Sam at the Gainsight conference and I'm really excited to have Sam on the show with us today. Sam, can you give me, maybe give us a little bit more about your background and how you got started with DataPlant?

Sam Cummings: 1:32

Well, gentlemen, I'm going to need one of those shirts. Let's start there. That shirt is, there we go. Merch game is popping. Well, Sam Cummings, CEO Datalant. I just cut my teeth at a time when customer success was in his second wave. Gainsight was founded. There were companies out there like churn zero to tango, et cetera, but companies were figuring out how do we build a data driven practice? And that's why I help companies do. And so I traveled the world, I was head of data science at Gainsight. So I work with companies like DocuSign, Marketo you know, RingCentral to realize the application of data science in the customer success space. So everything from predictive modeling applied to revenue retention, you know, reducing churn producing opportunity reports, the whole nine yards. And I tell you, I learned a lot of gems, and I took those things, then I went and I actually became a CSM. And I tell this to anyone who's building something, walk in the shoes of the people you want to serve. And that's what I did with my career. I then went on to LinkedIn to become a CSM, and I tell you, everything I learned as the data guy. I put the practice as an actual CSM, and that gave me a wealth of, a wealth of experience that doubled down on what I had before from a data perspective. And then I went on to go to Indeed to become Director of Implementation. So I've been in every seat from individual contributor, technical person, data science director, and that gave me such a perspective that I used to build this guy. Datalant.

Roman Trebon: 3:08

You got, you got, you got some polo game as well, Sam. I like that polo shirt as well. We're all, we're all looking good here.

Sam Cummings: 3:14

Just go to the website, get the shirt y'all, subscribe.

Roman Trebon: 3:18

So Sam, talk to us a little bit about Datalant. What is data plant? What do you do? How are you helping people in the customer success world?

Sam Cummings: 3:25

Yeah. And this is really just a part of what's going on now in our time. There's been this premise, the idea and the way customer success teams generally approach is. I'm going to lose business in my long tail. So my smaller accounts in my mid market accounts, and I'm going to make those gains with my hands on enterprise accounts. That's the nature of the business. Everyone expects to have higher churn in that segment. That was great when there was money printer season, and you could just try to make it to your next round. Now, companies are being forced to grapple with that problem. And if they want to really drive their business to the next level, Everyone has to solve this scale issue of being able to engage in their smaller segments. And that's what we're born out of. It's solving really one of the biggest problems of our time. Digital empathy. As companies are looking at more customers that they're managing per CSM. I think a phenomenal speaker I know as well named Irit. She has a CSM practice. If you don't know her, check out her podcast. She has shared some insights when she was speaking at a conference in Israel about how there's just a ton of innovation that's happened. But one of the biggest metrics that shifted over the last decade is the number of accounts per CSM. We've gone from 20 to 30 accounts per CSM to 40 to 60. And now we expect it to get up even higher. And so regardless of if it's enterprise, mid market or S and B. There's going to be a need to scale like never before. And the problem underneath that hood is when you're one to one, you can empathize with your customers because you're talking with them, you're engaged. But when you lose that one to one, you become one to many or one to a lot. That's where you have a challenge of, can I empathize through these digital channels? And that's what data plans helping companies with is not, Hey, I can't hire a bunch of CSMs to engage all these people. Even if I wanted to, I don't have enough to hire a hundred people. So how am I going to allow for the people I have to be great? And that's what data plan is doing. We're helping companies synthesize their experience of their clients and create hyper personalized messaging so they can no longer just be transactional at their segments that are smaller. I

Kevin Metzger: 5:33

love it. Yeah. So going down that path, what kind of information is it that you, what kind of data is it that you're gathering in order to try and basically get the information you need to have and create the digital empathy?

Sam Cummings: 5:48

Yeah. So I conducted a study and this is kind of going into my role here. Before we started, we were starting and synthesizing what data plant would be. We did a study of and reached out to about 200 COOs. CCOs and then CROs of like what kind of data they have. And you'll be surprised. The main data that people are using today is two sources. They're using support ticket data, which represents the first wave. The first wave of customer success was all reactive and it evolved out of going beyond customer support. And then essentially the second wave was product usage. So being able to look at product usage data, whether that's the Google Analytics, Pindo, you know, Mixpanel, all these tools kind of arose during that time when it was about, okay, what are people doing in the platform? So we take those sources as well as your CRM data and any other sources that you might have, marketing, outcome data like renewal history or churn history, and we bring all that data together. So if you have complex data relationships, you have multi layered hierarchies like regular child accounts that roll up to parent, that roll up to grandparent. We stake our claim in working in complex data scenarios and really synthesizing that into useful insight.

Roman Trebon: 6:59

So, so Sam, I don't need, I don't need to be a data scientist to consume the output, right? You're going to do all the heavy lifting and get all that data together. And then you're going to help guide me as a CSM. You mentioned 60 to 1 ratio. Oh, I'm getting, I'm almost breaking out in hives hearing that. But this output is going to help me be able to like, you know, scale and touch all these clients that I'm responsible for.

Sam Cummings: 7:19

Yes, because this is a revolution happening under to under between our eyes, right, right in front of us. The real opportunity is not something that's in a vacuum. We have been on a long arc of innovation from in the beginning when I came out of college. I'm a date myself, but I was S. A. P. Certified. And back then, you had the job with JavaScript, code the relationship between objects, code the jobs that do transformations. So if I wanted to convert a lead to an account, or a lead to a contact, I had to script that. But then there's this tool or this platform, some people might know, there's a little bitty company called Salesforce. That really took the wave when I was graduating. That was the same thing that Salesforce did for technology and ERPs and CRMs. The same thing happened with WordPress for web development and blogs, right? The average person can now do it. You didn't need a developer and that is what we're doing in this next wave that's happening. Cause even now, when you want to use a Salesforce or anything like a Gainsight, there's whole trainings, there's whole certifications because these systems are sandbox tools, which again, is better than coding, right? It's better than the JavaScript days. But you still have to be certified in understanding how to make these tools do any one thing. We're in a new age, and DataPlant is leading that age, where the professionals themselves will be able to drive their business. And so imagine, you could just talk to DataPlant and say, give me a report on download performance and put it in a bar chart format. And instead of like, writing the SQL, that's what most of the tools are still doing today, even with AI. They'll just write the SQL for you. But you still gotta know what to do. Here is, you can just talk to DataPlant and it'll, since it auto generates all the possible charts, think of it like a tree with the fruit or the insights and you're just curating with your voice and what you look at and passively as you leverage it. This nice curated bush of fruit that can bear fruit in your business.

Kevin Metzger: 9:17

Sam, from, can we take that down from an analogy level, a little deeper into what's actually happening with the data and how you're, how you're synthesizing it?

Sam Cummings: 9:26

Yeah, so real example, customer we work with called Restaurant 365. 40, 000 locations, so they're a point of sale system for restaurants. We're helping them analyze their customer usage data because they have a lot of products. So think about it. You have maybe four to six products that you might have. Each customer might be in one of three segments. So it could be an SMB client, a mid market, and an enterprise. Each of them have different desired outcomes based on who they are and what they need from your platform. And so we're able to analyze that based on what the size of the company, so if it's an SMB client and then what they bought. So do they buy the operations feature or the accounting feature? And so that tailored marketing of being able to engage them and communicate is what they leverage our tool for. So they can use, their CSMs can reach out to a restaurant and say, compared to all other pizzerias. You are spending 25 percent more on labor. If you leverage our payroll tool, you can lower that burn and really get a more efficient business and put yourself in the top performers. To do that in any other CS platform is impossible. The nature of the data is just too complex. You know how many restaurant types there are? Seafood, pizza. So imagine a rule chain that you would build to really automate that, which again, that's the underpinning of all modern CS software, which is rule building. And so we saw it like that itself was the barrier to unlocking these personalized interactions and these tailored insights. And that's what we did. We moved away from that and really transformed how it works to where now you can, as an individual contributor, as a leader, you don't have to pay the tax. That was required before. Before you had to pay a tax essentially of building out everything, going and finding out what's working, then defining the playbook and rolling out. We're duty free, baby.

Roman Trebon: 11:19

I love, I love that example. And like you said before, it wouldn't have happened. You didn't have the time. You didn't have the resources. You didn't have it all there to say, Hey, you know, your, your, your labor cost is up in X, Y, I, you know, look at these solutions. So for our audience, if you haven't gone to the website yet, you have to check it out. So you have to go to the Datalant. Com. Sam, you already have an amazing. Set of free tools on there. So I found this, what we had, we're getting new year and have you on the show, checking you out. You have these free chat, GPT bots on there. I've, I've already sent them to a whole bunch of people. I absolutely love them. They're game changers. How, how'd you guys come up with these? What was the concept behind this and tell the audience a little bit about what they are and where they can find them.

Sam Cummings: 11:59

So this is where we really saw the future coming forward. And I tell you what's coming next is going to be even more fun. I'll give you a sneak peek. I'm dropping the sneak peeks on this podcast only. So if y'all didn't hear breaking news, this is exclusives. So that's it in the background. So we saw that the main approach to this has been, we're going to create these all knowing bots. So if you look at most of the other companies, everyone's got a GPT out there, but their idea is like a chat GPT tool can do all things. What that leads to is they don't do anything. Well, And so what we saw was what's better suited is have smaller, more targeted modeling tools that solve the top problems. And we went and did some research as well. Big, you know, data guy, surprise. But I researched what were the top use cases of CS teams. And then we look to build the GPT around each of those. And we really tailored them. And then I know a lot of cheat codes around building GPTs, the Beck perform higher, once he caught, I'll give you two of them. If you tell a chat GPT to take a breath, that alone will allow it to be more diligent. I know it's not human. It's not going to literally go breathe, but that'll slow it down and make it think more deliberate. Another trick. If you tell it to proceed without instruction, it will allow it to make more assumptions. Based on what he knows in his premise, that it probably wouldn't have done before because he didn't trust that it would be accurate. And so those are just some of the cheat codes coupled with the learnings that we had over the decade of working in this space. That's why our bots are more powerful. And then we built them into the main core use cases like customer success transitions, big problem transitioning account. Being able to do customer success strategy planning as a leader. How do I plan how I'm going to approach my business? So we targeted these very niche problems and that's why you see that our free tools are way more high performing But i'll stop there before I give you guys the cheat codes for the future

Roman Trebon: 13:51

Oh, i'll tell you what. I so just just our audience knows I went there, but you have a success a success plan bought It's on there. Like you said, it's just that we had just met with the client first time. Tell us about your business Da da da Put in the transcript, put it into your success bot plan, spit out an amazing first, I mean, draft, I mean, an amazing success plan that we had within minutes that would have taken us, you know, you know, before, before the bot or even trying to do it ourselves, like you said, it would have been iteration, iteration, iteration. So Sam, you already saved me some time. So I'm all in. I can't wait to hear what's coming next. This

Sam Cummings: 14:27

is where Sci-fi kicks in. Kevin, let me share your thoughts before I jump into it.

Kevin Metzger: 14:30

I was gonna say, I want to take a guess. Tell me, if I, tell me how close I am, you're gonna have the, so with the bots, as you build out individual bots, is it that they get to, they'll start interacting with each other to basically build out the plans and the functions going forward?

Sam Cummings: 14:46

You got it. This is why this man is really, I'm telling you, I wouldn't have came to no other podcast, but this one, I mean this sci-fi kicks in gentlemen. Let me just take a step back in the concept. There is no rules. This is the wild wild west. There's no one that can tell you do this, do that, do this, you get that. And this is what this looks like. What we're going to have, and again there's some challenges with the price and the cost of these bots to be able to make it scalable, so there's going to be some time to this, but the other side of this is you're going to essentially pose goals To a chat room. In that room, there'll be six or seven different GPTs, and they'll all work together. And I've been experimenting with this technology already and working what we're doing with Datalant. And what we found was crazy stuff. So I put all those bots that we had already in the room to solve a problem. And again, the statistics have shown us over the last two or three years, the first version of GPT, they had about 80 percent efficacy across a series of human tasks. The more tailored version that people have been working on lately. Up in the maybe 90s. These new community approaches of bringing in multiple GPTs together. 98 efficacy above. Think of it like ants, right? Any visual ant can be smart, somewhat, but together they have emergent behavior. That's happening right now. Now what's crazy, we're putting in different types of GPTs that have no correlation with the other ones. So for example, I put a police bot that does interrogations in there, and the performance went up. So what that means is, there's no rules in what kind of bots you can combine. You can bring a bot that's a counselor bot. And have it talk to the other bots and it might actually improve their performance. And so that's the world that's going to be brand new here is it's not just like, I need all my bots to have some similar relationship, meaning be customer success. You can put six bots in there that all have a similar relationship and it put a wild card in there and get a 10 time better performance. The game just changed, gentlemen.

Kevin Metzger: 16:47

It's effectively, you're kind of driving an AI hive mind, is really what you're trying to, you're putting together. Is that accurate?

Sam Cummings: 16:54

Yeah. Think of it like one of those that might not be as, you know, they're like, well, that's so ethereal. You've, anybody ever played Uno?

Roman Trebon: 17:00

Yeah. Oh yeah. Oh yeah.

Sam Cummings: 17:01

Same thing. Each bot puts down his card. So it puts down what it thinks. Every other bot knows the history of what's happened. So they can all pick up from each other, edit it, and then drop the next card. So we're all essentially doing it essentially in the box and it's getting super fast speed of light, but they're playing UNO together. And from that, you get this emergent, better game.

Roman Trebon: 17:21

Love it. So Sam, you've been, like you said, you, you've been in, you've been in customer success for a while. You know, you're, you're in, you know, Gainsight, LinkedIn, you've been all around, right? So we talked about AI a little bit. Where do you see, what is customer success look like five, 10 years from now? I mean, what are we looking at? What do you think is you have to get your crystal ball out and dust it off a little bit.

Sam Cummings: 17:43

So for those working and I'll talk to a specific sub audience here because you are going to be the heroes of our future customer success operations. Those are the folks that have an opportunity to make sci fi real the way we've been thinking about customer success, not going to move the needle. We're going to have to go to a new level. The reason why is the thresholds higher. We're already seeing this in sales and marketing code outreach dying. Email spamming, not working. Good luck. Even reaching out on LinkedIn cold. Try it, see what happens for you. The issue is, that cold outreach type of mindset without personalization doesn't break through. And so that's where the future of this, what I see is, Is these hyper personalized engagements where again, I'm reaching out to people on a way that know from their perspective, they know that that message was tailored for them. If you can't tailor your message in today's age, when everyone's using these chat GPTs and our email inboxes have been blown up for the last 20 years, you are not going to be able to break through the noise. So as CS ops teams, that's the first level of where I see this innovation happening, where they're going to be the creative ones, they're going to become rock stars. Because they're going to have more ability to touch their clients than anybody else in the business. If I'm a CSM, I might own 20 accounts, 60 accounts, etc. Through digital customer success, these folks are going to impact millions of people. And so the way that the industry goes, if we decide to go further and be a little bit more risk taking in how we approach this, oh man, we're going to be the real, real heroes. So that's the first group that I see really benefiting from the innovations today. People that are going to be leaning into digital scale. Now, the other group that I think is really going to be the beneficiary on the broader sense is the medium to low performers. There was a big gap between those that were great at doing their job in CS versus those that weren't, that was really bucketed into the soft skills. Can you take good notes? Can you do good recapping? Can you distill your points into really clear messages? What's your follow up game like? Those all are now going to be democratized. So the distance between someone being really good on those skill sets and the average person is going to be way shorter. So that's going to allow for all teams to have really a higher performing group across the board. I'm excited for that time. And again, I'll share more as we go forward on some of those unique niches where the opportunity is. But overall, CS Ops is going to have more touch and become more of a hero and individual contributors are going to be more democratized in their performance.

Kevin Metzger: 20:21

Yeah, and I think the ops right now is huge. I think that's where you're, you're, you're getting a lot of the games and allowing the CSM teams to basically expand what they're serving. From we hear a lot about it, you know, it's not those who Use a error. It's not going to replace humans. It's humans using AI. That's going to replace humans not using AI. But I think as we talk about it, the AI is going to allow a human to scale much work. Right? So, and ultimately, it's hopefully as humans continue to scale. Outreaches and what they're able to do and how they think they'll be able to keep up with it and we'll grow more customers. And by growing more customers, we're going to expand. You're going to grow. You're going to grow the pot. And I think that's where there's there's an advantage. I do think that. As CSMs or CS ops folks, you got to start thinking about how you, what, where you're bringing value. What do you think on that? Like, where do you recommend people start thinking about how to think differently? So they're ready for the future.

Sam Cummings: 21:30

Yeah. Just throw away everything you've heard. All that stuff that worked in 2012 is not going to work today. You can show up with the same things you were doing then and be walked out of the room. And what I mean by that is. The approach to customer success was top down. The idea was we could just figure it out, we know what works and then we're going to make essentially and the reflex is the system has a rule that says if less than six logins do this, if more than 12 you know, interactions do that, like that was the core idea. Top down. The CSM was essentially was the meat bag executing on the systems intelligence. Innovation is too fast for that. We saw that in COVID. I tell you, I was in the seat, so I got the seat as first hand. When COVID hit, all these folks had all these rules that all of a sudden broke. If you looked into their cockpits, whatever tool they're in, into where they had these notifications for their teams, they had thousands of upsell records that said, this client's ready to upsell. Did everybody all of a sudden ready to upsell their business? No, since people were indoors, they logged in more. And so that was a perfect example of like the idea of I'm just going to hard code everything and then have it be top down in my business only works when stuff doesn't change. Now we might not always get a COVID again, again, something like that, that transformational, but what is true products teams are evolving their products faster than ever. So say I change how a feature works, all the rules I built around that now are no longer the same. And so there's just an infinite problem that we're going to deal with regardless Which is the rate of change in our markets. That means we need to switch that premise of top down CS to now a more agile version. And that's what data plan is enabling. Think of it simply put with an analogy here before it was turn based, we configure everything, we execute that configuration, what data plan switching to is real time strategy. Since we've removed the duty to pay the tax. Now, there's no issue for you to execute. If you on the Monday morning, think about, hey, we want to drive improvement of the downloads feature, you can execute on that same day versus having someone go build out all the reporting and then do analysis of what's working, then define the strategy, what we're going to deliver. We're able to deal with that in minutes. So now you as a leader can execute. So what was thought not possible, I can only run one campaign a quarter because it's going to take so much to mobilize or I can't run what we call perishable campaigns because we don't get certain things are perishable, meaning it's so short in the moment that it wouldn't make sense to automate it. It's like, for example, if I have an upcoming webinar this coming week. To put a lot of time to really customize a message around that and build that out for one outreach, the values not work to squeeze. But if I can move like this, now I'm able to do that. And that's the big difference here. The value and personalization is in the perishables. That's where you break through. And if you can't leverage that, which you can't today, you're going to miss out on the opportunity to really capture those people's attention and drive outcomes.

Roman Trebon: 24:39

Yep. The speed and the personalization, Sam, you know, that gets me excited. As a client, not even someone who's helping serve clients, but as a client myself, like you said, what, what, you know, what are we doing? Where's their opportunity to meet for me as a client? How do I compare it against others? And it, you know, you said that tailored towards me as my organization, that's real value, right? That's real value. And again, right now it's, it's slow. It's, you know, it takes time. It's, or, or you get something and it's you see a benchmark and it's, You know, i'm one of a segment. It's not me as an individual client. It's like, oh You're in this segment and here's how you can compare not me as an individual client. That's a real game changer. I love that It's gonna be so happy. We're either we ready to hit sam up with the the hard hitting questions here

Kevin Metzger: 25:26

Yeah, I think it's time. Let's go. All right, sam We got some rapid fire stuff for you. So i'm rubbing you want to take it or you want me to?

Roman Trebon: 25:34

Well, we can go back and forth here. Let's see. All right, sam All right, buckle up early bird or night owl night owl night out. All right movie Oh, I love it. You like, what about the sequels? You like, you're a sequel guy too, or just the original?

Sam Cummings: 25:50

Oh, yeah. The original is great. The sequels are marinade on you. You know, you got to sit into them.

Roman Trebon: 25:55

All right. You got a favorite sport that you watch or play? Boxing. Boxing. Who's your you got a boxer? You that's me. That's, that's my, you got, you're speaking my language. You got, you

Sam Cummings: 26:05

got a favorite boxer? Yeah. Yeah, so my favorite boxer now in this current time, Javante Davis, y'all see that fight this weekend?

Roman Trebon: 26:14

Let's go. Unbelievable walkout. The guy always delivers. So yeah, we will. I watched. So unbelievable.

Kevin Metzger: 26:19

Oh yeah. All right. So Roman's in the process of becoming a professional ref, I'm a

Roman Trebon: 26:26

judge, judge, professional judge here in Georgia. So I, I'm, I'm, yeah. So I, anyway, that's a whole different conversation. So.

Sam Cummings: 26:34

Get in the game. We need you.

Roman Trebon: 26:35

That's why I got into game. Sam, you've seen it. You're a box of being, you've seen enough bad scorecards. You know, there's help needed so much politics. Can't wait to see you there. Maybe I'm ever going to trip over to Saudi Arabia too. Who knows? All right. Book recommendation, favorite book or book recommendation for our audience. Smart brevity, smart brevity.

Sam Cummings: 26:57

All right. Priceless book. I think the biggest thing you can look up the author and all, but why I recommend it is we are in a time when you can communicate pointed and it's like being on CNN when they have eight boxes, you got to hit.

Roman Trebon: 27:10

A place you'd want to travel to, Sam, you've never been Bali, Bali. Nice. Nice. All right. And then to close out here, where can our audience find more about you and more about a data plan?

Sam Cummings: 27:24

Check us out on LinkedIn, Samuel J 3 1 4. That's my personal LinkedIn and then Datalant. You can just look us up on LinkedIn under D a T a P L a N T. And then also run a podcast called AI for diversity. We have 500, 000 followers on LinkedIn. You can check us out there.

Roman Trebon: 27:42

Awesome. Sam, thanks so much for coming on the show. Love the polo shirt. We'll have to find, well, you're the first up for our our guest shirts that we're going to queue up here soon. So really, really appreciate it. And for our audience, make sure to check out Sam. His stuff's amazing. Definitely check out his website, the data plant great stuff on there. And for everyone, thanks for listening. You can follow us on LinkedIn at Roman Trebon at Kevin Metzger, make sure to check out our customer success playbook page as well on LinkedIn. And Kevin, awesome.