200. What a Chief AI Officer Actually Does ft. Ziv Peled (AppsFlyer)

Our implementation is better. I’ll explain why. If you spoke with me two years ago, I would tell you, I don’t think dashboards are going away. I think dashboards are going away. When is the last dashboard?

Josh Schachter [Host]:
Say more. You’re listening to Unchurned, brought to you by the Gainsight podcast network. Ziv Peled has been in customer success for over a decade. He bought Salesforce for AppsFlyer in 2014. He built the dashboards. He ran the CS Org. Now his job is to tear down everything he built. As Chief AI Officer and still CCO Ziv is taking every workflow at Apps Flyer and moving it to AI.

Josh Schachter [Host]:
Not augmenting, replacing today what that actually looks like from the inside, why most AI ROI numbers are noise. And one question that’s been sitting with me since we recorded this. When is the last dashboard? I’m Josh Schachter. This is Unchurned. Subscribe to our substack@ Unchurned gainsight.com where we go deep on every episode. Like how one post sales team at cloudbeds built over 150 AI agents. That story and more at unchurned.gainsight.com everybody. Welcome to this episode of Unchurned.

Josh Schachter [Host]:
I’m your host, Josh Schacter, senior Vice President of strategy and go to market development at Gainsight. And I am excited to be here today with Ziv Pelad of Apps Flyer. Ziv, your title just changed. What’s your new title?

Ziv Peled:
Yeah, I’m now the Chief AI Officer. And thank you for having me.

Josh Schachter [Host]:
You are. So Ziv is now the CAIOow of AppsFlyer. The ciao. Oh, yeah, the ciao. That’s right. Is that how you pronounce it? Is that the Italian ciao?

Ziv Peled:
Yeah, yeah, exactly. But yeah, this is how my CMO calls me. Ciao.

Josh Schachter [Host]:
Ciao. Yeah, it’s pretty fitting for you. Okay, so Zeev, well, was, but really still is the Chief customer officer of AppsFlyer. So now you were explaining to me kind of in prep that you’re kind of in a chief operating officer role in some way. In some sense. Right? In your chief AI officer role. Because at the end of the day, like, what is a chief AI officer doing? They’re optimizing workflows and organizational transformation. Right.

Josh Schachter [Host]:
So that sounds pretty much like operations to me. Where do we want to start? You and I met like, six years ago. You’re one of the first people that I met in customer success. I didn’t know anything about customer success. I still don’t know anything about customer success. You know, that’s true. However, you were so impressive because you were very boastful in showing all these, like, amazing dashboards that your team was building in Looker. I know you showed, like, everybody your dashboards and Looker.

Josh Schachter [Host]:
You make us all feel special if it’s just us, but then you show everybody because you’re so proud of that. But, like, when I saw your stuff, I was like, wow, this guy, really. I don’t know what customer success is, but I know whatever it is, this guy gets it. You’ve always been ahead of the curve with your. Your beliefs and predictions and strategy and tactics and operations of it. So kudos to you. That’s why I hold you so dearly in my heart. And now with this transformation of your own, I can see that you’re still continuing to be ahead of the game with AI and bringing that type of culture to apps.

Josh Schachter [Host]:
Flyer, where do we want to start this conversation?

Ziv Peled:
Sure. So, first of all, thank you for the great word, but I’m still the Chief Customer Officer in addition to the Chief AI officer. And I think what’s really interesting is that the teams that were building the infrastructure and the Looker dashboards, for me, seven, six, five years ago, they are now the teams that are building AI, and they are now the teams that I’m leading in the new role. So I think that first of all, it’s definitely closing a circle, but it’s also, I think that this is an important connection. The CEO before, it was very well connected to go to market. And in the end, you know, I told it to our Chief Business officer, like, in the end, go to market is my biggest customer. It’s not my own, it’s not my sole customer, but it’s my biggest customer. And I think that the AI transformation is huge.

Ziv Peled:
It’s like I’m now focusing for the first time internal on that side, and there’s so much work to do here, and there’s so much, you know, the range of things that you can do in AI transformation is endless. So you need to pick what’s coming in the short term, what’s coming in the long term, what can we do very fast, what’s not, etc.

Josh Schachter [Host]:
Take me back to the moment where you were meeting with your executive team, you’re part of the executive team, but when you guys said we need this role, you need to be our chief AI officer.

Ziv Peled:
Yeah. So it’s an interesting transformation because when the COO decided to leave then we had a lot of different options. When is this? It just left just now. But that was a couple of months ago when we had those discussions. And there were a few options, like you can take this department and you can have numerous options of how to play it and places to put it under product, parts, under engineering, whatever. But in the end, in the important era of AI transformation, the consensus was that it has to be in one place, it has to be the same form and then it has to be AI native. So the organization wants to transform to become a native. And this is the.

Ziv Peled:
My core is to take this department and to build it the most AI native in the company. So it’s many things. I’m getting better.

Josh Schachter [Host]:
What department Success or no, no.

Ziv Peled:
The Chief AI Officer, biz ops, business operations. So I’m taking them first to become super AI native. And to the extent that I’m hands on, I’m there. I’m in the tasks, I’m in the strategy, I’m doing product, I’m doing some engineering. I don’t want to say I’m coding, but I am in the details of how we write, what we write and I’m trying to find solutions. And also it’s not always can we develop it faster? So it’s one aspect to look at it, but it’s also are we making the right architecture, are making the right choice? Because something that we are building today in the old world and you know it also from Salesforce development, et cetera, like if you do some decisions today, you might have them still maintained four years from now. So I’m now making the right decisions and I think it’s a great. This is why my journey is valuable here is that like I saw I bought Salesforce at AppSlider in November 2014.

Ziv Peled:
I know all the mistakes that we’ve done in the implementation and then I know how hard it was to integrate some of the very tough integrations and I know how easy it was to integrate some of the much easier integrations. And it’s the newer companies were the easier integrations and now, I mean the era that like we released our MCP over our data probably like four, five months ago, our MCP and like only now Salesforce is they released their mcq. I think we will always Have a better implementation. It’s not to say that.

Josh Schachter [Host]:
What do you say? You say you’re better than Salesforce?

Ziv Peled:
No, no, our implementation is better. I’ll explain why. By the way, I wanted to say that the end of the sentence was, it doesn’t mean that Build is better than Buy. Buy is in most cases better than Build. I’m saying that. But like, the. The advantage of Build is that you will always get something more customizable to you. So, like, I am getting.

Ziv Peled:
I started by getting the GONG recordings from the GONG API, but now getting the transcript and processing the transcript ourselves. We’re getting much better things. And it’s in the cost of. I think I told you that in the prep. Like 20. $20 a month.

Josh Schachter [Host]:
$20 a month for. You’re taking the transcripts from where? Directly from your Zoom calls?

Ziv Peled:
No, no, from Gong. But then we do the briefs, we do the TLDRs, we identify the signals, so we get more signals. And this is the important point again, like, it doesn’t mean that it’s always good to build, but I am getting the much more precise signals for Apps flyer. So it’s not for every customer. Like, every company will be able to build their custom better than the vendor gives them the generic one.

Josh Schachter [Host]:
Do you know UDI Lettergore? I’m sure you do.

Ziv Peled:
Yeah, of course, of course.

Josh Schachter [Host]:
Israeli cmo, former CMO of gong, now chief of. He had a post the other week. I thought it was great. I texted him, like, great post, udi. It was, you think you can build gong? No, you can’t. And then he went on, even Anthropic is our customer. Come on. I loved everything about that post, udi, like, shout out to you.

Josh Schachter [Host]:
Read Courageous Marketing. He wrote the book. It’s a great book. So what’s interesting, though, is that what you started saying, 20 bucks a month, I can take the transcripts. I can do all this amazing analysis and synthesis. My head immediately went to, oh, so, okay, so you turned on gong, like, 20 bucks a month. You could just take the transcripts. You just use it.

Josh Schachter [Host]:
Use granola or Zoom Zoom Recordings or whatever. But you said, no, no, no, no, no. I’m still working within gong. So this is a case where, like, you still have that. That. That big mega platform that is. That’s still within your tool stack. I’m not hearing you say that you’re churning on it, but you’re building on top of it.

Josh Schachter [Host]:
So tell me a little bit about that.

Ziv Peled:
Yeah, absolutely. I’m not churning from gong. I think GONG was very, very smart to push more value. Like, this is customer success. Like, if Gong were not there to build the deal board and the forecasting, like, not only that, I’m not telling you that we churned. We did an upsell with GONG because of the forecasting and we didn’t get, by the way, the full product. Like, we were waiting more for existing customers management and more account management and lots of other features. But like, I believe that Gong and I told them that, you know, from all the main vendors I work with that for the last two and a half years are saying they have AI.

Ziv Peled:
Maybe GONG is the only one I believe.

Josh Schachter [Host]:
Yeah. Oh, this brings me back, Ziv. I never even thought about this. Take conversations that you and I had years ago.

Ziv Peled:
Yeah.

Josh Schachter [Host]:
Where you’re like, so proud that you are like the, like, the most discounted GONG customer in like, the entire world. I know you’re gonna make me take that out of the recording, but I, but like, I, whenever I know you beam with pride about that because you’ve been with them like forever for like over a decade. So wait, so let’s stick with gong. So, okay, so now I’m playing a little bit of a straw man here because I think it’s really interesting. This is. We didn’t intend to talk about BONG at all, but this is, this is like, this is. This is the industry. Right.

Josh Schachter [Host]:
This is a perfect case study right now. So, okay, so. So forecasting kept you, has kept you and even expanded the opportunity for them with you guys. So just build forecasting and then you’re going to churn in four months.

Ziv Peled:
It’s a possibility. This is. I saw a few posts lately, I don’t remember by who, but like, probably three different posts that I saw about people speaking about, like, the modes are changing to big canals. And then like, I’m not sure I understand the concept of it fully, but it’s like, it’s. Instead of just making something secure, your uniqueness is like, it’s not about uniqueness. There could be a partnership. It can work with the LLM, it can work with different things. But like, you are.

Ziv Peled:
You’re making sure that you are in the canal and that you are flowing value. And I think that if, if, if GONG intended to or not, I think they are, by the way, continuously always pushing value out in AI, in this forecasting, in deal rooms, et cetera. It’s really important because I have other vendors that are speaking about some of these features, but they are not providing, not delivering in the end. And I Think it’s important. I think that it will be very interesting to see the next two years of LLM because, like, you know, we are, we are very shocked of what we saw in the last two years and the progression and the speed of the development of these models. But we don’t even believe or we don’t even think, well, like, what is Mythos? What is the next version after Mythos? So Mythos is the next version that anthropic ones, the next major version, anthropic ones to release on Claude. And they are, they fear that it’s so powerful that it will break a lot of it will be able to hack a lot of different systems, find a lot of loopholes and issues in security in many, many companies. So they released it for very low number of customers, partners to test it first.

Ziv Peled:
Most of them are in the security space in order to secure it, to, to make sure that it’s secure. But by the way, I think that, you know, if we talk one more thing about quantum computers, this is the next area that is really scary because, like, all the security that we’ve built for the last 40 years is worth nothing in quantum computers. So, like this, oh, we will need also to release quantum computers. Really, really measured and to whom and et cetera. But I think it’s. It’s great because it says, for me, it’s saying there’s a lot of so much more technology to come, so much more disruption, so much more innovation that we can do. And then it’s. It’s going to be really interesting.

Ziv Peled:
What I wanted to say, continuing my previous sentence, is that, like, I had a lot of videos, a lot of video recordings of screen sharing of internal people doing things, and I wanted to process. And then, like, I asked Claude, where should I process it? Like, it’s a lot of. It’s long videos. It’s a lot of gigabytes of videos. Sorry. And then Claude gave me a prompt and he said, you should go and do it in cloud, in Google AI Studio. And then I was shocked that, like, yeah, I was shocked that 90 minutes of video was processed in less than two minutes with the shocking output of exactly what happened in the video and what they identify to be. And I looked for agents.

Ziv Peled:
Do I look for. What should agents do?

Josh Schachter [Host]:
So, so, okay, so. So back to gong. So the thing there is that the canals. Canals. I kind of. I get what you’re saying. It’s like, I think I know GONG invested heavily in AI from, you know, for many years now.

Ziv Peled:
Right.

Josh Schachter [Host]:
And probably like, For a while it was just like, what are they up to?

Ziv Peled:
Right?

Josh Schachter [Host]:
But, but it seems like that’s starting to pay off with the, with what they were doing. So I think the lesson is you don’t like companies, just don’t know where it’s all going to net out. But just like, move quickly. If you’re a legacy type of organization, I use that term loosely, like, move, experiment, fail fast, but just come up with these speedboats, I think is what you’re saying, like, don’t just be the cargo ship. Deploy your speedboats and just make sure that you’re in the game. And I actually had coffee with Jen Zhu and I pray that I’m pronouncing his name correctly. He’s the founder and CEO of Zora earlier this week and he was talking about his AI transformation. You know, he’s in the process of transforming the whole company.

Josh Schachter [Host]:
And I was like, oh, you know, I said something along the lines of like, so you’re transforming all the different organization, the entire organization, all the different functions, etc. He’s like, yes, but the place that I’m most, that like, really matters most to me is engineering. Because if I can get engineering to be so efficient and productive and move so quickly, that is how I will stay ahead. And that, it just kind of reminds me what you’re saying with gong, like not diminishing your efforts as is AI transformation within the business operations. But if companies can stay so far ahead and be efficient with what they’re building, that will continue to be a moat for them. Let’s, let’s talk about the chief AI Officer role. How are you going about this? Like, you guys were born over a decade ago, so you’re not a native AI company, but you do have that, that mentality and that mindset and that spirit and that desire. So how do you transform into an AI native company?

Ziv Peled:
Yeah, it’s, it’s the question, definitely. And there’s two parts for my vision around that. So like for the short term, it’s about a lot about the mindset of people, but a lot of the connection between leadership, managers, employees, and how to transform that. It’s not only enablement, it’s like it’s a mindset shift and it has to come top down and bottom up. Still, in the short term, the biggest thing is to look at the main workflows and to change the main workflows to be AI. That’s, that’s. It sounds very simplistic, but like, it’s like, yeah, you sit, you you have

Josh Schachter [Host]:
an off site, you sit at a whiteboard, you say, what are the top 10 main workflows? And then how do we transform them?

Ziv Peled:
Sure it’s pretty, pretty easy, as you said, but then you need to make sure that you take the biggest workflows. So like onboarding of an employee, change it to AI, different areas of data input, change it to AI, a big one in sales. The opportunity, you create it in Salesforce. Create it now from cloud. When it works well, disable the opportunity. Disable the action of creating an opportunity in Salesforce from that moment on. You want to go tomorrow?

Josh Schachter [Host]:
Tomorrow, Series B company establishes a chief customer. Excuse me, a chief AI officer. Because they’ve listened to this podcast and they’re like, no shit, I’m going to go and I’m going to prioritize my top use cases and workflows within the company and AIFI them or identify them. What is like concretely, how do they do that again? Like do they call together a meeting of all the different functional leaders? Is it a memo sent out? Like what? Like give it as concrete, as pragmatic as possible.

Ziv Peled:
Guys, I still spoke about the short term, but yeah, definitely department by department, role by role, jobs to be done from A to Z. Watch. This is what I described before with the video. The video shows what are these specific functions doing for entire day and what can we automate with agents, but the agents, I live for the long term, but the main things that everyone are doing two, three, four, five times a day. And then not only it’s not just to force AI to be in the workflow, it’s because in most of the workflows I don’t think of one workflow that we enabled AI that didn’t add any value. Like the intelligence in creating opportunity is a lot of value. Adding that AI, adding that stack step include. And by the way, it’s many things and I see that like, you know, we launched so many processes in the last 13 years of me here in apps.

Ziv Peled:
I saw some processes that my employees liked and I saw some process that they didn’t. And like when there’s no value, they don’t like it, they don’t adopt it and they don’t do it. The great thing about this short term action is that it forces them to, to use AI. Forces them to use AI.

Josh Schachter [Host]:
Okay, we have, I have four questions in four minutes. Ready? Sure. One minute each. Adoption. What’s the trick to adoption? You just talked about it. What’s the trick to adoption?

Ziv Peled:
From FOMO to FUD you start with fomo, you bring a lot of value, the FOMO works and the rest is

Josh Schachter [Host]:
FUD that if they’re not in the game, then they’re going to be missing out, then they’re not going to have a job, all that sort of stuff.

Ziv Peled:
Fomo, like most of the people will go towards FOMO and then the ones that don’t like. Sorry, fud.

Josh Schachter [Host]:
Yeah, okay. Measurement.

Ziv Peled:
That’s the very tough point. First of all, like finding the right KPI. Like everyone are speaking about like 2x5x10x50x50x what? 50x what? Like we enable, we push now $10 million investment in AI and then like we get 50x of revenue. No, it’s. It doesn’t work like that directly. If, if the AI transformation will work as what you said before, and it will in the end also push product engineering, go to market and then you’ll get a lot of value out. And that value is also as a product market fit, then it can do 5x, 10x whatever. But besides that, it can be an increment in the.

Ziv Peled:
Some increment in the revenue, I put it aside. But a lot of more things in, much, much more efficient. So it can be. Time can be less friction in the end being measured by revenue and.

Josh Schachter [Host]:
Yeah, yeah, it’s tough. It’s a tough one. I want to know like the exact dashboard that you have that measures these things. I know you have one, but we’re going to move on.

Ziv Peled:
We should do another episode about when is the last dashboard. Say more when is the last dashboard? Because I don’t know. If you spoke with me two years ago, I would tell you I don’t think dashboards are going away. I think dashboards are going away.

Josh Schachter [Host]:
Yeah, I get it. I get where you’re coming from. Expense related to measurement, expense. How do you control expenses? Or do you.

Ziv Peled:
Yeah, so this is a big one. I think some companies understand it, but most companies doesn’t. We are going to have a huge increase in AI cost if no one sees it. But like all these companies are now changing their business models and they’re increasing the prices significantly. And worse, they are changing the business model so it will be consumption only. So you don’t even have the limits coming August. You will not have the limits that you have today in the models, in the plan, in the plan that you currently have. So first of all, you need to get ready for that.

Ziv Peled:
So get ready for August. Make sure that you have the system to know and to control people. But I think that you have to have token maximization. So not to use the max amount of tokens, but the best quality output and maximize tokens. You have to use a lot of tokens, but it has to, it has to be efficient. And this, this, this, I think it’s an equation that you can control. You can find the right balance for

Josh Schachter [Host]:
that and close measurement of that, I assume as well, keeping, keeping a pulse at all times. Security, how do you, how do you maintain security?

Ziv Peled:
From what I hear in the industry, there’s not enough solutions today to the risk. But then companies also needs to make great decisions about like versus what is my risk of using Cowork and then versus my risk of not using Cowork. So there’s has to be a lot of common sense in making decisions around security. So we are investing a lot in security. We’re putting everything that we can out there and on the other side in some specific areas, we make very conscious decisions.

Josh Schachter [Host]:
Ziv, we got to get you back on the program again some other time. We got to go into more details of your use cases and everything. I think you, I mean, you’re clearly a step ahead of the game here. And every time I speak to you, I learn a lot. And so I did this time as well. And I’m glad our listeners have that opportunity. Also, Zev Pelled, the Chief Customer Officer and Chief AI Officer at AppSnier, thank you so much for being on the program.

Ziv Peled:
Thank you, Josh. Thank you for having me.