Latest / RevOps 500 / Why AI Is Giving Your Teams Different Numbers | Jose Forte
Transcript
- Hey everyone, Sajil Qureshi here with another episode of RevOps 500. Today I am super stoked to be talking to the RevOps manager from Infraspeak, Jose Forte. Welcome to RevOps 500, sir. Hey there, how is everything? It's Sajil. Yeah, good, good, good. So, what is your biggest mistake? Okay, so when I was thinking about this question, I thought any person that is invited by... or any ops role will have tons of funny stories. Otherwise they are not having the scope that they should. It's too small. We are seeing always impactful things. We are always playing with that. But I wanted to make something more recent and more related to the topic that I'm bringing today. So I thought on a current thing that is happening about the AI approach that I took. And I think when I say I took, I think most ops roles took that approach of looking to AI on first on a vertical perspective. How can AI help ops? And as we were doing this, all the other verticals were doing the same. So they were building their own stuff. their own artifacts, dashboards, reports, insights. And nowadays, I think I'm tackling this challenge a bit too late because all the other things are already done on the other verticals. So the first question, or at least a quick second question that I assume made was, how can I ensure that everyone will play with AI, but will we ensure compliance and accuracy between everyone? Why that? In the past, someone was looking, a non-observant was looking for a dashboard or a report. It was quite obvious that it was broken, the numbers were not super accurate, so they were quickly coming to us and asking if we could help. With AI, all the answers are super confident. I see nowadays tons of verticals talking about the same thing. both with the wrong numbers, just because they believe on what the AI is saying. So that was one of the mistakes. Nowadays, we are tackling by creating rules, plugins for everyone to look to the same things, for the AI to always count on the same architecture to a more global vision on what they are looking. We don't want to push them back. So we need to do this as they are already playing with it, doing tons of things, amazing things that sometimes are not super accurate. Okay. So AI issues, people using it the wrong way with the wrong sets of data, those sorts of things. How are you correcting that? What is your... How do you correct that? How do you find the problem? How do you correct it? So finding the problem, going to meetings. Going to meetings. Yeah, going to meetings and start hearing like managers being super, once again, confident about what they were saying. And since we are always playing with data, we were that numbers seem solved. And then looking for the first source on this solution. And okay, that solution doesn't have the whole scope of data. So he's not bringing all deals, all one customers, blah, blah, blah. Then the other part was we were making Digest. So it's something quite common on the places that I work for to create Digest to help managers to manage. So tons of insights and so on. And the arguments start to appear more usually. And once again, with the wrong data that sometimes was coming from the old data sources. But since we have complex architecture, data architecture, the AI was not able to understand how all the objects are correlated. And to fight that, I split in two important things. Security. So also something that at the beginning of AI, we're giving not to everyone, but big scopes to a lot of non-ops roles and with the right permissions and something that than talking about mistakes, something that in the past, probably to delete 100 deals, everyone will have access, but they have to click 100 times in some buttons. Nowadays, it's one problem. So that is the first thing to ensure that the scopes with more critical permissions are set. okay and we are 100 sure who gets that if they can get we discuss if they can get that or not and the second one by creating in our case we are working with cloth so we can make plugins and it's way easier because we can simply update the plugin explaining new rules new context every time that we have a new object and that will also update for everyone so we are sure that they have that installed and we are sure that they have the most updated version but things like gemini also have solutions like gems in those cases i don't like that much because we need to waste people to install and when you are on a more than 200 place company you are sure that there will be people that will not install not because they don't want but because they are missing that and eventually if someone has a a bad number that comes back again Yeah, exactly. So when you have, so you're obviously in this stuff all the time. You're an expert AI practitioner. Do you feel other companies or other businesses understand the issues in their AI adaptation? I don't think so. Aside from Infraspeak, I do consultancy in less mature companies and it's normal, less mature projects. We only think on structure and scalability and so on when things are running, when they are already based on something really clear and so on. So it was quite common. And by the way, also at Infraspeak, I started when we were... uh okay so uh through that through wow and at the time everyone had access to to everything because the biggest focus sorry because you were small the focus is on growing and the focus was growing velocity everyone was doing tons of roles uh so we couldn't make these things because security compliance and so on At first, mostly for ICs, it seems like a blocker of velocity, what we should explain. Of course, trying to make everything the most or the least blocking possible to everyone. That is not our objective. But they need to understand that if someone can make a huge thing on critical things without our control, eventually... everything could be broke and then if everything is is broke uh they they will have way more impact because the system could be down for for some days and so on and we saw things near that and is that why nowadays it's a big topic for us okay so what would you what would your advice be to to like a business that is that you feel like how can they know that their data is is not right or not clean or not accurate, how could they know? They just might be even just the blind talking to the blind, right? How can they know that their data is not right? One of the things, and that could also be a huge problem, you need to have a source of truth, not split through several spreadsheets, not split through several solutions. And even though you work the sales team usually or the marketing team works with upspot and in some cases the sales team wants to work with Salesforce, then the success team with other solutions like Planet, and that's okay. But at the end of the day, everything needs to converge on the same place and have someone, normally more than one person making that housekeeping, ensuring that there are mistakes. And if there are mistakes that, as I said to my team, that are... especially it happens one time or something like that. We simply solve that manually. If we see something being recurrent, we need to automate, to create new structures, to create new flows, to ensure that that works. Regarding the use of AI, since everyone can do tons of things by themselves, it's working less on authority. more on adoption with the teams. So we went two ways. I could simply say, no one can look to AI and believe that they were not looking to AI. And then saying to them, your source of data is a Power BI solution. You could only look to that. But for sure, I'll catch a manager on a weekly call with their team looking to a dashboard made on cloud. So the idea, instead of doing that, is that. We know that you are using AI to do that stuff, and that's good. First thing, we create tons of enablement, how to make prompts and all that security and plugin stuff, then enablement, and then global artifacts. So it's like if you want to go individually on Cloud, on Gemini, you could. But then since we are giving you weapons to tackle exactly the same, but on a global perspective, the official thing should be this way. So you make your own things for insights, you use the global ones for indicators performance analysis. Why is that? Because otherwise, nowadays, you are talking with your IC or with your dashboard numbers, then your IC will have their own dashboard numbers, then you are talking about with the acquisition source marketing outbound and so on with their own dashboard sources. And At the end of the day, all of you, if you are not using something global and fought for everyone, your numbers will not match. If they don't match your meeting, that should be to request more performance, to get the solution ends up on checking the numbers. And when the meeting starts like that, the meeting is already ended because the goal of that is not there. And I saw that happening tons of time. like this is my number that is your number they're trying to understand on the call why it's different and the the big goal of that meeting was already gone so how do you get how do you solve those problems if you're in the meeting you're the rev ops guy someone says the number is one the other person's number is two and they're fighting about it no i'm right you're wrong no you're wrong i'm right no no you can't be wrong i'm right i just so so what is how did you solve that i mean how did you find out that there was a wrong issue how did you solve it okay so the first thing as i said is to to give provide everyone common numbers one source of truth where they can do their own reports they can do whatever they want but we are sure that it comes from the same base that is validated by us okay and and the second one and making a bridge to the topic that i'll bring today is to get one single unit of measurement So a big challenge on companies, and that is quite common, is that when they are planning, the main goal starts on ARR, usually, and then they make an inverse engineer that will eventually end up, will pass from euros, dollars, whatever, numbers to 1MQL, 10 leads, and that is the place where it starts clashing. The first place where the acquisition sources look to a number that then... directly translate on the goals of the next vertical okay okay okay so let's try and change gears a little bit now you want to say so what is something that jose before the age of ai would tell something jose in the current age of it like what would you tell yourself jose pre-ai what would you tell yourself now like the you know like the younger version of yourself what would you tell you what would you tell yourself what is some some advice so that the AI ensure that unit, common unit for the whole structure, for the whole framework and to tackle, to ensure that... before everyone getting super expert and making their own advanced things, that we were able to construct things for them. So as they were growing, they were showing us new things and we were able to keep updating the global things. Because nowadays, as I said, it really relies on the, let's say, political role of ops. convincing people to follow us and that is is nice not for all profiles more uh analytic profiles was really uh bigger issues with this so when you were i think i think years ago you worked at a in like a in startup boot camps right where you were like a a marketing analyst if i'm not mistaken is that right yeah yeah yeah so yeah obviously that's before you know AI and all this stuff came into the mass world, right? What would you think that life would be like now? If someone was like a junior marketing analyst and a startup boot camp now, what kind of work would they be doing and how would that look? So from one side... i think way easier way easier because nowadays it's easier to build appealing stuff okay so okay you can easily create appealing dashboards part inside the only thing is that if since he's a junior and i see that a lot with junior profile is that it's not mature enough to validate what ai is doing way easier to create stuff It's way harder to set yourself as an expert because you really need to create your own stuff to start understanding what is failing, where are the holes, where are the high performance performing trends. Because if you have someone doing that part of the ops role for you, if you are always using AI, take this data and tell me what we could do. It will be way harder for you to start to know how to look for patterns. And it was something that I really enjoyed on starting from the marketing side. It was quite easier because it was the first moment of entry on the company. So prospect account was entering, a lead was entering, and then I was able to understand what was happening after. And that's why nowadays I make the full scope from prospect account to finance, everything in there. But the base of understanding. good performing patterns, opportunities of growth. Instead of growing revenue by new orchestrations or new revenue strings, also being able to grow revenue through efficiency. And that is more the financial part. And that was possible by starting from the base of a pipeline, let's say. The funnel is still a thing. And like the younger Jose. How would he have approached AI? Would he have been really interested in it? Would he have been afraid of it because it might make things too easy? How the younger Jose have reacted to seeing AI for the first time? As a bulldozer, as I call him. As a bulldozer? I always liked that approach, even though nowadays... totally understand that every time that, and if this is for a junior ops role, you need to understand that it takes time for an organization to accept you because it will be a new ecosystem outside from the usual thing when you are entering the marketing vertical, when you are entering the sales vertical, that will start giving insights, telling people. if they are performing well or not, because sometimes they don't know. Of course, the key indicators, if you have a manager or a team that doesn't know if they are achieving the key indicator or not, probably something is wrong on their side. But everything in between. you will start looking for and telling people don't do like this, you should go that way. And if they don't trust you as an expert on the matter, if you don't have some quick wins to gain their trust, that will be a challenge. But once again, if you are able to do advanced and huge things, and sometimes more than starting with super accurate and super complex structures, you just need to enter the big table. okay as i say most of the time when i started that way i was on the leadership table before having a leadership role because i was there not and that once again it could be a a method of dealing with ops but in my case i i always prefer that since we don't have authority on the teams i'll not be the one saying the strategy for sales but i'm the one convincing sales to go that way and go that way so is that why it's so important to to be on the table and to have a voice so in our case we always prefer to have the ops roles on the revops team it's part of a big operations team than to be a role inside of the vertical they work way more with the vertical than with the ops organization but their weekly things their career reviews come from me to ensure that there are alignment They are on, each one of them are on their own vertical. They align with them, give insights and so on and so on. But at the end of the day, we all align to ensure that the whole company, the whole framework is looking for the same. Interesting. And staying in the same, I think, interesting, now you've been there about five years now, right? How do you, how do you... I met you for a speak for the past eight. Eight years. I lost you. I lost you. It's been that long. So eight years. How do you stay in a RevOps role for like eight years? I mean, how is that? What is the secret to success in that? One, the company. And it's also important to give kudos to, I'm included too, since I'm in here, but to the company overall, since we are on playing the investment. the investment game time to time. Nowadays that we are getting bigger, it takes a bit more time because it needs time to happen. But we are always, every investment round, we are kind of a new company. New challenge, way more new people, way more new sources of growth engines. So it was when our goals were to reach 1 million ERR, probably simply doing ads was enough to get that. Nowadays, we have several segmented teams for sales with super specific things, tons of different acquisition sources, tons of ways of clients to pay, models to have, then different verticals on success. And all of that nowadays combined on one team. They are doing an amazing job on the strategy and so on. But at the end of the day, we are the ones looking to the whole framework and ensuring that they are aligned. They are together. They are not missed parts. And when people are on podcasts and things like this, sometimes it looks that everyone has everything together. And no, what I'm saying is that we are always doing this path. There is always something new to add that up. There is always a few things that if they are super new, like a new vertical will enter. It's not the time to act. So if we act, we'll for sure go with the mindset of making that on a scalable superstructure and so on. And if it's something super on igniting mold. So if your company is starting the product-led growth motion, ops should only appear. It could give some insights at the beginning and so on, but really approach that as a known project and so on when things start to get in shape. Okay, so when it's more mature, because basically what we are saying is that we are implementing you on the main framework. And then there are rules, there are criterias, there are things to follow. Got it. And Jose, what is something that worries you today? What is something that keeps you up at night technically or in the job or something? What's something that's really concerning for you that you're trying to do but you're not being able to get done yet? As I said, a bit as I answered when you asked me how I am and I said, always good. Yeah, exactly. When you asked me what worries me. So, of course, we have tons of challenge and it's hard sometimes to play this role. My team, not talking about only myself, but on one side solving tickets, on the other side planning. big things on the other side, implementing huge implementations, new verticals that we need to do that because we have different faces. Sometimes we are project owners. Sometimes we are system architects. Sometimes we are simply a support ticketing solution. And that could be super worried if we're not seeing a path. So one of the things that we always tackle, the first one, and that will be really, really annoying. Having tons of tickets as the teams start growing, everyone needs to port, everyone... one's answers mainly on the sales side because they are dealing with the client super urgently. So they start calling you and so on. Just give you an example. The moment that I start planning to hire two roles, two senior roles for my team to convince that on the budget, I simply made one thing. I went to ClickUp at the time we were using that. Every time that I was getting a ticket, someone calling me, someone sending me a message, I was clicking on on a counter to know how much time i was spending per vertical and how much times i was solving in a week that's 140 something tickets not all of them were huge but it was a lot and then imagine that i was doing that but at the same time on meetings talking about new projects to know how to implement that and of course things were failing and i was getting near not near to burnout but near to be near we're not so for seeing hired senior profiles and that is quite important to ensure that they they will be able to sit on each vertical leadership table and the second one to look for that why everyone is asking for support so much like of enablement direct direct answers like of a global channel if everyone is sending you a direct message they are not seeing the questions from others that we already replied So we started solving that. Nowadays with AI, for example, we can also tackle that. I always tell them, if this is becoming a routine, we need to optimize by automating, by something. But if every day you need to do something, you need to decrease the amount of time that you are doing. And this was the first case. Nowadays, we still have to support in some tickets, but mainly complex things. and with complex things it takes way more time to solve but is not as urgent as the small ones but because people understand i'm asking something super complex so i'll give them time it's not a big amount of things so you can simply focus on that one and things are getting in shape so that is one of the things that i could be really worried about but i'm solving the other ones on adding everyone to the same framework and so on. I was a bit with this new transaction to AI because I was really seeing people doing advanced things and I was worried on how can I convince them to come to my global things at the beginning since I'm doing this quite quickly and for everyone will not be as pretty as there. There will be some things missing. But once again... After the enablement, the teams understood like, okay, those guys first are trying to explain us how to do that. And then, even though the first prototype is not as advanced as mine, if I keep giving them feedback, if I keep working with them, at the end of the day, I'll not be the one doing that. I'll not be the one or I'll not have walls against data access because those guys can do. can access to everyone that I get. So that was also solved. So once again, there are tons of things on a daily basis that can worry myself. The way to tackle that and to be more calm is first to write everything down. It's quite important for anyone on the ops role. Otherwise, you'll have tons of small tasks to do. And the second one. to plan ahead even though it's not for tomorrow because the projects start to to pile themselves if you have something to solve people will be way calmer with you you'll be way calm and you can take the right decisions jose where can people learn more about you and talk to you more Okay. I'm quite active on LinkedIn, but also on HubSpot community. So I made some articles and time to time when I have time, a routine that I had Monday mornings, taking my coffee, was to reply to tickets on the HubSpot community. Basically because one thing that I had for working for Infraspeak for so many years, nowadays it's not super normal for someone to be working at the same company for eight years, was that I was afraid, and that would be something to worry about, that I was getting... to specialize at Infraspeak, not at RevOps, at RevOps at Infraspeak. So is that why I started to make consultancy to see other type of projects and also these small things. So by answering tickets, and I have more than 100 tickets to reply themselves on the AppSpot community, I was able to see other type of challenge, things that were not happening at Infraspeak were happening to the others. And of course, you need to focus on that because if you are on the ops role, that could be as a game. kind of vicious and then you confuse your work with replying to others. So you really need to set like, I only can play for like 30 minutes, 45 minutes every week. If there are more tickets, it will be for the next week. Otherwise it's like playing PlayStation, Game Boy, whatever. Game Boy? That's an old one. But that works quite well. And that's why even though Infrospeak is always turning into a new company, the base is always the same. So I wanted to ensure and it's been quite nice. Keeps me way motivated to keep in here because I see what is happening outside and I still like what is happening inside the way we are growing. So it's also an advice for profiles that are always jumping from one thing to the others. It's a strategy, career strategy and nothing against. But sometimes it also takes you from some opportunity to evolve. Once again, it's quite important that you have a voice on the leadership. And if you are always jumping from one side to the other, you don't get the time to have that voice and to really do things. Jose, it's been really interesting talking to you, getting to know more about what your background's like, how things are interspeak, and what you feel about AI. Thank you so much for coming out and jam with me today. Thanks, Agil. I have tons of topics. Just invite me for more. Yeah, absolutely. Absolutely. This has been another episode of RevOps 500, everybody. Thanks for listening.