Janis Zech: Welcome to another episode of the RevOps Lab. We are here with Olga. Nice to meet you.

Olga: It's great to be here. Thank you so much for having me.

Janis Zech: Yeah, it's going to be a fun episode. But before we dive in, could you give us some context? Who are you? What do you do?

Olga: I run revenue operations at a company called BirdEye, where reviews management, listings management, overall online reputation management company, software company, B2B. We're global, we operate in US and in UK and in Australia, and we have global teams spread across all those regions, plus India and then Philippines as well. So I manage revenue operations, covering the entire customer journey, starting with lead inception and having my marketing ops team focused on that, managing sales process with sales operations and post sales with our customer success operations. We'll also touch on enablement, analytics, go to market, tools management and all that. So we're busy.

Janis Zech: Yeah. I mean, the company has around twelve hundred to thirteen hundred people, right? So quite high complexity across multiple segments: SME, mid market, enterprise. And you have a fairly large team, right? It's around forty people or so?

Olga: That is correct. So we're pretty large in size when it comes to the team. And mostly because revenue operations at BirdEye are responsible for account sourcing. So essentially half of my team is responsible for generating new accounts for our upmarket sellers. So we use tools such as ZoomInfo, or Lusha, or Cognism, or Seamless, or Invan is one of the latest ones, or we just use Google, another great tool in our hands, to source new accounts and then create them on Salesforce. So we are responsible for generating new accounts and also keeping existing accounts clean to ensure that the data is intact. Every now, every ever so often, probably once a quarter, go through all of our existing upmarket accounts, meaning in enterprise and commercial segments, and verify them, and verify that the data points are still correct. So that is somewhat a manual exercise. So half of the team is essentially busy with that important work. And then I do have marketing operations team, sales ops and customer success, so sort of more classic RevOps part of the team.

Janis Zech: Yeah, yeah. I mean, so I think anyone who is a bit familiar with your industry knows there's probably a lot of accounts, so obviously that keeps you busy. Love the true RevOps, end to end bow tie, including everything, plus the systems and tools analytics enablement that you mentioned. But that's actually not the topic for today. I think it just sets the context of who you are and what you do. What we wanted to do is actually dive into AI because it's obviously a super hype topic, right? Obviously, Philipp and I, we're building a company that builds AI into various different workflows. You have a great perspective. We chatted before this call about it. Maybe let's kick off with one question, and that is: How has your view on AI changed over the last six to twelve months?

Olga: It definitely has changed. And I'm trying just to pace with all of the developments in the AI world, right, and be up to speed with everything what's happening. And it's just becoming so quickly, you know, changing the landscape, the use cases are new and new use cases are added pretty much every week. LinkedIn is exploding with, you know, sales and RevOps leaders, you know, using AI for different kinds of plays. And I met with some of my peers, RevOps leaders, in October, and we talked about how, you know, AI is here to stay. If in twenty twenty three and twenty twenty four, we were sort of like gauging this and not really sure if AI is a new shiny thing and it's just going to go away as, you know, NFTs, or is it a thing that's going to stay in here? Now it's obvious that it's here to stay. The question is, how do we, what are the use cases that we start focusing on? How do we leverage this from the get go? What are the major pain points that we can start solving with it? Being cautious and not compromising, say, on quality, but still helping with the process, helping with the everyday tasks that our sellers go through, our BDRs go through. So I used to say that I'm cautiously optimistic. Can't afford to be cautiously optimistic anymore. I need to get in there, right? I need to get on that train of AI and start implementing it and start using it in our everyday life.

Janis Zech: Yeah. Yeah. I mean, this resonates with me personally. I'm, like, obviously, I've been following it closely, since the first news of ChatGPT and so on became, like, more broadly public, but then I had a really hard time to actually really start using it, didn't really have so much trust in it, all these hallucinations. I, you know, I experienced them as well, not like I myself had hallucinations, but definitely saw the other—

Olga: That too, I mean, could happen as you're using it.

Janis Zech: It could be. Particularly over the, I would say last two months, I've been really just like diving a lot deeper into it, trying to make it really like a part of my daily routine. I myself, I'm like a Claude user, it's just like this just works best with me. Started to set up projects, incorporate workflows, just starting experimenting with it. Because like you said, yeah this is not NFTs and I think NFTs is such a good example because I think we all came from the whole crypto NFT like, you know, this was sort of like the big hype before. So then came AI and then obviously a lot of people were just very pessimistic about it in the beginning. But yeah, you see this changing now. I'm curious about like now that you are, like, you know, really believing in it and understand, okay, you need to get on the bandwagon, like, what sort of, like, outcomes would you want to drive with AI? Like, where do you have a high belief in it that it actually can create positive net results?

Olga: Yeah, and I think it's important to, first of all, understand the company goals and the pains, right? Put yourself — there can be multiple work streams there. RevOps is one. A day in life of a salesperson is another one. Analytics could be another one, an executive. So there can be multiple work streams. So what sort of drives the North Star and the revenue, right, the bottom line? For us right now, for me right now, that's the daily life of a salesperson. My goal right now is to maximize the sales time. As I look into sales process, our sales process, and actually hear and read through the same experience from other companies, right? There was some study that said that almost sixty percent of sellers' time is spent on admin work, filling data in Salesforce, just going through some steps and motions within your systems, right? Which is insane. We pay them to sell, not to do the admin work, not to do the data entry work. So my number one goal is to maximize the sales time and get them out of the everyday filling in the data in the CRM. So I will be looking into the AI tools, apps, or apps that will help us to speed this up and maybe fill in those fields and fill in the gaps, listen to the calls, aggregate some summaries, aggregate some findings on the account level or on the opportunity or on the lead level, and do the work for the sellers. The question is, how far do I want to go with that? Is that, you know, an AI agent that sits on every call and then sort of fills in the fields and takes through the steps? Or I want to be, again, like cautious and make it a more iterative approach and not compromise on the quality of the data in Salesforce. So we're going that direction. How we will approach that is still the question. So back to your question, maximizing sales time. So mapping to my pain, mapping to my number one goal within the company, mapping to the pain of increasing sales time and getting them out of the admin work and then plugging the AI capability there.

Janis Zech: Yeah, really love what you said about starting with the problem first. I mean, we are like both Philipp and I, we're like very much builders, so we build product. I think sometimes you feel like, okay, you try to create a use case with AI versus — and this is, I think, our core belief — like AI won't replace AEs fundamentally. If you look at the SDR companies, they are churning like crazy, and it's just not working right now. It's not better than the new outbound stack that's kind of signal based, I would say. I think it's our core belief that AI will create 10x AEs, but still there will be AEs. They will be able to spend, of the sixty percent, will probably automate a lot of the time, and there's obviously very obvious use cases you outlined, like how do you follow-up on meetings, how do you prepare meetings, how do you extract information and do CRM data hygiene. Many interesting use cases. But I think there's this debate in the community, right? Is it fully autonomous or is it basically faster and easier and automated workflows that leverage AI, but still the AE is in control. I think the best example I can think of, and I'll stop talking, is as a buyer, there's a lot of information already happening outside of the conversation with an AE. But if I then join a Zoom call, do I actually want to talk to an AI agent? And I don't believe that. I cannot see it for deals that are ten to a few million, that that's actually the experience you want. You want somebody who guides you through the process, but that person should be able to spend eighty percent, ninety percent on selling, and all the rest, all the annoying stuff, is basically taken care of. And whether you then call that agents or you call it AI workflows, that's a different discussion, I guess. But I'm curious what you think about that. Is that the belief you have as well, or do you see it completely different? Just curious.

Olga: I mean, let's go back to human nature. People buy from people. We need community around ourselves, right? So at the end of the day, not everything is a data equation. Not everything can be solved via some math problem. There are a lot of intangible things around us, right? Relationship building, trust building, thought leadership. Sounds like a cliche, but still, right? How do you sell right now? Now we sell through becoming a trusted advisor, right? That's why you guys are recording this content, there are so many companies investing in content, because you want to be the trusted partner, the trusted advisor, first of all. I don't believe an AI agent can become a trusted advisor, can actually build relationship with the person, right, and sort of gain their trust and understand their pains. I truly believe that, you know, as humans, we need communities around us, we need people surrounding us, and we buy from people. So I'm not, again, I might be wrong, but I'm not believing, especially in the upmarket world. Small business world, when you think about your average deal size of a few thousand dollars, this is where it is. It is a huge opportunity to drive efficiency for any organization, right? Do you even need five people closing your five thousand dollar deal, right? If you're also focused on small business deals, that's how you drive efficiency. Bring in AI, make it self serve, make it, as I say, a Netflix experience, have those customers just come in and onboard themselves, pay themselves, and then use the platform themselves. And this is where you plug in AI to help them navigate through your product. But as you go upmarket, and we'll talk about six figure deals, there's no way you can, like, escape building relationships, going on sites. During COVID, we're sort of like we're all stuck in our homes, right? And everything was on Zoom. But I think this world of, like, having field sales traveling and visiting their customers or visiting their target accounts is coming back because we do need that relationship building. We do need dinners. We do need sports games, right? We do need to build the trust between the parties. So, as in many cases, I think it depends. Small business, I would love to explore this being sort of fully automated with AI agents all over. But in an upmarket, I don't see how that's gonna replace account executives. When I'm signing a six figure contract, I want someone, like, real life sitting next to you.

Janis Zech: You know? I mean, I think we're in good company here, right? Like so Salesforce themselves said they're not going to hire any new engineers. Basically, I think taking the step to say, hey, now we got all the tools. Now, you know, let's stop adding more people. Let's stop throwing more people at the problem, and let's see how far we can get with AI, to sort of, like, you know, improve the efficiency of our existing workforce, and then let's see how far we get with that, and then iteratively take it step by step. I think this is also how you were thinking about it, if I understood correctly, sort of like, okay. Hey. Let's just take the possibilities that we have now. Let's see how far we get step by step, and then let's not necessarily create worse quality, but let's keep that level of quality at the same level, and then just be more efficient and effective.

Olga: Be more thoughtful as you roll this out. I also don't want to end up in a tools fatigue, as I think we all, you know, ended a couple years ago, several years ago, when pretty much everyone was in this B2B software game, and we had a tool for pretty much everything. From there, we went to more like a platform approach and centralizing a lot of systems. Now there's just so many apps, right? And I don't want to end up in having a user based app for every different step of the sales process, right? I want to come from, like, what is the biggest pain? And can I solve this pain with AI and make sales lives easier and more efficient, you know? And then from the RevOps perspective, I'm always looking for data quality and actionable insights and prescriptive actions. And if it can check boxes from either as well. For example, if this AI can sort of analyze the data that I can then use for coaching, if it can come up with some coaching exercises, that already solves my problem. The more boxes we can get to, to cover the use case for sales or for the operations, for executives, for analytics, you know, sort of the better chances this AI app has to end up here.

Janis Zech: Yeah. Let's take a look at some specific use cases, I think, just to make the whole part more tangible. You mentioned, right, like a big part of the team is busy with account sourcing, account creation, and that's obviously super important. Data quality needs to be more or less on point there, and actually to really drive growth. So is that like a use case that you're thinking about with AI to improve that part?

Olga: And we do actually have — we actually onboarded one of the apps to help us get there. It is also an AI driven app that essentially crawls the web and allows you to generate accounts based on various criteria, filters, etcetera. It plugs into your CRM, it plugs into other systems to sort of suppress and generate net new accounts. So we're doing that already. The next step that we're looking for is the meeting quality. Pipeline is the king, right? So we're looking to have enough pipeline coverage. Pipeline comes from — just based on our funnel, each company's funnel is different — comes from meetings, right? And as we go into the meeting, we want to ensure that the meeting is a high quality meeting. So there are different use cases. From the meeting, I want the details to be entered in Salesforce, and that's how I can save time from our BDR org or sales org by having AI enter details and populate fields based on the call outcomes. What I can also do is to use it for coaching to ensure the high quality, right, of the meetings, how we're running discovery, how we're asking certain questions, how we're addressing the pain, are we able to identify the pain, first of all. So that's another angle that I'm gonna be using it for. And so we're exploring, but we are very ready to sort of make next steps in that direction there.

Janis Zech: Yep. Yep. Yeah. That makes a lot of sense to me. And I think this is probably something where, you know, an LLM can get very far already. I mean, this feels very tangible to me, less experimental, where it's just about, like, really tweaking the underlying prompts really, really well so it fits to your use case and to have a tool that is sort of, like, as flexible, to actually really embed it into your workflow and really make sense. So—

Olga: And as we're talking now, I just had a genius idea, you guys. Can I run my genius idea by—

Janis Zech: We love spontaneous genius ideas!

Olga: Yeah. Can I run this idea? So say I'm onboarding an app that will listen to all of the calls and the number one exercise it's gonna do, which is gonna fill in the fields, right? Then taking the next step further, can I use this app to be able to predict the outcomes of this meeting? Does it make sense for this meeting to be even routed to my account executives, or should this meeting still stay with the early stages? BDR org and requires further qualification or further discovery, right? Because again, to maximize sales time, I don't want to send the meetings that are not yet qualified, right? So can I have this app sort of predict and forecast the close rate or the pipeline yield rate from this meeting? And maybe this app can then go and analyze similar calls from the past that either went closed lost or dead or went to successful opportunities that were closed. So it's not just the data entry, right? It's this full scope of analytics that we can go and crawl into the past data and then create actionable insights. We went from like gather data, store data, ensure data quality is correct, to then create insights and then actionable insights and then prescriptive actions based on that. So as AI analyzes all of those steps, can it then recommend if this meeting should be forwarded to an account executive or solution engineer or whoever in the next phase? Or does it require more qualification based on the past signals that it sees in Salesforce?

Janis Zech: I think it's such a good point because I think it goes into the routing challenge. Is it genius enough? I think we don't have that many genius, spontaneous ideas on this podcast, to be fair. But I think what I find so fascinating here is that, obviously, you have the routing use case, right? Who should take the call? Usually there's a data quality problem. You had all information about the account and about the stakeholders that are in the meeting and the signals, and you can take all the past data, but then also you enrich the context data for those meetings, you could route them a lot more intelligently. Then I think what happens a lot, and I think everyone who has been in sales knows this, you basically go into a meeting and often the top performers do a great job of preparing the meeting, but many don't. What you want to do is you want to obviously go through the LinkedIn. You want to look at the account data. You want to understand. You want to have an account prep. All this is automated. That's basically part of the meeting prep. You run the meeting. Obviously, as an AE, you often have a solution engineer, so there could be an AI that helps you answer during the meeting technical questions based on context data from your help desk, from your product usage data, and so on and so on. And then you go out of the meeting, and then I think what we are facing right now is like, yeah, we have all these meeting recordings. Nobody looks at them. They're all siloed somewhere. They're not well integrated into Salesforce or your CRM, and they're actually not using all meeting data or email data to then level that up to the opportunity there. I, as a manager, want to quickly understand whether there's risk, and then I want to either inspect and have a conversation about it, or I want to take action. We're building this, so that's why I'm speaking so passionately about it, but that's basically something that is going to happen. If you think about how the reality looks like most of the cases, it's just not that way.

Olga: And I'm passionate about that because what I'm experiencing is we go and listen to calls when there is a problem.

Janis Zech: Exactly.

Olga: And we're already so deep in that problem, that listening to calls will not help us, right? And you need to spend time and resources crawling through all of those recordings or transcripts to actually understand what's going wrong there. For example, if we're talking about setting meetings and listening to calls, right? What I'm looking for is to be more proactive, identify early signals, right, that we can address, and we actually have the time to address and fix and pivot. And that's where I think, you know, this AI can help us. You know, not just the actionable insights, but also like prescriptive steps of this is what you need to drive your attention to or your sales manager's attention to, right? Again, from like gather data, ensure data quality is there, ensure like you have one source of truth, then insights, actionable insights, and then so what? What's next? By the time, like, the world we lived in — or still I am still in there — like, when I'm being alerted of the problem when the problem is significant already, that it's everyone's aware of this. So it means that it has been going on for a few months, right, and to uncover that and untangle that, that's gonna take another few months to solve for.

Janis Zech: Yeah, I think there's the efficiency gains, but then there's, okay, how do you actually drive win rate? How do you reduce sales cycle lengths? How do you focus everybody on the right things? Because if you hunt a deal that is actually dead already before you're hunting it, then that's also a lot of time wasted. So I think taking the unstructured call data, taking the activity data, taking the CRM data, bringing that all together and then having essentially ways to signal, hey, look, here your pain discovery was really poor and there's actually no pain. You're single threaded and there's actually not multiple. So you go through the standard signals for opportunity health and you surface them to the right people in the right moment, and that's different to the AEs versus the managers versus the VPs. But I think this is right, that's—

Olga: And something more — like, if pain is something that can be maybe quantified, or ROI can be quantified, you can assign some value to it. Things that I cannot assign value — going back to us being humans and building relationships — I wanna see the sentiment. Is the customer excited? Or are we twisting their arm to agree onto this meeting, right, into the next steps, right? Or is it just like a call fatigue and we reached out to them fifty times this week and they just said yes, right? I want to understand that as well because that's gonna then impact my pipeline yield, my win rate, etcetera. Were they excited from the get go? So if I can sort of work that into the opportunity scoring or some sort of forecasting, you know, that is helpful. So understanding not just some data points, but also the sentiment around the activity. Now we're moving into AI scanning our faces and looking for the right wrinkle when you're smiling or not. Can you do this, guys?

Janis Zech: That's certainly something we cannot do today. Yeah. As a European based company, that would also be fairly complicated to do. Yeah. This is actually a super interesting topic because I think from the text alone, just talking about limitations, right, I think just from the text alone very hard to do, because like if the AI reads in a transcript that the potential buyer says this sounds great, I'm so excited, looking forward to the next step, right? But then the buyer actually, you know, like next step will be like ghosting the seller massively because they just said it to have a good ending to the meeting. Wouldn't be able to read that from the transcript so easily, right? So and then you need to build like a new layer of sophistication and I think that part then would be really really hard. So yeah, I'm wondering how far we can get with this whole like text part. Just working with like text, with transcripts, with just like input into fields and then sort of like when we reach sort of like the limit there. Then you just said like next steps, right? Another genius idea. Can then AI fill in these next steps? And if a customer or prospect says, okay, excited about next steps, let's catch up sometime next week. Can I have a task created automatically to catch up next week in the next step field updated, to connect with them on a certain date?

Olga: So then we can — that is possible, right? Like, I think that's already something we're running today. And that is very possible. Then the question always comes down to how accurate is it? Do you want a human in the loop? Do you want to automate that or not?

Janis Zech: I, for example, am not a big proponent of letting the AI update your close date because I think that goes into your forecast. You want accountability around that. I mean, you want clear accountability. But do you want the AI to give a suggestion and maybe say, hey, you put the close date here, and actually, the way I see it is like this. So I think there will be also, in that sense, an AE coach that is not a manager, but it's actually the AI that helps you to chat about specific opportunities and analyse specific data points. So I think that's very feasible as long as you have access to all the right data points. You need to be deeply integrated into email, calendar, meeting transcripts. And this is, I think, the multimodal, analyzing faces and everything. There's also compliance issues with that. I think there's many issues with that. Is that technically possible in six to twelve months? That's probably very possible, but I think we're not yet in that phase, in my mind. In my mind, we're in the phase of, look, we are in tech, so we are probably already running fairly sophisticated revenue engines. If we look at most of the setups, they are far beyond great. Think really starting with this, let's kill all the manual data entry as much as possible, reduce that time, improve the meeting prep, meeting follow-up quality across the board, and then actually get everybody to have better visibility into deals that are at risk and then level that up into the forecasting. I think if that would be the reality today, a lot of companies would have better win rates, better accuracy, better predictability, a lot more efficiency. I don't think — at least I talk to RevOps and sales every week — and I don't think that's the reality today. I think that's the year where this is starting to happen. And then I think in the enterprise, it will probably take a few years, actually, until this is fully deployed. I know LinkedIn sometimes reads like everybody's doing it that way, but that's at least not what I'm observing. I'm super curious what you think about that, because you know so many RevOps leaders that probably talk about this.

Olga: Yeah, and I think — we're talking now in January of twenty twenty five. I think if we meet in December of twenty twenty five, the conversation would be way way different, alright? Because the world and the landscape is changing so much, right? Most of the bigger players, CRMs, they roll out those AI features within the platforms themselves. You already have ChatGPT functionality coming into Salesforce, within Salesforce, in spring release of twenty twenty five, right? So there's account summaries and aggregation of, like, activities on account level that's coming into Salesforce. Then there's gonna be other major players on the market that will be introducing these AI functionalities. But to your point of how far do you wanna go for now, again, still wanna be cautious, right, on things like close dates or some other key identifiers on my deals. I want a human to look at it. I love the idea of AI coach. I think back in the day, there was something called WalkMe that would actually walk a salesperson through the sales process, but it was something that you would have to code, right? And it's something that you just put in there. If I would have an AI coach, you know, sitting sort of next to an AE and listening to their steps and the actions that they're taking within your CRM or other tools, and sort of suggesting some edits, that is gonna be major. So another thing that we can explore. Again, if we go forward twelve months, I'm pretty sure the picture is gonna be different. For now, what I'll be focusing on is transcribing calls into insights and prescriptive insights in Salesforce, and then thinking about how can I also leverage that from the ops perspective, and then leverage that from the analytics and executive perspective as well.

Janis Zech: Yeah. Yeah. I think this podcast episode definitely proves that you can quickly come up with a lot of good ideas — genius, right?

Olga: That's right. Geniously. Thirty five minutes.

Janis Zech: Olga, thank you so much. I think this was super helpful. Hopefully we, you know, sparked a couple of ideas to our listeners, with our listeners, and definitely worth exchanging, getting together, use the next RevOps meetup to talk about these topics, exchange on it. I think it's just helpful to create more transparency here and yeah, keep an eye on also what the big players are doing, what the small players are doing. I think there's lots of innovation happening right now. Olga, we always ask our guests one final question and that is what book would you recommend to our listeners? Doesn't have to be a RevOps book. It's totally fine if it isn't. But yeah, curious what you would suggest.

Olga: What comes to mind is, well, there's Simon Sinek, that's classic right now, right? I think everyone has read those books already, like Eat Last. I have a book right now. It's called Strategic. It's about how you elevate yourself and connect tactics to strategy. It's especially important for revenue operations because we tend to be very much in the weeds and the tactics a lot and the manual steps. It is very important to stay elevated and see the big picture and understand where the ship is going and to navigate it in the right direction. So, that is super helpful to just, like, stay up at the top, and also helps you to speak the executive level language, and have that representation at the top.

Janis Zech: Yeah, perfect. Yeah, I think it also fits very well to this episode. So thank you very much. We put it on our list that you can find at getweflow.com/resources. So keep an eye on that. Lots of useful resources to help you become a better RevOps leader. Olga, thank you so much. This was a great episode. Really enjoyed it.

Olga: Thank you so much for your time. Love talking to you guys.

Janis Zech: Yep. Same here. Thank you so much.