Rachel Krall: The RevOps Mindset That’s Reshaping Strategic Finance

Business. Business. Business. Business.

Welcome to Go-To-Market, a podcast for RevOps, sales, and marketing leaders to dive deep into industry insights and best practices. Go-To-Market is sponsored by Fulcast, the go-to-market cloud. My name is Amy Cook, and I am so excited to introduce Rachel Krull, Senior Director of Strategic Finance at LinkedIn. Hey, Rachel, how are you?

Great. Thank you for having me.

Oh my gosh, thank you so much for being here. It is such a pleasure. You and I have had a couple of discussions before, and every time I chat with you, I learn something new, and it is so interesting and fun, and it's exciting to talk with someone who is really at the cutting edge of what is being done in RevOps.

You're very nice. No, I've also appreciated our conversations as well.

Yeah, it's been really fun. So for all of the listeners here, could you tell us just a little bit about your background and how you got started?

Yeah, happy to. So I've been at LinkedIn for about 10 years now. Prior to that, I dabbled in market research for a bit at Nielsen, working on global projects across a variety of different kind of industries and then went to business school and have been at LinkedIn ever since. And so I actually started my career at LinkedIn in go-to-market operations, sales ops, rev ops, bounced around for a few years across multiple different kind of domains within rev ops, and then actually just recently transitioned to strategic finance about 4 months ago. So I was in rev ops, in different capacities at LinkedIn for about 10 years.

That is awesome. So coming at it from a financial perspective, how has your perspective about RevOps changed?

Oh, it's a great question. I think one of the first things that I really noted when moving into finance was actually just the aperture widens so much as it relates to the things that you care about and you're looking at on a daily basis. In RevOps, I mean, I personally felt like I had a pretty good purview into our challenges and our opportunities as it related to our business. But then coming over the finance side, you realize that there's a lot of other things that you don't really think about, like the cost of our data centers or just, you know, like various parts of our overall business that I personally like hadn't really had exposure to in RevOps. And so I think that was one. Thing that was really unique or interesting making the shift over. I would say though, like in terms of the skills and the, you know, things that you're getting asked to think about or do and how to problem solve, it's very similar across the two disciplines. And maybe one thing I would say that I realized I appreciate about RevOps is you really do become very stakeholder-minded and in some ways you're always thinking about your customer, which is like the go-to-market team, and really ensuring that you're understanding what they need and how you can support them in their jobs and in their daily lives. And so one thing that I've found is like I've actually started hiring a few people from RevOps into strategic finance. And what I love that they bring is that real like problem-solving but customer-oriented mindset, really trying to understand like who their partners are, what their needs are, and how they can solve them for them. And that like partner orientation, I think, is not maybe as common in the finance org. And so that's like one thing I've really enjoyed observing as well as I look across the two teams.

Yeah, I think that that's really awesome. It does seem like finance is a step removed sometimes as you're looking at the overall— you've got a great bird's eye view of everything and how it's working. But it's the thing that I think is so fascinating about it is that RevOps is really in the details, right? There's not another— besides marketing, there's not another discipline that I know of where you go so deep into everything from, you know, did this pixel get placed appropriately so our ads are showing, all the way up through, you know, how is this being presented overall and how is our market positioning going? That's like on the marketing side. And from the RevOps side, to me it seems very, very similar. Like when I do RevOps, it's like, is, is this field between HubSpot and Salesforce integrating correctly all the way through? Some, some of the ways that you're connecting with finance in terms of, look, I'm showing all of the dashboards correctly for them from an executive level. So I think that, you know, the way that finance is probably appreciating you guys is they're able to get data that they have never really been able to get before. Do you think that's true?

I think spot on actually on that as well. The training that you get in RevOps is really remarkable as it relates to really truly understanding all the different elements that underpin our go-to-market, the dynamics that are at play, the various variables that we have at our disposal to like kind of try to tweak or change across years or teams or industries based on kind of what we're trying to achieve at any given time. And I think you really understand like those underlying dynamics and then the full funnel as it relates to our like product creation and R&D motion all the way through to like how we're bringing things to market and how our customers are receiving them and how we're, you know, thinking about driving growth or efficiency over the years. And so that's also something that I do think is highly valuable as it relates to then moving into other functions as well. And it's, it's remarkable how it all ties and connects together. And so last week, for example, I was going really deep on our credit and collections process. I'm trying to understand how it works and how I can optimize it and where there might be opportunities for AI, for example. But through that process, you start to realize like, oh, this thing that we were trying to solve, you know, in go-to-market ops because it was a known pain point about making sure we understood our customer information correctly and had the right contacts and the right data and that, you know, customer hierarchies when you think about these massive conglomerates like Google and that we know the right person to talk to and we have the right accurate information across all of our systems. Well, that ties exactly to, you know, the collections process or the credit check process or other things as well. And so a training in RevOps really provides a ton of value and foundational information pretty much for any other job you would ever take in a different function.

I love the way that you said that. And one of the things that's interesting to me from a RevOps perspective is that like RevOps is having a moment. All of a sudden people are like, dude, this is such an important function in our business. We didn't even know what to call it before, and now we know what to call it. But now it seems like different teams are starting to fight over who gets to be managing RevOps, right? So sometimes they report to the CFO, sometimes they report to the CTO, sometimes they report to the CRO. And so I'm wondering where, where do you think they should report?

I don't know if there's a right answer of where it should report generally. I think it really depends on where you are in your journey, your growth journey, what your goals are and what you're trying to achieve. With the company that I talked to that was thinking about RevOps under the CTO, What I thought was so fascinating about it was that they, they were a growth phase company. They were, you know, small series, maybe B venture-backed, and were in this current state where the market had some uncertainty and they were trying to find ways of like try driving more efficiency in their spending and go-to-market while also continuing to grow. And I think as part of that, they realized, well, we have an opportunity to figure out how we scale our business, but through technology as much as through people. And I think that was where this idea came about of having the RevOps team under the CTO, because the CTO was going to be building that technology that they needed to really try to take their product to customers in a way that's efficient and effective, but also through technology rather than people only. What I have found is that one of my closest partners over the last probably 5 years at LinkedIn, both in go-to-market ops and now in strategic finance, has been our, like, internal productivity engineering teams. And so those are the teams that are developing tools and delivering tools for our teams to use internally. And I think you realize that that partnership is so important because those are the tools that your teams are using as operations and strategy, but they're also the tools that the field and go-to-market teams are using, you know, to do their jobs and to grow your business and scale. And so I think that's where there's a really strategic partnership between RevOps or any type of internal strategy and operations team and these technologists that are developing the tech that enables teams to do their jobs. And I think that's becoming only more important as we look at the new age of AI and, you know, technologies just rapidly changing and evolving, and there's a need to stay on top of that and stay close to it as it relates to our go-to-market.

So do you think RevOps teams need to become more like developers and more technical?

I do. I personally do. And I think it's an exciting opportunity actually for people to get closer to some of these technologies. And I would say there is maybe two themes that I personally find pretty interesting and exciting. So one is this shift from not just being like process stewards or process-oriented to being more product-minded as well. And I think the last two teams I've led have really started to shift very formally into being more product-minded. So actually having people on the teams that were like business product owners, I'm about to hire a person in my new org that'll be leading our product roadmap and strategy as it relates to financial technology. And we had a similar model where that type of persona existed in the business in RevOps at LinkedIn. The other trend I think you see as well is more towards citizen development. And what I mean by that is how are you allowing your business teams to remain agile and actually you know, test things, iterate things, and innovate themselves using technology to, you know, speed their ability to drive value for their teams, their partner teams like the sales teams, or for the business broadly through developing of technologies without needing necessarily like an, a formal engineering team. And so, mm-hmm, I would think about that as like kind of starting to think about more of a federated model where you have your engineers that are developing the platforms and they're, investing in things like foundational data, foundational platforms that talk to each other and are synced. They're developing layers that allow for API connections for different applications. Another big one that we're seeing right now as it relates to AI is aggregators. So something like Glean or Microsoft Copilot, what it's allowing you to do is it's like allowing you to actually connect to a lot of different resources you have across your entire company. Search for them and kind of aggregate them. And then these types of platforms then enable business teams to actually start developing on top of those technologies in ways that are maybe a little bit more scrappy, but also a lot more agile and flexible. And it, it's great. And so I think this concept of like citizen development as it relates to business teams getting a little bit more interested in, you know, developing within these types of technologies and, getting their hands a little bit more dirty in these types of areas is actually really powerful. And so, one example we had in RevOps is that we started really investing in the Microsoft platform, the Microsoft Fabric and Power Apps, and we started actually being able to build like low-code or no-code applications to solve like real, often like very common use cases for our sales team, like forecasting, the ability to write forecast notes and like do scenario planning. In a dashboard that our team owned. And we started playing with, you know, connecting that to then the OpenAI API and being able to start doing things like coding the notes that reps were adding to kind of say, is this positive, you know, neutral or negative based on that commentary on the account? And then you can start also then collecting data on that and over time saying like, oh, let me actually normalize it. Based on recognizing some reps are more pessimistic or some are more optimistic, and you can actually start to really play around. Now, all this is very new, but I think this concept of having ops teams be a little bit more technically minded and get excited about this type of innovation is really powerful.

That's so interesting. It reminds me of a time at a former company when we would joke around about how this one salesperson's pipeline was like 70% made up because he would put everything in pipeline. And then this other salesperson was like 90% win rate because he waited too long to put it in the pipeline. And so that even just that one use case of sentiment data, like, and like understanding the rep and, and how soon they put things into pipeline, that, that's fascinating. And I think that's something like we really haven't explored that much yet. So there's a lot of—

exactly, yeah, that's exactly the point, right? So we've historically had to rely far too much on manager kind of adjustments and then sales ops adjustments on top and then VP adjustments on top of that because they're the ones that understand this. And like, you really recognize that like sales forecasts are never going to be perfect. It's human-entered data and it's based on a lot of different things, you know, personality types, optimism levels. It could be all sorts of stuff, but you've historically had to rely on that kind of like human-level adjustment. And at LinkedIn, what we do is we manage like a top-down forecast, which is led by a team that's completely separated from it, just looking at the data purely and making these like kind of adjustments at like a very high level. And then we have the bottoms-up forecast, which had always historically been more you know, art than science, but based on these people really having to like go in and say like, oh, Carl always overestimates, I'm gonna take him down 20%. And now it's the AI that's maybe doing that rather than the manager.

That's very interesting. And, you know, back to AI piece of it, eventually the AI will get to know each of the sales reps very well and be able to make those adjustments automatically, et cetera. Where do you see AI going in revenue operations? Do you think that this is going to eliminate jobs? I think it's going to enhance the importance of RevOps. What's your focus and what's your thought on, on how AI interplays?

Yeah, it's a, it's a great question. The first thing I'll say is that I've recognized you can't make hard predictions in this space because it's all evolving so quickly. I've personally gotten really excited about agentic technology and agents, and I've been doing a lot of, you know, interesting demos and having lots of conversations with people in that space. But one thing that's like important to remember is like this agentic technology is really only a few months old. And often when you talk to people about, you know, their use cases that they're using or how it's been going so far, you realize pretty quickly like, oh, this has been going on for 3 months with one customer or something like that. So it is, really, really new. With that said, I, I will say that I personally think of AI kind of in two broad categories. You've got AI tools and then you've got this concept of digital workers. I think the AI tools are tools very similar to kind of how you think about personal computers or the internet or the iPhone. They're tools that are helping people be more productive in their current lives, and it may be fundamentally changing their workflows, right? I think the iPhone actually fundamentally changed how people do work or consume media and get tasks done, but it's also like a core productivity tool. And I think that this concept of digital workers is one that's a little bit harder to predict and fully understand. And that's the concept of how are we gonna start onboarding technology into use cases or onto teams in ways that we may have had to rely on humans up until today. And so that then comes with a host of different questions, including just like, how do you provision a digital worker? How do you think about it? Is it like end-to-end workflows or is it specific tasks that we're trying to automate and how do we tie it all together? And so one thing that I've heard said recently that I really like too is this idea of you know, process documentation and just understanding what work people do and what work is required for different jobs is going to basically become a currency because it's really the foundations for how we can then start to bring technology into different spaces and understand how technology can augment. Um, and so I think that is like one thing for us also like recognize and understand is that you really can't just add AI on top of something. You have to make sure that there's a clear process and that there's, you know, clear foundations already in place, whether it's data or just more clean process documentation or just broader like standardization of what you're trying to solve. You need to have goals that you're trying to bring this technology in to solve, because otherwise I think it can be very disorganized and it's probably not going to drive a lot of value.

It makes so much sense. It's interesting because I was the CMO of a company a few years back that had a fascinating, uh, workflow of digitizing skin cancer slides. And ostensibly we were talking about digitizing the digitizing them just to improve the workflow and syncing them to the EMR. But there's also, we also did a lot of AI research on the possibility of, and we, we did this with our AI research, was the derm, the dermatopathologists were doing the interpretations with supervised learning and then they were uploading everything into AWS and we were able to eventually understand, train the algorithm to understand what was skin cancer and what wasn't for certain specific types. Now, that would never, ever, ever have worked if the top dermatopathologists had not been very, very carefully logging what is and what isn't, what is and what isn't, what is and what isn't for years. Yeah. And finally, we got to the point where there were enough use cases and there were enough instances that, and there was enough of a dataset that we could understand it and the computer was able to develop the algorithm. But I do think that what you're saying is so important because if there's not enough of a dataset, if your dataset is dirty, if you're unable to interpret the dataset correctly, if you teach the algorithm incorrectly, all of those rely on humans to be able to teach and build that algorithm.

So, yeah. And I think what you're talking about is spot on. And so that's related to some of these like algorithms we used to have as we think about AI, which is like machine learning, right? And often people think about trained versus untrained models with machine learning. And one of the problems with untrained models is they're very hard to explain. So when you think about explainability as it relates both to just then adopting a model, but also getting others to believe in the outputs and things like that, untrained models can be very complicated for that purpose. I think this new type of AI that everyone's talking about now with agents is— it brings, you know, new opportunities and also new challenges, right? And so what I do think we're realizing is that you often want to focus each agent on a very specific narrow skill set, and you want to have a very specific goal as it relates to that skill. So it might be, I want— overall, I want to improve my contract review process and make sure that, you know, we're just kind of streamlining how we're ingesting customer contracts or supplier contracts. Well, that may involve multiple different skills that I actually would want AI to support me with, including like quick process, you know, review and pulling out things that are anomalies are unique on this contract for me to look at closer, confirming that all the information is complete, running, you know, like basic checks as it relates to this customer. Are they already in my system? What information do I have on them? That type of thing. But for each part of that process, you may want like a different technological solution. And then you have kind of like an overall agent reviewer or supervisor that's kind of looking at what task are we trying to complete and what skills do I need to deploy based on the overall broader task or goal. And so I think that overall is, you know, like really exciting and powerful. But it also brings up this point that we need to have our processes like standardized and we need to know what are our goals when we are completing a process and what are— what's the most important information for us to know about customers and where are the bottlenecks in the process today? And those types of things are really important. And so what we found is that our biggest opportunities right now for this type of agentic technology is in processes that are already relatively mature. So we have this incredible work that the team has already done and really like centralizing, optimizing a number of processes and then even shipping them out maybe to, you know, suppliers and vendors. Who are just standardizing and running those processes like through outsourcing. And now those might be the processes that are most easily disrupted by technology because they're already very standardized. We have rules, we have SLAs, we have success metrics. We know exactly what good looks like for that process and we can now hold technology accountable to it. And so I think that's the important part is you need to find these areas where you already kind of have some level of rules, you already have some level of success. Criteria that you're able to, you know, hold either a person or technology accountable to achieving for you. And the final thing I'll say is like, I do also think people are really enamored by this opportunity that this new technology presents. But you also have to recognize that not every use case is perfect for agentic technology. And there's actually a lot of like more old school technology that still exists and still brings a ton of value today. And so that's been another thing that's been really helpful as you start to evaluate these types of solutions is like, where do I need an agent to be doing something and acting on, you know, my behalf and running a task versus where was there a technology that already existed that works really well, right? And so document extraction, there's actually technology that exists that works really well today, but maybe like translation is where LLMs can be adding, you know, an extra value to the equation that like we just weren't seeing previously. And so I think it's important to also like constantly be recognizing you don't have to disrupt everything. There's a lot of stuff that was working today that still, still will work, that works really well, and is probably a cheaper or more economical solution than trying to apply AI to everything.

Okay, you said like 10 different things that I was like, Yes, yes, yes, yes, this is amazing. So a couple of things, I had an opportunity a few weeks ago to MC an AI agent conference in New York. It was unbelievable. It was amazing. There were, you know, the technical leads from Meta and Anthropic and the RAG technology co-founder. I mean, it was unbelievable. So a few things that I learned from that conference is I sat there for 2 days and listened to these developers talk with each other. Number 1, the public data has been consumed already by LLMs. The future is in private data. Number 2, what you were talking about with infrastructure and processes being set and standardized is so important in my opinion. And that in terms of having, how we technically build the LLMs, that's pretty much being standardized right now. So as we ingest private data and as we build these AI agents, the opportunity for innovation really lies in the data and in the processes like you were talking about themselves. Number 3, when you were talking about with the AI agents, each one being individually built for others, I love what you said there because it's almost like, like separate brain processes for, and that, that build on each other. And so one of the main themes of this conference was, as we were talking about, you know, we know how to make an agent now. Now, how do we string them all together so we can start doing more and more complex work? And I, what you said there about, you know, having one agent for this and one agent for this and one agent for this, I really think that that is the future. And as our agents become capable of doing more complex operations, it's going to be because we're, we are consolidating a whole bunch of separate ones together and as we build on each other, right?

Totally.

Yeah, I mean, I think everything you said about that was right on and really the future of what we need to be looking at in terms of how we use AI agents in RevOps, but how important the RevOps, it just really underscores to me how important the RevOps function is because who's going to build those processes? And also it underscores what you said about them needing to be more technical in working with product, because how are you going to build those agents if you don't have the processes and the data correct in order to build them? You know, we have the product team technically building the, the agent, but then you need the process and you need the people and you need the data in order to ingest it in there to create something great.

So that's exactly spot on. No, I agree with everything you said. I, First off, on the conference, it sounds great. Maybe, you know, I should attend next time. I love the takeaways. I think you're right in terms of how we think about it. A lot of these LLMs or models right now are relatively like commoditized. Now we may see like breakthroughs, you know, over the years that we can't predict, but I do think you're right. It's about fine-tuning with your data. And so where you're gonna be getting value, you know, you know, when you're bringing in an LLM and thinking about how to utilize it in your own company, it's like less about, oh, I'm gonna build my own LLM and train it. It's more like, let me fine-tune it and connect it, you know, using APIs connected into my ecosystem. And you mentioned RAG technology, but like, that's a huge one. So retrieve, augment, and generate. That's a perfect type of technology to use as you're thinking about agents that should go out and read policy documents for you. We've been thinking about it for, you know, expenses or, yeah, customer deals that reps are trying to understand. Like, you know, what's our policy for this and can I quote this to my customer? How should I be thinking about this? Well, you know, that's a type of technology that can quickly answer questions conversationally for employees on a variety of different things based on your documentation and things that exist at your company. And then, yeah, as it relates to, you know, just different skills and how you start to onboard this technology and the role of RevOps, I couldn't agree more. I think RevOps really sits at such a unique connection point across these, you know, operational-minded, you know, kind of processes and programs and the field and our customer engagements and our products and the solutions that we're trying to bring to market. RevOps sits at the core of all of that. RevOps is often considered kind of a connector across the various different go-to-market functions and players. But I think that's a really powerful place to be as you're thinking about these types of technological solutions, both in terms of being able to think about the processes that we're trying to optimize or ingest or automate, but then also like what is the value that we're going to be getting, right? Because a lot of this, like we shouldn't be disrupting just for disruption's sake. We have to be thinking about the ROI we're expecting to get as well. And, you know, I think RevOps has always played that role as it relates to go-to-market as being kind of that thoughtful, more quantitative-minded voice in the conversation as it relates to just kind of our go-to-market motion, our investments, and how we can get better collectively.

Oh, that's so— it's, it's so fun talking with you every single time, Rachel. We've never had an AI discussion before, but this was fascinating. I loved chatting with you. So my last question for you today is You are clearly at the top of your game. It is what you're doing right now at the intersection of RevOps and finance and then bringing in all the technology is just fascinating. How do you stay at the top of your game? What do you do to keep yourself charged and interested in innovating? And it could be anything. It could be yoga, it could be reading books. What's your secret?

I love it. I would say, first of all, I have kind of push myself out of my comfort zone with regards to really embracing a lot of like networking and events and going out there and attending things that maybe don't sound like the most fun events to attend on a Tuesday night, uh, like a networking dinner. But it's been something that I personally get to benefit from because I am in the Bay Area and there's so much going on at all times. And so building personal connections, learning from new people, hearing how other companies are doing things has been really like monumentally helpful for me. As mentioned, I just recently like kind of made a career switch from go-to-market ops to strategic finance, and that was kind of the conclusion of a year and a half plus a search and just industry analysis and trying to understand what's out there, what problems are people solving, what's important to people, and what's exciting, and how can I get involved in whatever seems like it is the most exciting place to be. But I do think like putting yourself out there is always helpful and just having those conversations and building relationships. And there's a number of relationships I've now been cultivating over the years from people that I met, you know, random events or worked with for a brief period of time, but then realized that they're actually, you know, now doing something really cool. And I would say that's something that people talk about a lot is like those second degree connections are the ones that get you opportunities in your career. And I think that's really something that is powerful. I mean, it's easy to talk to the people you know and kind of stay in your lane, but I do think, you know, we're at a place where no one really knows what we're doing. And so having those just broad, open-ended conversations with people that you may not be in the same exact close circles with has been very helpful for me personally.

That is awesome. Well, thank you so much for being willing to come on this podcast today and have such an awesome discussion. I really appreciate your time and your insights and your perspectives on everything. So thanks again, Rachel.

Thank you so much for having me. It's been a pleasure. Business.

Yes, that's business.

Business.

Yes, that's business.