How to Forecast within 5% with Paul Shea and Chris Lowry

Sean Lane 0:01 Hey everyone, Sean here before we get to the show today, I have an exciting announcement for you all. My go to market services business mind that light Consulting has a new name, new look new feel, beacon GTM. And we are still doing all the same type of work it'd be in GTM. We are helping founders and revenue leaders improve their go to market execution. One of my partners is a former sales leader, one of my partners is a former customer team leader. And with my background in OPS, I've got a pretty good chunk of the go to market spectrum covered so we're excited to unveil this new brand for everyone to check out head to Beacon gtm.com To learn more about it. Alright, that's enough of that on with the show. Hey, everyone, welcome to operations the show where we live under the hood of companies in hyper growth. My name is Sean lane. I want to read you a quick excerpt written by Kevin Canarium, who is the President of Strategic go to market at Clary, Kevin writes, sales forecasting is one of the most important business processes to running the business. It determines how the company invests and grows and can have a massive impact on company valuation. The sales forecasting process is so much more than just calling a number, it represents the entire operating rhythm of the whole company. When the team is hitting their number quarter after quarter, the company can invest in grow with confidence. Done right, sales teams perform at their best and everyone wins done wrong. There's finger pointing and distrust and everyone loses. So this excerpt has always been one of my favorite explanations for why forecasting matters. It represents this Kevin's as the entire operating rhythm of the whole company. I share this with you because on today's episode, we're going deep into forecasting specifically how operators can contribute to the forecasting motion at their companies. And to do that, I'm bringing in two people who I know understand the ins and outs of forecasting because I did it alongside them. Their names are Paul Shea and Chris Lowry. Paul and Chris both worked with me at Drift. And while they didn't actually overlap with one another at the company at the same time, they both were the architects of our ops teams forecasting model. When it comes to forecasting, it's a blend of art and science. And so in this episode, we're going to explore that blend and how we went about building a forecasting model that regularly forecasted within 5% of actual results. In my conversation with Paul and Chris, we talked about the crawl, walk run approach that you can follow to building your own model at your company. We cover the tough conversations between sales and ops when your forecasts are different. And ultimately, whether they thought all this work was actually worth it. Now, Paul was the original builder of our forecast at Drift, and then Chris inherited the model and improved it in later years. So in this conversation, you're gonna hear from both of them and their perspectives will definitely resonate with folks differently based on the maturity of the company that you're at. Let's start with Paul, since he was there first, I asked him, What did you need to do when you were first coming into the business to assess what was available to you to kickstart this journey towards forecasting? Paul Shea 3:19 Yeah, of course, I think there's really two main points that I think of when you're attacking new business and coming in and trying to learn as much as you can. Because kind of like you alluded to, you can't start a forecast without knowing how the business works. For example, when I was at Drift, we're mainly targeted at kind of the SMB market segments, but then you go by, it's almost strictly fortune 500. And so those are obviously very different sales cycles. So you really need to understand kind of that part first, like, what are you going after in terms of the actual ICP, and then to really kind of get that understanding, there's really two ways to kind of go about it. One, I think, is with the people, and really trying to spend as much time with different sales leaders with the sales reps, understanding how they view the business, and then seconds with the data. And so getting into Salesforce, seeing how opportunities are progressing, seeing how the stages are set up. And really kind of getting into the weeds of the data and being able to balance both the data and kind of the human element, the kind of art and the science of it is really kind of how to get a well rounded perspective on on the business before you really jump in. Sean Lane 4:32 And Chris, I feel like it's almost easier if you get there super early, there's little to nothing and so building it from scratch is almost easier than having to inherit something but you very much had to inherit a bunch of stuff. When you got to drift. Like what do you remember about kind of that learning curve for you? Chris Lowry 4:48 Yeah, definitely. I mean, when I joined the forecasting model, I already had a great proven track record. So it was very important for me to, you know, not get complacent with what was already set up within the forecast. model but also learn to adapt with it as the business changed and really be proactive about it rather than reactive. So, you know, it was really important for me to kind of look at how our different business units in particular performed really what was rolling into our bookings on a quarterly basis, whether it be you know, from open pipeline that we already had, at a certain point within the quarter, or coming from create and close within that quarter, as Paul alluded to, it's really important to just kind of get to know the data as best you can, and then be able to optimize it moving forward. Sean Lane 5:30 So two things, spend time with the people spend time with the data. Already, we're seeing the balance of the art and the science required to pull this forecasting thing off. So as you come into a business, or as you approach forecasting with a fresh set of eyes, bringing the business knowledge and context that you have to the problem is critical. But of course, it's also critical that you understand the data that you're working with, and that that data isn't going to cause more harm than good. Here's Paul, if Paul Shea 6:01 it's bad data coming in, it's going to be bad data coming out. To be honest, that kind of takes a bunch of different functions to really make sure that data is accurate, you need to make sure you have to build from a sales ops perspective and infrastructures there, you have the sales managers and the sales enablement, making sure the reps kind of know exactly what they're supposed to do, I think everyone's kind of worked with teams were getting reps to fill out things in Salesforce, it's kind of like pulling teeth, Sean Lane 6:28 I don't know what companies you're talking about. Paul Shea 6:32 Yeah, it's few and far between. Making sure kind of everyone owns the data in the company is really kind of the pillar to make sure that data is clean to start. Sean Lane 6:43 And I think you can set the tone from a leadership level about the expectation of that, right? And that we all have to play a role in the cleanliness of that data and in the accuracy of it. So we could do a whole separate show about entry and exit criteria for stages and forecast category definitions and how you enable people around that. And maybe we will, but today, I want to talk about exactly like assuming you get those things, right. And assuming you've got the inputs in place, how do you then move from? Okay, we just put all this stuff in place, like, what do we actually do with it? Right? And how do we use that for all the things that we tell people we're going to use it for, right? Because we make all these lofty promises about why the inputs are so important. And we're going to use them to kind of flex this important muscle of forecasting for the business. But if you don't ever get to actually flexing, then I think people will start to question the utility of all those inputs. And so, Paul, I think, if I start with you what we kind of had as a business problem, you and I, in the early days of drift was like, Okay, how do we start to kind of call our own shots based off of all this data, for, you know, how the business is going to do from a bookings perspective on a monthly, quarterly, whatever basis. And so it was pretty rudimentary, I would say in those early days. And so if for folks that are listening to this, who are trying to kind of take those first steps towards building a system, or a model that they can use to forecast their business, take me back, and what do you remember about kind of those those early ingredients to making that happen? Yeah, Paul Shea 8:20 so I think when I got there, we basically had a weekly meeting with the sales leaders, we had a spreadsheet, and we had a column for each week. And we just said, we wrote down, here's what each leaders call was that week. And that was pretty much the forecast. And I'm sure that's how a lot of companies start relying on the sales leaders who are closest to the business to give that call. And so I think once you kind of get comfortable with that, like you said, you've built up the data to a point where you trust it, adding that second layer of hey, here's a more analytical sales operations view, to kind of supplement the leadership call, I don't think you ever want to kind of think of it as one call versus the other, but kind of how do the two together really, really support each other to come up with kind of, hey, here's your total complete view. And to start, I think, I think the best throwing a little shout out to G connector, I think that's the best few 100 bucks you can spend is to get Salesforce hooked up to Google Sheets, or to excel and really start playing around with the data. And that's where you start to build the models out. So the first thing we did was started to take weekly snapshots of our open pipeline from Salesforce by using that tool. And so we have a spreadsheet with dozens of tabs that were each snapshot of our entire pipeline in Salesforce each week, which you can imagine was was quickly grew to a pretty big spreadsheet. But by using that you can you can really start to see hey, how's the data actually change week over week? How are deals progressing? Which deals are from what stage are they kind of pushing out a month and coming out of the forecast? How quickly are they converting in basically using that was kind of our first step was like, let's look at that data. Let's see how things are converting from stage to stage and using the history of that to see, hey, if we ran that forward for for this month, what would that look like. And from there, we kind of took the crawl, walk, walk, run approach, for sure, and then start adding more layers of complexity. On top of that, Sean Lane 10:20 as Paul points out, a simple spreadsheet got us pretty far in the early days of building out our forecasting model. And when he talks about staying simple, and then adding more layers, I promise you, he means simple. One of our layers of complexity was literally just new business versus everything else. Now, we had a pretty intimate understanding on how new business worked early on, thanks to all of those snapshots that Paul mentioned. And then we just tacked on some sort of run rate average on top of that, for our upgrade and expansion dollars. This works when new business is the fuel of the company. But it can be dangerous if you don't ever get around to developing the same level of maturity or understanding for your customer base. So for us, our basic building blocks were one pipeline, we would create and close in a single period to pipeline that would carry over from one quarter to the next, and three upgrade and expansion dollars. And when Chris joined the company, those were the ingredients we handed to him. My ask of Chris was go make this thing better. So how did he do that? Here's Chris. Chris Lowry 11:27 So when I first started taking ownership of the forecast model, there was a great layout for basically looking at how our pipeline converts historically using conversion rates, specifically from each stage. And I felt like honestly, that was already optimized at a pretty high level. I think one thing that I started to realize, as I did it more and more throughout the quarters was that there was a kind of a give and take on Do you want more data? Or do you want higher data quality as the business changes in particular, so in particular, if you're moving upstream into a more of an enterprise division, your conversion rates might be going down a bit, and you have to kind of make that change of Do you want to look at the last four quarters and have you know, maybe a little bit less data but more accurate conversion rates, or do you want to use the last two to three years. So that was one piece in particular that I had to make my best decision making at another piece was specifically how we looked at creating clothes. So you mentioned, there wasn't necessarily a great way of looking at expansion forecasting. And a lot of our Korean closed comes from expansion. So that was one piece that I wanted to try and optimize moving forward as well. So the way the previous model had it working was the creating closed was just basically taking the same amount we had in that timeframe from the previous year. Well, we added in a couple of different layers of looking at it from a rep capacity standpoint, and also from how big is our renewable base going to be for that quarter. So for instance, if we had an X number of reps, the past two quarters who were getting create an average Korean close per rep of why then we're able to apply that to the number of reps we could have for that quarter. And then additionally, if we had a large renewal, renewable base for that quarter, we were able to say, you know, we're most likely going to have an uptick and our Korean close for that quarter. So really, those are two, in particular points of data that I wanted to try and improve as I took over the model. Sean Lane 13:22 So let's dig into each of those separately for a second, because I think they're both really important. So on the historical conversion rate example, to your point, what we basically would do is we would say, alright, this is what we typically see as the movement from call it stage three to closed one during a given period of time. And there were two significant changes in our business that we had to contend with when we would look at this data. The first was that we were changing who are come ICP was right, we went from a very much a small business focused company that did transactional deals, low ASP short sales cycles, to a world where mid market and enterprise became our bread and butter, longer sales cycles, more complex buying processes, larger ASP s, and so the wind rates associated with those groups were very, very different. And then the second thing that we had to contend with was COVID. Right? And basically, starting in March of 2020. A lot of your historical win rates meant nothing. Right? And so those were two kind of dynamic factors that we had to play with. What do you remember about kind of how we addressed those changes in the business? Chris Lowry 14:39 Yeah, it's another good question. I mean, in particular, I think the COVID boom for a lot of sass companies was definitely upping their win rates, and we had to really try to adjust for that. So that's kind of what I was alluding to when we decided to start looking at the previous four quarters rather than multiple years back because I felt like it was the more accurate way of Looking at our win rates. Sean Lane 15:01 Paul, what about you like when you move from drift to Hugo, by the way conversion rates work, the way sales cycles worked must have been pretty different. Paul Shea 15:10 Yeah, yeah, it was night and day, I think that drift, when I initially started to the point of having a lot of create and close, our sales cycle is less than 30 days at the low end of the market. And at YouTube, right, we've had sales cycles that have been 12 to 18 months. So it actually kind of really circles back to knowing your business and saying, and you to buy right now we don't have a ops pipeline driven model, because a lot of it's quite binary, whether the deal is gonna happen or not. And so that's where it's a your spend your time where you can provide the most value, and it's actually being built closer relationships with the sales that you're here, trying to help them and support them and really understand what the ins and outs of the deals are, to have my own perspective to say, I think this deal is going to close or push for X, Y, or Z reason whether that's the same opinion as a sales leader or a different one. But I think it's really kind of knowing the business you're going after is really, like we said the foundation of it all? Sean Lane 16:11 I think it's also probably, and I don't mean this in like a derogatory way. I feel like it's almost easier. Why do you only have one segment to go after? Right, as opposed to totally different dynamics in different segments of the business? Right? I mean, I think Chris, the way we ended up handling a lot of those variations between different segments as we basically ran unique models for each one. Yeah, Chris Lowry 16:38 definitely. I mean, we had specific conversion rates based for each of our different business units, which definitely helped improve the accuracy of each of those, and then be able to get a rolled up some of each of the individual business units. So it was definitely a lot of data that we were working with when you break it down by different segments, and then breaking it down by new business versus expansion, and then essentially breaking it down by creating closed versus open pipeline, there's a lot of different ways of slicing that data, Sean Lane 17:04 there's so much there. So let's quickly hit on three key lessons here. One, if your business sells to different segments of the market, or is shifting the segments you sell to, you should think about your forecasting differently for each segment, there is a danger of a zoomed out view, if you just look at one big number. Number two, the more upmarket your businesses, the more likely it is, you're going to have to rely on more art and less science, especially if you're just starting out. Number three, the trade offs between more data and more accurate data is a tricky one. When business conditions change, sometimes a shorter look back period is actually more accurate than considering a year's worth of data that might no longer be relevant. So we've talked about a bunch of the different factors that we considered for inputs in our model. But I also think it's important to share tactically, how we got better at storing, accessing and manipulating the data itself. Chris was the one who leveled us up in a big way from those early days of our CI connector that connected Salesforce to Google Sheets. Yeah, I Chris Lowry 18:09 had no idea you guys were doing that manually before I got there. So I appreciate you living in the stone age's and doing it that way. You know, when I had got there, we already had the snapshot of data on a daily basis in our data warehouse, and we were able to push that right into looker. So essentially, we could take a What were our opportunities looking like on this specific day historically. And then we could basically compare that to the live data that we had in Salesforce to say, you know, X number of deals in stage one at this time of the quarter ended up becoming closed one within that quarter. And that's really how we were able to get those conversion rates, and then just apply that to our live open data that we have today. Sean Lane 18:46 And the other thing you talked about as an ingredient there that I think is pretty important for people is you talked about the idea of sales capacity, Chris. And so I think this is probably a very different answer, depending on whether you're in more of an inbound heavy business versus an outbound heavy business. And I think it's also probably pretty different when you're talking about new business versus expansion. So how did you think about that sales capacity, wrinkle and build that into the model? Yeah, Chris Lowry 19:14 the sales capacity piece of it was really to help with the creating close, which again, is like, you know, a big piece of the expansion levers that we have. So for the most part, when it came to new business, I was letting the open pipeline and the historical conversion rates just kind of do its part. But as far as the creating close piece, I wanted to have a better understanding of what we could actually expect. Because if you have a different number of reps than you did, you know at this time last year, then you can't expect that same amount of create and close. So that's why it was important for us to get basically over the last two to three quarters, our total amount of Korean clothes that we had within a quarter and then divide that by the number of reps and be able to get that kind of average Korean clothes we can expect per rep and then be able to Apply that moving forward. Sean Lane 20:02 And I think Paul like this is where knowing your business really well will help you kind of pick and choose which of these lessons you pull from this conversation, right? Because a very common mistake people make is like, Okay, people book 100k per person per month. So therefore, if I just add more people, I'm gonna get 100k times the number of reps that I have, right? And so like, that's not the sales capacity, that that we're encouraging people to do here. But understanding which of those levers are truly driven by people versus the ones that are like, Oh, this inbound machine is just gonna feed a certain number of people is probably a way to think about that. How would you think about kind of what people lairs people should be incorporating into into their own models, to Chris's point, Paul Shea 20:45 knowing kind of what is coming from inbound? And what's coming from outbound? So if you do you have step one is do you have the inbound capacity to add new reps in with a high competence feel like there'll be enough volume there for them to get enough at bats and be successful. The other piece from a personnel standpoint is understanding, okay, you have a new rep come in, there's there's a ramp time to how quickly are they going to learn your product, learn the sales motion, and really get up to speed to where they can start start delivering. So you definitely need to factor that in. And that's obviously gonna be different by different segment to, at the lower end, it might be two or three months for them to get up and running. But at the enterprise level, it might be nine months is kind of your standard Ragtime. So understanding those different factors, at the personal level is really important to kind of getting the full picture beyond just kind of what the data tells you. Sean Lane 21:40 And the other ingredient you mentioned, Chris was eventually you reach this point, where you do have a pretty dynamic renewal base that is up for renewal in a given quarter. And I think when you're early stage, like we talked about before, and you're pretty focused on new biz, that number probably doesn't have quite the volatility that it might as you as you get bigger, especially if you're a really seasonal business, or have some seasonality built into your year where you have a monster q4 Every year or depending on who you sell to, there might be a budget cycle for your customer base. When we got to that point, that renewable base one, we had to kind of have the early prerequisite of having the data, right, which is not something that people you know, should take for granted that having the the amount up for renewal in a given quarter was, you know, a big deal. And then we also kind of started to think about kind of the time horizon with which we could predict those renewals. And I think one of you mentioned crawl Walk Run before, how did you kind of like start to build that renewable base into it, and then also start to change the behavior of the team, such that the data we were getting about those renewals started to become more reliable, because it certainly wasn't at the beginning. Chris Lowry 22:59 Yeah, so as far as looking at the renewable piece of it, it became really a game changer. When we started to add a new field, will you call it pipeline date, rather than create a date of an opportunity. And essentially, what that would do is look at the first time we start seeing positive value on that renewal. So if we have, you know, essentially an opportunity already created in Salesforce for renewal, but it's going to be listed as flat, or no change, etc, we can't really do too much with that, we don't really know if that's going to end up being you know, a churn that could end up being an expansion. But what ended up really helping us was being able to start stamping the day of when the reps started to see a potential of expansion on that opportunity. And that's what we use this this pipeline day. So being able to start looking at that and analyzing that data helped us to understand, you know, if we have a really large renewable base this quarter, and we already have X percent of them showing positive pipeline potential, that were able to kind of increase the uptick of creating clothes and expansion that we would have for that quarter and be able to have a little bit more accuracy for larger renewable quarters rather than smaller, renewable quarters. Paul Shea 24:08 Pretty slick, right? Sean Lane 24:10 And if you're sitting there saying, Shawn, I don't even have a way of tracking all of my renewals yet. Don't worry about it, I get it. This is exactly where the crawl, walk run approach comes into play. I'm a strong believer that how you pair all of this data with the behavior expectations of your go to market team is what really matters. For example, in the example that Chris just described about forecasting the future expansion potential of a renewal, if your team isn't looking out for that potential, and more importantly, they don't know how to then signal that potential. Then the beautifully crafted pipeline date solution that Chris just explained, doesn't matter. Everyone wants to get further ahead more proactive on their renewals, but that doesn't happen overnight. So maybe you should just set a goal that says we want to get 20% of our renewals signed up more than 30 days prior to the renewal date, then it's 40% of renewals, then it's 60. But you need initiatives to drive that behavior. And you need to reiterate the why behind those initiatives over and over and over again, and you need to measure them. Otherwise, there's not going to be any evolution in your model. Okay, up next, let's talk about how you make the shift from building your model to actually using it. For us in operations. It started with who made forecasting calls in the business and how we became one of those calls, and how that compared to the calls of our colleagues, like I Paul Shea 25:36 said, at the beginning, it was just hey, let's roll up what all the sales leaders are forecasting for their business. So that's, that's kind of your your sales call. When we first made the model, I think we kept kind of our ops call in the background for the first few months just to kind of test the data and make sure we were at least getting close before we shared that with everyone else. But once you feel comfortable with that, you bring that to the table. And I think part of its going back to the relationships you have with the sales leaders is you don't want that to be a kind of a combative of hey, minds, right? Yours wrong, it's one you need the sales leaders to tab and understand the value of the data behind it in that perspective. And really then kind of, we worked with the CRO at the time to say, hey, I have both these two models, they're both these two calls, how can I take those together to really make an informed decision about what I think is going to be most accurate. So for example, I think for the first few months, especially as we're, that's at the time, where we're really starting to go up market into enterprise, we felt comfortable using the ops model at the lower end of the business. So we have more data, we used it for, I think, the SMB segment. And maybe there's some sort of hybrid between our call and the sales call for the mid market segment. But early on, we fully relied on the enterprise sales call to kind of layer on top. So you're really kind of taking the perspective of, hey, we need to use these two different points of view together to come up with the best answer and not look at them kind of in a vacuum of each other. Sean Lane 27:09 And both of you were actually really good at this, those conversations with the sales leaders when their number was different from yours, right on a per segment basis on a per leader basis. You both were really good at you know, I wasn't in those meetings, but the manager would come out and be like, Hey, I met with Chris, I met with Paul and like, I feel so much better about my call. Now, Chris, what do you remember about those conversations? Or what were tactics that you use in approaching them so that you were building these like really trust driven relationships with sales leaders? Paul Shea 27:45 Yeah, Chris Lowry 27:46 honestly, it was providing and explaining the data to them in multiple different ways. And getting to a fairly similar number. So we obviously are forecast model is kind of our single source of truth for when we're inputting our operations forecasts. But there's obviously different ways of getting to that number as well. So what I would try and do was essentially walk those managers through the specific inputs that were getting us to that number from different ways. So for instance, I would take the number or the each rep individual rep that they have rolling up to that manager, and we would look at their starting pipeline in previous quarters and what their historical conversion rates for that quarter was and be able to get all right, this was the amount of bookings we can expect from their starting pipeline. Historically, each of these reps on an individual basis have had this amount in Korea and close so we can add that in as well. And essentially, you can get to a number where they're starting to see Yeah, you're right. This is the actual performance of these reps, who I am now managing, and it really built that trustworthy relationship. Paul, Sean Lane 28:50 anything you'd add to that with those conversations, Paul Shea 28:53 the approach I would take to start as is really kind of tried to first understand how they got to their forecast number, and what kind of their perspective is on that in it, that's going to kind of help easily kind of narrow down Why are two numbers different? Now, it might be something if the sales calls higher than ours, maybe they're a little bit more optimistic on new reps kind of getting ramped up than we've seen historically. And we can kind of say, Hey, here's, here's our perspective, what the data tells us, maybe we pull that back. But then on the other side, maybe there's a big outlier deal. That's maybe two or three times the average sales price of that segment. And on the modeling side, we're kind of discounting that quite a bit, because we haven't seen that come to success very, very often, but they have the inside knowledge of that deal to know, I know, this is where we're at the finish line here. This is 95%. So that's why I'm calling so much further ahead because of this one outlier. So I think it's kind of once you understand how they're building their forecasts, you can use the data that we have from our model to really help them wonder Stan your point of view, and then kind of, ideally, you're walking out of that meeting with, with one number that you both believe is the true call. And it's kind of coming from both sources there. Sean Lane 30:11 The core takeaway for me, and what you just said is like, there are known strengths and weaknesses of each of those calls. And the more you can lean into each other's strengths and weaknesses, the better, right? Like, right, I completely agree with you that a rep, and a manager will always know their deals and their pipeline way better than we will wave at certainly way better than a model will. Right. And so if that big outlier deal is checking all the right boxes, and that's going to be the thing that makes or breaks their number, but they've done all the right stuff along the way, great. But then where we can kind of counterbalance that and say, okay, you know, what we know from our data is that the legal procurement stage, whatever it's called, at your company, on average in enterprise takes 33 days, you're not even in that stage yet. And you're saying this deals are gonna close and 80. Right. And those are the places where we can we can poke and push a little bit in a way that is helpful for everybody and makes everybody better. The Beauty and the curse, Chris, of having our own model is that like, it is wildly objective, you put version rates and live pipeline data into it, and it spits out a number on the other side, no matter our own feelings about what the output is, or what about the deals that are inside of it, which can be great, but the curse is real, where it's like, well, offices is calling something wildly different from what the team is calling. We've had a lot of conversations about this, like, how do you feel about how do you reconcile that when, you know, the rest of the team is looking at the office number and saying, Well, this is this is just not right. Yeah, Chris Lowry 31:45 I mean, it's truly a numbers game, where we are just taking the number of opportunities we have and the historical conversion rates that we have, and basically being able to get that analytical number, whereas the managers and the reps are going one by one for each deal. And being able to say, you know, we have an executive sponsor on this deal. So it's going to become closed one. And it becomes really tough, you know, when we were getting the stakeholders asking, so in your forecast model like this, this deal end up being close, and you have to kind of explain to them like, we're not looking at it by an individual deal basis, you know, this is basically a rolled up some number that we're expecting here. So we can't get to that level of granularity, I think we have definitely had the pressure of trying to raise our call at times to get closer to the managers and the higher level stakeholders number and kind of what they're hoping for. And we've had to really, you know, stand our ground and say, We're sticking to the math. That's the type of work that we do. And as operators here, we have to be nerds and listen only to the numbers and in the I really stick to what we believe in here. Sean Lane 32:49 What these guys are saying is way easier said than done. But the beauty of this show is that they've actually also done it. So you know what's possible. And the results that they achieved speak for themselves. Paul and Chris, along with a long roster of sales, CS and ops folks created a model that was regularly forecasting bookings within 5% of actuals. And they touted a trailing eight quarter average of forecasting within 7%. But wow, does everything we've talked about sound like a lot of work. And I know everyone listening to this has a ton of competing priorities. This is not a set it and forget it project. It requires maintenance, attention, and constant upkeep. So even though it worked, even though we're proud of what we built, I asked Paul and Chris, looking back now, do they think that this type of effort is worth it? Paul, first, I've Paul Shea 33:43 actually asked myself that in my current job, and you'd wait kind of in I think part of it is understanding your team, you only have so much time in a day yourself, and how do you want to prioritize what you're focused on it at your job day in day out so that you're providing the highest value to business? And so when I look at Drift, I said, Yeah, I definitely think it was worth it. At that point, we had grown to a pretty sizable renewal base and ARR base, and that forecast number was coming more and more important. And so having that added perspective, really helped the senior leaders kind of understand, hey, where's the business going and making decisions off of that? And you might right now we're earlier stage in where, like I said, focus on the fortune 500, where the deals are more binary. So and from a resource perspective, I think I was like, the 10th or 12th person on the greater ops team adrift when I joined and we were 300 people and I do right with 300 people and we have an ops team of two. I think you have to kind of take in those different factors to kind of understand when's the right time to actually roll that out and understand what that work behind it to maintain it's going to be what's those? What's that day to Operation look like after you build it? And so I think it's gonna be different for every company, it goes back to understanding your business, how valuable will this kind of second view of a forecast be? And I think there's no question coming from from the upside, we probably all believe like, there is going to be a point where, where it's going to be worth more, it's going to be valuable. And you're going to want that. But what we kind of said from the beginning is we're assuming all the inputs of this are right, and the data is right. And the other pieces are you at the stage in your current company where you're still building up an infrastructure, and you're still building up the sales enablement, to get reps on board with, Hey, this is the definition of this stage and in trying to get consistency across the field. So I think it's really about knowing kind of when's the right time to pull that trigger? Chris? Yeah, Chris Lowry 35:45 I think palsa for the most part, everything there. But one thing I will add, for me, it's are you going to be listened to as well and I drift, we luckily have a very good relationship between our operations team, and you know, our sales stakeholders and sales reps as well. So they do value the input that we have when it comes to our forecasting and the methods that we're talking about when we're forecasting. So the fact that we can put out a number, and our managers are asking us, how is that going to change? Like, how does it change this week? How does it look now, like as we're halfway through the quarter, the fact that they are coming to us and curious about it, and as you said before, even kind of compete against our number as we move throughout the quarter is kind of a fun, and, you know, competitive challenge that we have, and it's all in good spirits. And you know, we appreciate it from our side as well. Sean Lane 36:44 At the end of each show, we're going to ask each guest the same lightning round of questions. Ready. Here we go. Chris, I'll start with you. Best book you've read in the last six months. It doesn't Chris Lowry 36:54 have to be just business. It can be anything. Dope, anything. greenlights Matthew McConaughey. Book, Sean Lane 37:00 nice. Did you read it? Or do you listen to the audio version with him? They're both. Nice. All right, Paul, favorite part about working in ops? Paul Shea 37:10 I'd say it's, for me getting to work with the business itself. And so having that kind of data, analytical background and kind of bringing that toward to the business to the sales leaders and leaders of other organizations who might not have that same kind of perspective and working together to really make the business better. Chris, Sean Lane 37:31 Flipside least favorite part about working in ops? Chris Lowry 37:35 I don't have a good answer. Because I'm not even technically in operations anymore. So Sean Lane 37:40 all right, all right. That's a cop out Paul Lee's favorite part about working in OPS, Paul Shea 37:44 least favorite, I'd say it's pretty timely, with a lot of companies here flipping into the new fiscal year. But that flip is always a grind. getting everything set up for the new year can be real tough. So I'd say that that's a good one. Sean Lane 37:58 Chris, someone who helped you get to the job you have today, I'll go Chris Lowry 38:03 with shout out to Kyle Feldman, who basically was able to pull me more from the operations perspective into a true bi role where I was able to take that go to market in front of the business, you know, context that I had, and then be able to get a little bit more technical with the data on the BI side and really be able to combine those two skills. So I appreciate Kyle being able to, you know, take a chance on me without having like true sequel capabilities at that point, and being able to kind of teach me and bring me into it a true bi manager role. Sean Lane 38:30 Paul, how about you, if someone who impacted you got into the job, you have to give Paul Shea 38:33 a shout out to my current manager, Eric Landon, we work together in the past at Pivotal for a few years, never directly on the same team. But he was an ops, I was in finance at the time when he went over to you. But he reached out to me about kind of joining him. And we've been a two man tag team ever since. That's Sean Lane 38:51 awesome. All right, last one for both of you. One piece of advice for people who want to have your job someday, Chris, I'll start with you get Chris Lowry 38:59 really into the weeds of the data. There are so many different tools and skills out there that you can learn, but get really good with Excel, start working through SQL, get ahead in the coding game and start learning your Python and R. I think if you learn those three skill sets, you'll be in a really good spot. Sean Lane 39:17 You are a great advertisement for that. And so if people want to figure out how to do all that stuff, they should reach out to Chris. All right, Paul, one piece of advice for people who want to have your job someday. For Paul Shea 39:25 me, I'd say it's trying to go early in your career as as broad as you can. I've actually my role now is both operations, only half of it. I also kind of run out of finance and FPN a team here. And so I've had roles on the finance side on the ops side. And I think getting the perspective of of different roles early on in your career. One will kind of help you decide, here's the path I want to go down and what I like doing best but but also when you're in that role down the line, you have different perspectives from your prior experiences that really can We can add value there Sean Lane 40:09 thanks so much to Paul and Chris for joining me on this week's episode of operations. Also, thank you to both of them for all the work that they did to make this model happen and all the other work they did alongside me for many years. Special shout out to Wilkie Collins, by the way for instilling in us the mentality of why this was so important and also instrumenting a lot of this in the early days so that we could be snapshotting data from very, very early in the company's history. If you'd like to be heard today, make sure you subscribe to our show we get a new episode every other Friday into your feed. Also, if you've learned something from Chris and Paul today, which you must have, make sure you leave us a review on Apple podcasts or wherever you get your podcasts six star reviews only. Alright, that's gonna do it for me. Thanks so much for listening. We'll see you next time.