Sales forecasting & predictability during times of crisis with Harrison Rose

I previously co founded a company called Paddle and I'm now working on a new venture called good fit. My first company paddle I started 10 years ago, hundreds of staff recently named Unicorn with our most recent fundraise. Then after Paddle I co founded this company called Good Fit with a with an ex paddler and I've been there for about eight months or so full time but I started working closely with notion in 2017 and when they led PADAW Series B fundraise and now super happy to be working as a Notion expert and sharing learnings with folks just like you guys and today we're talking about Forecasting and Sales Predictability. I've run quite a few webinars for Notion at this point and covered a really broad range of topics. I think we started by talking about how to navigate the tough decision of whether you should be a product led or a sales led company. We've also talked about what I guess I believe to do some of the most high leverage things we can be doing in our go to market. That being contacting the right customer at the right time, as in when customers are showing the highest likelihood to buy and delivering the best message that we can. And all of that content is available on the Notion site. Maybe we could stick some links in the chat at some point. But going into this year I really wanted to be guided more on what to discuss with you guys based on the questions I was most commonly getting from the portfolio companies. And something that was coming up time and time again recently was how do we improve kind of predictability in sales performance and by proxy, how do we get better at forecasting? I had about three or four conversations with folks about this in just the last few weeks and that's the topic of today and I don't think it's super. It's not a huge surprise to me that people are asking this question because during these uncertain times I think people yearn for some stability unfortunate enough as a founder to not have endured too many crises which have impacted the running of my business. I remember very confidently telling an investor around two years into Paddle's life that we'd never be impacted by macro factors which like nine to ten years down the line sounds incredibly naive. If anything, I think some of the macro conditions will be softened for us over time. Given we started in 2012 specifically access to capital. The first kind of macro crisis I had to lead a company through was corona in 2020. It was a particularly wild time for me. I moved to and from the US and got stuck in an apartment for eight weeks, but that's a for another day. But I always relish kind of the challenges these crises presented to us as a leader, as leaders, I guess, and how they impacted our businesses. And to have the opportunity to lead the company through this stuff was always really interesting for both yourself and to lead your employees through it too. But it's one of many moments that I think define our company and I'm sure others too. But in terms of how that relates to 2023 and forecasting, lots and lots was written about how to navigate the uncertainty around coronavirus in 2020. Notion shared some amazing advice with founders and were super supportive during board meetings whilst we were trying to work out what was going on in the world. But I don't know if anyone remembers the Sequoia kind of Black Swan memo that was released during Coronavirus. I think it was probably one of the most talked about bits of content at the time and I don't know if anyone can remember any of the things that it suggested that we do or even perhaps guess, but there was a ton of info in here on how we should try to navigate some of these uncertain conditions I guess that coronavirus brought about. And when I was preparing for this webinar, it was quite funny to see how many of those suggestions and actions were really similar to those that we're being asked to carry out today. They talked about extending cash Runway, asking could you withstand a few poor quarters if the economy stutters? And do you have as much Runway as you think? I know that a lot of folks have been asked to extend Runway during this particular macro outlook of 2023. It talks a little bit about fundraising, stating that private financings could soften as happened in 2001 and 2009. I think this again is true in 2023. We're seeing less activity, especially at latter stages, and there's lots of data around that that you guys might be aware of. And folks are ultimately looking to extend Runway when capital might be easier to raise or on better terms. They talked a lot about sales forecasts, which is our topic of today. The memo said to anticipate buyers changing their spending habits and they really talked about being prepared for this and not wanting to get caught flat footed. And today we're going to talk about only how to build an effective forecast full stop for your businesses, but also how do we ensure that it remains a useful tool during these turbulent times and how can our forecast actually guide our behavior as the world around us is changing? Going to spend more time on this in just a few minutes the memory talked about marketing. It said with softening sales you might find that customer lifetime values have declined. And in turn thus we need to kind of rein in our customer acquisition costs to maintain consistent returns. I was listening to the to the all podcast over the weekend. I don't know if anyone else listens. Maybe someone again could further link in the chat, but they explained that AOL back in the day really failed to respond to kind of market conditions. And I might have the story slightly incorrect, but I think they said AOL at their heyday were willing to spend hundreds and hundreds of dollars and funding massive TV campaigns to win like $35 subscriptions for their Internet service. And they failed to foresee how this competition in broadband was going to impact churn for them and that's LTV of their customers. And suddenly there was this real imbalance between how much they were willing to spend to acquire a customer and their failure to anticipate what impact a competition was going to have an ltv. And this put them in a real sticky situation. The memos talked about Headcount asked us to evaluate critically whether we could do more with less. Again, this is something that I'm sure all of us have been thinking about in our startups. When I was preparing for this webinar and putting some notes together and earlier in February, I think it was the 12th, I looked at layoffs, FYI, and I think it had 340 tech companies with layoffs and over 100,000 employees laid off. So this is definitely something we're thinking about in 2023. And last but not least, it talks about capital spending. So until we kind of said, until we charted a course to the financial to financial independence, we should examine whether our capital spending plans are sensible. And this is similar advice to today when fighting for budget seems harder than ever before, when selling and all expenses are being scrutinized, I think more heavily than ever before. All of this stuff seems super familiar when operating in kind of these market conditions. But one thing really stood out to me during coronavirus at the time I read this and that paragraph that really stood out to me can be seen on screen and it said having weathered every business DownTurn for nearly 50 years, we've learned an important lesson and that's nobody ever regrets making fast and decisive adjustments to changing circumstances. And I think that probably rings true both now and 2023, but probably even peacetime. Quite frankly, it's quite rare that we we regret making kind of decisive action provided the data available to us we can rely on. And that was a message that really resonated with me during the kind of only crisis I really had to navigate a company through. And I wanted to act upon this advice, but I didn't feel like I was confidently able to do so. And it was a result around our lack of predictability and poor forecasting in terms of sales performance in terms of kind of where we're at and the status quo. Talking about paddle, we were kind of burning millions of dollars per month as a fast growing kind of VC backed startup. @ the time, finance owned our financial plan and our revenue forecasts. And these were really, really basic, right? So we were predicting bookings and thus revenue to a large degree based on a headcount A's and C and an assumed percentage attainment of those account executives. I don't think that percentage attainment figure at the time was even based on historic attainment. I think it was just like a flat 80% figure. So super, super frustrated me because it didn't me the indicators I wanted to look out for to understand how coronavirus at the time was impacting my sales performance. Instead I was getting this quality feedback from my BPO sales at the time on where he felt we'd land in terms of bookings. And I didn't love this because I literally had to wait till quarter end to know where we'd land. Particularly because we had this real hockey stick style quarter usually whereby the majority of our bookings would take place in the last few weeks of the quarter. This might be something that's familiar to some of you guys, but certainly results in going gray very quickly. Like it's horrible, but no matter what we do, it seems to be a very difficult to break with this kind of hockey stick. But this frustrated me, right, because I wanted to see more leading indicators and get a better handle on what's happening with my business so I could take some of that decisive action that have been suggested and in the past. I left all of this financial planning to the CEO and cfo, but it just felt too, too simplistic. We'd furloughed a bunch of staff. I don't know if everyone remembers that. And I wanted to bring these folks back when I saw green shoots and I felt I could even kind of accelerate past some of our competition if I could really take decisive action and if I knew how our business was performing. So we decided to take the business and commercial led to planning off of finances hands for the first time and see if we could do better within Rev Ops. And the first thing that I wanted to do was build a basic bookings forecast. Right. And what I found was there wasn't a bunch of content out there on how to do that super effectively. So we muddled our way through it and stumbled upon something pretty effective but easy to replicate that I think you guys today can go away and apply. So I'm going to start by talking through how to implement this basic kind of bookings forecast and then we'll discuss some of the tools at your disposal to keep this forecast useful and relevant in a changing world. And as I run through, keep throwing your questions in the chat. We'll definitely have some time for those at the end. I think we're not going to need the today, so we'll address those at the end of the call or if Stephen or Claire spots anything super, super relevant as we run through and perhaps we can answer that at the time. I'll see if I can also get the chat up here on the left in case I can spot anything. Just a quick one. The name of the podcast I mentioned, Stephen, was the All In Podcast. If you want to dig that out and stick a link in the chat. But keep the, keep the questions coming, guys. So in terms of the bookings forecast, the starting point for bookings is forecasting based on opportunities. And opportunities are created and live in the CRM typically and against an account with like a dollar value. I imagine this is something that most of you are familiar with. Now. The very starting point for me to building an effective bookings forecast is to start by not just forecasting from opportunity created to opportunity closed. And you could maybe get away with this if the time for opportunity, opportunity creation to close is super short for your business, but that tends to be quite RA and B2B SaaS and higher value deals. Instead, you want to break the opportunity down into various stages, basically the stages that the customer actually goes through in order to evaluate your products before buying. And these stages are going to be super, super critical for us to get very early visibility into our forecast and how it's performing. And this is going to be really critical during these times of uncertainty and change. And I'm going to discuss more on that later in terms of how to do that. And I'll come back to this slide. These are the opportunity stages we used at Paddle and this was the very first set of opportunity stages we created. It's kind of heavily outdated now and that they've improved a lot, but I wanted to give you guys a real Accurate picture on how we the steps we kind of went through to get to where we are today. And the first stage that we created an opportunity level was the discovery stage. And this was once we booked in a call with a prospect, we create our opportunity. And we'd only create that once the call was booked and you needed specific things from the discovery. In order to move an opportunity forward, we needed to know the size of the opportunity, the time at which the customer wanted to make a decision by the pain being solved. The demo call needed to be booked in, et cetera, et cetera. There are various things that needed to happen in order to push forward from that discovery phase. Then we had the demo. Nothing groundbreaking there. The exit criteria for the demo was a proposed implementation plan based on the demo that was delivered and a list of objections we knew we needed to overcome. And once overcome, you'd send through your pricing proposal and integration plan. Now, we were kind of expecting different objections at each stage during this kind of opportunity creation. To close process at Discovery, we needed to nail our kind of revenue based champion addressing issues. It was usually about overcoming the technical Persona and the technical challenges of implementing paddle. And at the proposal stage, it was all about overcoming finance, budgetary constraints and demonstrating ROI to the financial leader. Now you should really spend time on what these stages are for your business and the entry and extra criteria for each. I think Jen's done a great amount of content on this in the past for Notion, but these will look really different for your business. So a good fit. My new business, we've got Discovery calibration, data feedback proposal close. Right. They're going to look different for you and your business. The thing that I'd recommend is that you set these opportunity stages up based on how the customer buys. Right. They should map to the internal hoops, decisions and evaluations that the customer is actually going through before buying your product shouldn't be the path that you want them to go down or the path that makes it easiest for you to forecast. And PADDLE certainly evolved a hell of a lot on what's on screen. It doesn't need to be perfect, but the sooner you do this the better. If you haven't already done so. Make sense so far. All with me. Just slow down. Yep, got some thumbs. Cool. So once you've got these opportunity stages and you've actually been, I guess, performing or selling for a period of time, you want to look at your historic funnel analysis, as in how opportunities are moving through those stages. And you want to look at two things. So firstly you want to look at the historic conversion rate for each stage. Right. What percentage of deals or opportunities are moving through the stages that we've just defined. And there's some sample or example data here on the right just to give you an idea of what that might look like. But then you also want to look at the average time it's taking to move opportunities between these stages. Because both the conversion rate and the time taken to move between the stages is going to be critical for us to build our forecast. And what those two things allow us to do is build what's called a weighted forecast, whereby we look at the total value of the aggregate opportunities in any stage and the percentage conversion allows us to see what percentage of those dollars are going to proceed to the next stage all the way through to close. And the time taken allows us to anticipate when those specific opportunities or those dollar values are actually going to fall. At what point are they going to close? Right. So you even want to know the time and stage at a per opportunity level. And what you end up with is knowing the number and the dollar value of the opportunities you expect to close within the quarter and those you expect to fall into the next or into the future. Right. This is actually relatively straightforward to implement once you've defined those op stages and you've got a consistent level of data actually running through your kind of sales process. Right. It's just some basic maths. I've seen people build these weighted forecasts based on, sorry, in spreadsheets or even bi tooling, lots of different places you can do this. Pretty simple. And are you with me so far? Keep, keep throwing questions in the chat if you, if you happen to have any. So that was the weighted forecast we built to understand how opportunities were going to turn into bookings. Now there are a bunch of learnings that we had in doing this which hopefully are going to be helpful for those that maybe already kind of having built out some of these weighted forecasts before. So on the previous screen, maybe to just jump back a second, I listed two different conversion rates here, some fake stats for my S and D funnel and my mid market funnel. Now, when your conversion rates, your time taken to close dramatically differ based on deal type, that deal that could be a change based on size or maybe you see very different performance in different regions. You should consider independently forecasting opportunities of those different deal types because they behave so differently and then aggregating them at the end. So you want to forecast them independently and aggregate them at the end. And that's going to deliver you a much more accurate forecast. To start with at Paddle, we had independent, we had one forecast, then we had independent forecasters for our S and P mid market and our enterprise funnel, we actually called it the strategic funnel and then we even broke that down into geospecific funnels at some point because the US behaved very differently to the rest of the world. But you want to start to forecast these things independently if you're seeing different performance and then aggregate them at the end and that's one going to tell you how your different segments are performing at any one time and are they on track, but you're just going to get a much more accurate forecast at the end. So that's definitely something I'd suggest for you guys. The next kind of pro tip that I could give you is that we also really wanted to allow reps to qualitatively feedback on the deals that they were working right. Whilst my VP sales finger in the air forecast wasn't something I wanted to rely on. They're in the deals day to day, right. And they can certainly give us a lot of helpful information and input as to where they think we're going to land. So what we did was we encourage reps to kind of rag status their opportunities. They gave it a red, amber, green. We didn't call them those exact names, but that was the idea. I think we largely avoided the red and the amber status. But when reps mark something as confident, which was basically our green status, we talked to this effectively like a commit for the quarter. They basically said that this deal was absolutely going to close this quarter. And when this was done, we overrode the traditional conversion rates that we were seeing throughout the funnel to like 90%. And it actually improved the accuracy of our forecast because someone was marking something as committed, they were damn sure that was going to close by the end of the quarter. And as we started to layer in some of these qualitative inputs to the quantitative, we got more and more accurate over time. So think about how you could do that for your own businesses. Next pro tip is that over time you should certainly check the accuracy of your forecast. Like it's a complete waste of time us forecasting if it's just completely wrong. And what I'd suggest that you do is look for the maximum error rate that you're seeing and normally you're more erroneous earlier in the course of them later. So say if in month one of the quarter you forecasted ending up at say X amount by the end of the quarter and you were 10% over or under this when you actually finish the end of the quarter, this is kind of your margin of error. And once you're seeing some sort of consistent level of error rate, you can actually start to talk about an upper bound or a lower bound forecast. I think we got to within about 10% error rate over time at paddle. And what it is is you see the number that forecasts are spitting out and then you say, look, based on our margin of error, we could be 10% above or below this figure, depending on how things go, right? This is how you start to think about an upper bound and a lower bound forecast based on how accurate the forecast you've actually built is. So you want to reflect over time on what that error rate is and hopefully narrow that margin of error over time as you get more and more data. The more data that you have to inform this kind of weighted forecast and everything that you're doing, typically the more accurate it becomes. Last but not least, on suggestions for this stuff. The favorite meeting I ever, ever rolled out at paddle, I talk about all the time. I might do a separate blog post on this independently was our weekly opportunity review. To the extent our chairman, who's a very busy man, a guy called Carson, he would often join this meeting on a weekly basis. It was on a Friday afternoon. I think he's still on the invite now. I think I am too. Despite not being full time at paddle. And what we did in this meeting was we built a view in Salesforce which listed every single open opportunity that we expected to close in the quarter. The opportunity value, its current stage, its owner, its rag status. We had an agreed next steps field which was just like an open text field and the days in stage right. So it gave us a real view of the opportunities we were currently working. And all the AES would attend this meeting. Eventually when we had too many, we split it out based on size and region. But at first it was all of them. And the sales manager would pick out opportunities at random that they were looking at in the shared screen of opportunities because we were having to be remote at the time in coronavirus. And they would challenge the rep on is this opportunity in the right stage? What's your plan to get them through to the next step? Have you done A, B and C criteria that needs to have happened in order for it to be at the current stage that it's at? And they challenge the rep around the health of the deal. The AES would often come once they got confident to this meeting and actually ask their peers to brainstorm with them on how to move a deal forward or how to unlock it. And then eventually we'd ask the AES themselves to chair the meeting because it was, it was really helpful for them to practice to critically assess the help of other people's deals and then reflect on their own deals and critically assess the health of their own deals, which resulted again, in a more accurate pipeline and a more accurate picture of how we were performing. But it even coached those AES to be better managers themselves over time. And most importantly, in the context of the forecast, it builds the practice into the sales managers and the reps to manage their day to day and their life and their deals through the CRM and through the data in it. Right? Like if we know an opportunity is the next stage and we're challenging whether that's the right stage based on whether things have happened and we can correct that, it's going to result in a much more accurate forecast. And as your sales manager start to manage people through the data that they're seeing in the CRM rather than the conversations they're having by the water cooler, the more accurate your forecast is going to be. Because if all your opportunity stages are completely inaccurate in your, in your CRM, you ain't going to be able to build a accurate forecast. And again, this weekly cadence of the opportunity review resulted in more accurate data for us. For me as the founder, and I'm sure there's some founders on the line right now, I loved it because it gave me insight into the deals I had in the pipeline, the names that were coming through. Do I know these people? Do I know some of their investors? How can I help? Help me understand what the market was struggling with, why people were buying, objections we were hearing. It was just a really great time for me to get a load of quality feedback on what was happening in the market. And then we even cycled in things like product marketers into this meeting over time. So there's some tips for you on how to do this more and more effectively. So often more and more of us are also taking more of like a product LED motion. Right. And before we talk about the impact of uncertainty in 2023 on this bookings forecast that we just built, I wanted to talk about those of you that maybe have more of like an inside sales that motion. Right. So the focus today has been on how do we build a forecast for a high value kind of B2B sales network. But the principles can also apply to like a higher velocity Motion as well, like, yeah, that inside sales team converting, you know, trial users in a product led kind of startup. Now, in these circumstances, whilst the time from OP creation, maybe when they start that trial or that card present trial, maybe the time from OP creation to close, maybe shorter, say seven days or 14 days, the length of the trial, maybe the principles are actually really similar. And here rather than breaking down opportunity stages, because maybe you're not navigating through someone through a complex sales process and there's more of a level of self evaluation, what I suggest you could do is you might want to look at the steps in the product that the user is taking or has to take before converting into a paid user and think of those steps in a very similar way to OP stages. Right? So maybe someone has to set up a product and then fill out some details of stage two. So maybe create the product at stage one, fill out some details in stage two, X number of actions complete at stage three before they become a paid user in stage four. Right. You can effectively build a very similar kind of weighted forecast based on these actions taking place any one time. I've got X number of trial users in stage one, two or three and we know why percentage of them convert within Z time period. So I think of the free trial users that I've got today, we're going to end up with X amount of revenue. Right? It's a very similar concept to that which we've just talked about. It can also be used to inform how you actually set up your trial, by the way. So if you know that like 75% of users convert into a paid customer once they've completed Z Action five times, maybe we just let all customers use it for free until they're completing Z Action five times every month. Because we know that if they do that, they're really likely to go on to become a paid user for our product. Andreas has been building a great content library around how to think about these things much more intelligently than I can suggest to you, but it can even inform some of that stuff. So taking a step back, we've established there are a number of ingredients to your bookings forecast. Right? So we've got the number of opportunities, how many opportunities are being worked, we've got the value of the opportunities, both individually and in aggregate. We've got our win rates, what is the percentage conversion rate for each of these opportunity stages which we've defined. And then lastly, we've got the deal length, like how long does it take on average for a opportunity to Move through these stages that we've defined in order to work out when these deals are actually going to close. Right? Simple enough. Now things start to get complex when we're operating in uncertainty, right? So the problem working in 2023 or during coronavirus is that you build a forecast based on the historic performance of those ingredients or those inputs. But as the Black swan memo warned us, buyer behaviors change during uncertain market conditions. Right? So this historical data that you've had, which is informing your forecast, may no longer be reliable. So what do we do? How do we react? What I'd start to do is scenario planning. Right? So we can start to plan for the impact that changes to our forecast inputs, or ingredients, as I call them, would have on our performance or on our business outcomes. You can start to project forward through this spreadsheet that you've built or wherever it lives, what a change same win rate would do for you. If my win rate is going to halve in these macro conditions, where do my bookings end up with the same amount of pipeline or how much more pipeline do I need to create in order to reach the same bookings goal of my win rate is going to have? You can start to play forward some of these scenarios based on how you think the market is going to react in the conditions you're operating in in 2023. Something we're definitely seeing, and I've seen some data on this from a bunch of folks out there, maybe like Ebster, is that when budgets are being really kind of hard fought for, and we're seeing this a good fit too. My other company, we're seeing the time in the proposal stage where you've presented ROI on pricing massively increasing. People are spending ages trying to get these proposals, all these deals approved with finance. And when I've got this forecast, it's really easy for me to project forward what longer deal cycles are going to mean for my brick and skull. If all of my deals are going to take X amount longer to land, what does this mean in terms of bookings this quarter and in subsequent quarters and thus revenue and thus cash in the bank or whatever it may be. Most companies tend to set on like a base, a lower and an uppercase scenario, and they bake in tweaked inputs or tweaked ingredients based on the current knowledge that they have as to how the market is responding and then they financially plan accordingly. Right. You might want to budget based on that lowercase or the base case scenario, and the upper is kind of your stretch goal. You might even want to target the company based on the upper based scenario, but financially planned based on one of the lesser scenarios. But then week on week, month on month, whatever necessity dictates, you can compare actual performance against these scenarios that you thought outright. And then you continue to continue to be more and more informed as to where your business is actually going to end up based on how these ingredients and how these inputs are actually evolving over time. And this is why the leading indicators are so, so key. So rather than me waiting for opportunities to actually be close one or close loss, in order for me to understand whether my conversion rates are the same from creation through to close, or whether the time taken from creation through to close is differing, I actually get a much earlier view into how the market is reacting to the macro conditions. Because I can look at is the time taken between my OP stages changing compared to my historic averages? Are my conversion rates between those op stages different to my historic averages? And you can even look at the average order value or the average contract value of those opportunities too. Is that consistent and is that changing? Makes sense. So far all good. Got some nodding heads, got some thumbs. Cool. Sadly, I'm going to talk more about problems next, now that we've, we've solved scenario planning in a bookings forecast. So at this point, suddenly I have an idea of where I expect to land in terms of a booking forecast for the quarter. If we're massively kind of expecting to outperform our plan, I can begin investing more heavily. Maybe I could bring some of those folks back from furlough taking decisive action, as was suggested in the memo. This actually happened a good thing. We built out a bunch of paddle. Sorry. We built out a bunch of scenarios after coronavirus that were relatively pessimistic. And when I compared our actual performance to those scenarios which we were running the business based off, we were actually outperforming them, which allowed us to bring back some of our staff earlier than we anticipated. Or if your performance and forecast is looking bad, you can set expectations around how your deal cycles are changing and sound the alarm bells to the business. But the real big problem that arose for us when I, when I first built out this forecast was that, and this is something that I think is even talked about less than how to build a good bookings forecast, was the big problem we had was that huge amounts of our pipeline was actually created and closed within quarter paddle as much as like 40% sometimes. It actually was a really difficult challenge, particularly when we were going through fundraising and people were Trying to understand the nature of our pipeline. So as much as nearly half of the deals we close in any given quarter are often created within quarter. So relatively speedy sales cycle, you could argue it was 40% by number, probably not by value. But the issue is that the forecast I just encourage you guys to build out allow me to effectively forecast based on the opportunities that I had in play. Right? I could only forecast based on the opportunities that were in one of those opportunity stages right now. And it didn't account for the opportunities I expected to create and close within the quarter. So I started to approach that problem in a very similar way. I started by building out a number of what became to be known as pre opportunity stages. I don't know if they're still called that. And these were largely worked by the bdrs, right? These stages were owned by the bdrs. They were responsible for moving our deals through these stages or a company through these stages if you like. The first was contacted, we'd also move an account or a company into this stage and we drop them an email for the first time or actually engage with them from an outbound or account based perspective. We had the in conversation stage. So we moved an account to in conversation after two positive responses or interactions with the client. It's quite difficult to maintain sage actually, especially given the different places in which you can actually engage with a customer. Like getting LinkedIn data into the CRMs is a real pain, but really critical. So we had the in conversation stage or the call book phase where we scheduled a call with a client and then finally the discovery phase. Right. Knows at this point that we converted this pre opportunity into a true op as you as you heard about before. And then we start running through that second set of stages that we had on screen earlier. And in building out these pre opportunity stages allowed me to do a few things. Like I could start by building a weighted forecast of the opportunities we expected to create based on the opportunities we had in play in this funnel. So how are the conversations that I have right now expecting to turn into opportunities? I'd look at the number of them, the value of them, the time the pre opportunity had been play in this funnel to determine how many of them would become actual opportunities during the quarter. But then I also needed some view into how many pre opportunities I'd expect to create before the quarter ended. And here I looked at the historic kind of six week rolling average of conversation creation or pre op creation if you like, to account for the conversations which also were yet to happen. Right. So suddenly in my forecast I'm thinking about both the opportunities that are in play now, the pre opportunities, or the conversations I'm having right now and how they're going to relay into opportunities and even the conversations I'm yet to have, and how they're going to turn into free opportunities and then opportunities and very similar concepts that apply to earlier and earlier in the funnel to give you the earliest visibility you possibly can as to how your business is performing and how that compares to the historic averages. The great thing about the pre opportunity funnel is that typically you have tons of data on it. Like normally we're having quite a lot of conversations with clients, much more so than in the latter stages of the funnel. So you can start to get really, really confident in this forecast really quickly. The interesting thing about the kind of we called it the conversational, the opportunity creation forecaster panel. The interesting thing about it for me was that where in the opportunities kind of phases we saw really different performance between SMB mid market and what we call strategic. At this kind of pre op phase, we actually saw quite similar performance across all deal sizes. So we actually just had one forecast for this with behavior, any changing at later stages in the funnel that could be different for you guys. Having built this out, you want to aggregate it all together. So we've built our bookings forecast based on opportunities and we add our pre opportunity forecast to account for the ops net not yet created to work out where we're going to land in any given month or any given quarter. We aggregate them together to work out where we're going to land. So on this example screen that's relatively basic, the light blue columns and the pipeline we expect to close based on the opportunities that already exist in our funnel. This is my weighted opportunity forecast basically. And the darker blue columns are the ops we expect to create and close in the quarter based on the conversations that were taking place right now and the conversations we also expect to take place between now and the end of the quarter. And as you get farther and farther into the future, given your pipeline has an average time to close or typical shelf life, you'll see more and more of the pipe to be closed to be made up of conversations you're yet to have. Because we're so far into the future. Right. And of course this is super useful because you can start to see if we continue at our current rate or current trajectory of demand of like conversation creation and our historic performance of closing, that demand remains the same. Where do I end up in terms of bookings month on month, quarter on quarter and then again you can start to scenario plan as to how that might evolve over time and even inform where you invest. Makes sense so far. Before I jump into some pro tips, got some nods. Cool. Again, feel free to throw some questions in the chat. We will have some time at the end if you've got any. And again, we did this aggregation I think in a BI tool called Periscope, but you could do it in spreadsheets or wherever you want. Quite frankly, the cool thing about doing it in the BI tool is that you can start to feed it with data directly from the CRM programmatically so you have this kind of more real time view of how this stuff is evolving across performing at the time. Albeit I did have a relatively data savvy kind of rev ops team, I'd probably encourage you to start in spreadsheets before moving to some of the more complex tooling. In terms of pro tips that we learned having started to evolve and build out this kind of pre opportunity kind of forecast or conversations forecast, we've got lots of learnings here too. So one of them is that you don't actually know the size of the opportunity you're working or we didn't until the point of discovery. This may be less of a problem for some of you who do like a seat based approach for example, but often you speak to a big company and you'd be really hopeful that you're going to be able to go and capture their whole business. And like maybe at the pre opportunity phase before you have that discovery call you'd size the office this massive opportunity. But in reality maybe they wanted to like trial you with a particular business unit or set of users or have some limited usage for example. So what we decided to do in the pre opportunity kind of pipeline is rather than look at the the value of the opportunity that the record input and use that in our weighted forecast, we decided to normalize the OP values up until the point of actual discovery where the the customer told us the the dollars they expected to spend or the utilization of seats they expected or whatever it may be. We normalized basically the S and B, the mid market and the enterprise into a standard opportunity value and use that to project forward. It was just more accurate for us over time. And yeah obviously as the opportunities move later and later in the funnel you get the actual dollar figure that they expect to spend with you. So that's one thing I think about. The next is that there are tools that can help with this like at Paddle. We, we really often probably didn't strike the balance right between like, I guess like I've forgotten the word for it, Adam, and eving like creating something from scratch because we thought it made sense versus buying that the tools are out there. They're already did this thing and there's some really great tools out there to help you build some of these forecasts. I'm really good friends with the folks at Clary. They're. They're a customer. Good fits and tools like that are really great at forecasting from the opportunity object. They tend to be more difficult if you're not using things like Salesforce's default opportunity objects to do some of this stuff. We ran into issues with them because we had the we were forecasting bookings based on the opportunity object in the CRM. We built this custom pre opportunity object which they could have factored into their forecasts which meant I couldn't deliver that aggregate view that I just shared with you guys, which was pretty critical for panel given how much of our pipeline was created in closer than quarter. So the less custom stuff you can do in the CRM, the more likely it is you're going to be able to use some of these out of box tools which do a really good job of this. Again, if anyone's using any cool tools like Clari or otherwise as a part of this process, feel free to throw them in the chat for everyone. Cool. And then this is just the beginning. Like again, this was all born out of wanting to better understand how the business was performing and how it was being impacted by the changing environment around me. And eventually Revops took on more and more of the commercial planning part of our kind of financial budgeting and financial plan for the year. But that work became quite a big comprehensive, scary, scary spreadsheet. You start to model for things like revenue realization, like how bookings actually converts to revenue and then cash in bank. You start to model for expansion and contraction of customers. You build some pretty complex commercial models off the basis of all of those things with scenarios in their own rights. But hopefully today gives you some basic takeaways as to how you can go and build. Firstly a basic bookings forecast and maybe a pre opportunity forecast too. If you have a lot of opportunities created and closed in the quarter, then how you can start to tweak some of those inputs and play them forward in order to understand how changes to your to your performance are going to start to impact you and your business. Because in the limited time that we've been operating in 2023 for the most people I'm talking to, they are seeing some quite significant changes in average order, value or time taken to close or win rates. And you're able to play these forward with a forecast to see where you're going to end up. But.