 You'll smash it. Good morning, everybody. Can you guys hear me OK? Cool. Wait for this. Such cool. That's a title slide. Thank you so much for the introduction. Yeah, my name is Christy. Just like she said, I am the business and marketing side of Caldera Labs. We make Caldera forms, which is a form building plugin for WordPress. Yes, I am from Lima. I live in New York. I'm very far from home. So it's really exciting to be here. And yeah, you can't find me at at ccchirinos on your social media platform of your choice. Twitter, Facebook, Instagram, LinkedIn. I like to talk to people. I do this for the social media followers. So yeah, let's get started because I have a lot of content I want to share with you. And I want to lay out a foundation for what we're talking about. I talked to a lot of people yesterday. And they're like, yeah, if I enter forecasting, I'm really excited. And I was like, oh, that's so great. What do you want to know about so I can make sure I answer your questions? And they're like, I don't know whatever you want to tell us. And to me, that's exactly an answer of, I don't know. I just think it's probably a good idea. And so I'm going to assume that many people in the room that are here for the content are here because you find yourself in that situation, which is common. When it comes to WordPress products, the norm tends to be that we're self-started companies. So what I mean by that is that we invest our own money, or our own time, which is money, into our products, right? So you build a plugin and you put it out there and you try to make it work and get people to care about it and sell it on your own time while your income comes from your savings or your job or whatever. That's the norm that I've seen. People who have agencies and they start to get into products, that is, in fact, the origin story of caldera forms as well. And caldera forms continues to be a self-funded product. So this is the norm, which means that at least at the very beginning, most of you who have a product, your financial strategy is cash accounting. So what that means is you have a bank account. Hopefully you set up a bank account for your business. And you get the money, and then you pay for your business expenses out of that account. And that's it. That's the end of your strategy, which is a great strategy, because it is low cost, because it is simple, because you know how well your business is doing, because you look at your balance in your bank account and you say, do I have money or do I not have money? So that's really good. When you get started, that's why we always say set up a bank account for your business, because it'll give you the ability to engage in this very simple financial management strategy, which is, do I have any money? But presumably, you're here because you want to get a little bit more sophisticated than that. You want to deal with larger amounts of money, and that means more questions about your money. So a little bit more sophistication. It's sort of that financial modeling, but I don't really know whatever you want to tell me. And so that's when we get into this concept, because there are more questions, right? Cash accounting has a lot of drawbacks. The main one being, and I'm sure that if you are a freelancer, you've been in this problem, right, where your cash accounting won't really account for the fact that you sold a project, but the client hasn't paid you. So your account says that you are about to go out of business, but you're not. You have an incoming payment, right? You have what we call an account receivable. And so that's when we start talking about a cruel accounting, right? And so that's just a word for we think that we've earned the money when we've been told we earned the money, right? So we have the cash, but we also have the accounts receivable, which is the cash we're waiting on, right? And we also have the value of the things that we could do in case we needed more cash. And so this system starts to get a little bit more sophisticated and a little bit more complicated. Like I said, the reason, presumably, that most of you aren't doing this and aren't keeping complicated books is because it's expensive, right? And so that is where the idea of financial forecasting comes in. You are probably doing a, where is it? I'm missing a slide. Sorry. So you are doing what we would call a cost-benefit analysis. You are going in and you're thinking to yourself, well, I'm going to do cash accounting. And it's going to come with these drawbacks because you don't need a financial model to know that if you spent seven hours of your week taking care of your, well, that's a lot, but taking care of your books, then that is seven hours that you could otherwise be spending on more profitable activities no model needed, such as selling, such as talking to your customers and learning more about what they want, such as answering the questions and support tickets of your existing customers. So that's why you're not engaging in this, and that is at the core of financial forecasting. So financial forecasting is this idea that we start to think about what's coming in the future to make decisions about whether something is worth it or not. So you're already doing it, right? If you're engaging in some sort of smaller accrual accounting, right, like maybe it's not incredibly complicated, but you are doing some degree of it so that you have better planning, especially for that situation where you have clients that don't pay you and then you have to think about the money that's coming, the money that's here and the money that's going. That's at the core of what this talk is about. It's getting a little bit more sophisticated and using math and using the numbers that we have in our account to make those kinds of decisions about what we should be doing to move on to the next stage as opposed to just going on our hunches. So it gets about making decisions, prioritizing tasks, assessing performance and evaluating businesses. So a very important thing to keep in mind is that all of these forecasts have diminishing marginal returns. So what that means is I could spend the next four weeks of my life building the perfect model that accounts for every single thing I've learned over the past three years about how product businesses make money, account for trends in the season, account for trends in the market, account for the way that things grow, account for the rate of growth of different kinds of plugins. But I'm gonna hit a point where the amount of time that I spend doing this, getting it perfect, just isn't worth it. It's not gonna be worth any kind of use. And so when we move on, when I talk to a lot of especially newer people, they're like, how can I get a little bit more sophisticated about my business processes? And I say, well, what are you doing in sales? Because chances are that when we look at your cost-benefit function that whatever else you're talking about doing doesn't have as much of a return as you just talking to customers and trying to get them to buy your thing and trying to get them to explain to you why they're not buying your thing. And that idea is what we call the return on investment. If I was to build a mathematical model behind that question, I would say, well, how much is my time worth, right? Whether you wanna calculate that as your hourly rate minus the expenses that your freelance practice would have, right? Like if you were consulting minus what you would spend, then that might be your cost of time or it might be, I know that if I spend this amount of hours in X activity, then why will happen? And when we calculate that function, it turns out that we always wanna make a decision that costs us less money than it makes us. Otherwise, we're making bad decisions and that is a fundamental aspect of what we're doing. The other thing to keep in mind with this concept is the time value of money. So it's this idea that money in the future is worth a lot more, excuse me, that it's worth less than money today. So when you have your client, they don't pay you. That promised $10,000 is worth a lot less to you than having $10,000 in your bank account right now. If I go up to you and I say, do you want the dollar today or do you want the dollar tomorrow? What are you gonna say? And so when we think about future decisions, when we think about whether we're going to calculate, okay, how much time would it take me to build a perfect forecast versus what kinds of gains would I see from building the perfect forecast? I wanna make sure that I consider that the right now, the present is a lot more valuable than whatever's gonna happen in the future. And so we think about that when we make these decisions. Having $10,000 right now and then saying, oh, well, you know, but if I invest money now in five years, I might have $10,000. That's not a great decision because money right now is worth a lot more than money later. That's not breaking even, that's losing. And that is the idea of your discounted cash flow, right? So we'll be throwing around that word and that word means that when we start transitioning into these more sophisticated approaches to planning our finances and trying to grow our business, we wanna make sure that we're always considering that whatever amount of money I put into something, whether that's building the model or the thing that the model is actually analyzing, which may be a new product that you want to build, a new feature that you want to build, is actually worth the time that you're going to be spending doing. So we'll get into a little bit of the concepts behind how we would build a forecast in the context of WordPress products. The first thing to consider is that we have internal and external factors and then we also have internal and external reasons. So we actually have four things in internal and external. So your internal and external factors are your considerations for what you think is going to happen in the future based on your own product and based on what has happened outside of your product. So an external factor might be something like looking at products like yours and seeing how they do, looking at the market research that is available online on what web development agencies and freelancers make and trying to make a reasonable decision based on their income of what you could charge for a product depending on how important it is to market. And then you have internal factors when it comes to trying to make these better guesses about the future. So if you're going from this cash accounting start point, that's your internal factors. Being able to look at your bank statements, being able to look at what has happened, the way that we explain this is an internal factor is something that wouldn't exist if the product that we're thinking about didn't exist. And then you have internal and external reasons. And so we can have internal reasons to engage in building financial forecasts, right? We can have identified a planning problem. For example, with caldera forms, we found it worthwhile to create a module to predict the intervals at which we're going to need more support help. That was reasonable because otherwise people would... Support tickets would increase to a point where it got managed out of control and that would make customers unhappy. So that sort of effort was worth engaging in some sort of internal model building so that I can make better decisions for the business. When we talk about external reasons to do it, we're talking about other people want to see your forecast. So like I said, in WordPress, we don't see this often. I can count on my fingers a number of WordPress product companies that have taken out investments, have taken out loans. And if you're not aware of this, that's a really interesting thing about our industry is that many, many, many of the very successful products built themselves from the ground up, which I think is nice. But when we have external reasons, that means that someone else wants to see what you think is going to happen in your business. And these four corners of financial forecasting are important in thinking about why you're doing what you need to be doing. So we'll talk about the methods. Last time I gave this talk, they told me that this section was a little bit scary. Don't be scared. I'll be here if anything goes over your head and they'll also be available online. And also, you can Google this, right? Like your developers, you Google things when you don't know how to do them. Like you can totally Google any of this and it's all on the internet and you can just copy and paste it. It works the exact same way. So when it comes to building models, try and make better decisions. We have qualitative and quantitative concepts. So when it comes to qualitative, we're talking about the personal research, right? So it's last math phase. We're talking about talking to people. How much would you pay for this? We're talking about asking questions of our competitors. Something that I've found to be very true of WordPress products specifically is that your competitors will be more than willing to share their keys of success with you. Everybody seems to have a strategy that incorporates thinking that there is a pie for all of us to share. And I've had lots of success going up to people to do exactly what we do and saying, what do you think I should do? And they give good advice, market research. And then you have what people think of with financial forecasting, which is this quantitative aspect of it. And that's the part that really gets juicy, which is how do we take the numbers that we've collected through doing business to build things that will give us decisions based on statistics and mathematics. And so out of those, we have two things, right? Like one of them is we have the time series. So we have this idea of thinking about what has happened in the past and extrapolating into the future. And then we have the causal qualitative methods, which is a method of thinking about if this happens, then this will happen mathematically. And the very first thing that we do when we think about qualitative methods that have to do with time series is rule of thumb. So if my business made $5,000, I keep wanting to say dollars and I know that I should say dinars or euros, I'm sorry. My most basic and most unsophisticated method of building a forecast would be to say if I've made $5,000 last month, I'm going to make $5,000 this month. If I've made $5,000 last month and I've made $4,500 the month prior, perhaps this month, I will make $5,500, rule of thumb. And after we move past that, because that has some limitations, we think about smoothing. So like if you've ever seen like those graphs where they have like all the dots of the different things that have happened each month and then we draw a line among them, that's what smoothing is because we're able to sort of see the pattern behind the individual data points of what you're trying to look at. So when we have that, we use a calculation called a simple moving average. So I said this example about the month prior and then the month current. And if we want to get a little bit more advanced about that, we would say, well, what has happened in the last three months? So for example, say that I have this August, September and October thing and this is the amount of money that I made and I'm trying to, for whatever reason, make a determination about how much money I'm going to make in November. I would take the three months and I would divide between them to create my forecast. So I'm saying, well, you know, based on what I have made, based on the average of what I have earned in revenue in August, September and October, this is what I think I would make in November. But then the simple moving average is this concept that when you want to apply that idea to December, you would drop off the last month because you're engaging in a three month simple moving average, so it moves. So then this go around, I want to make this sort of assessment for December and November ended and I have figures on what actually happened. So I actually exceeded my expectations, which means that I would let go of August and I would take a look at September, October and November and take the average of those numbers to come up with the December forecast. That's a simple moving average. And naturally the example is that this can be done over any determinate period of time. You can do this with any chunks of three months. You can do this with any chunks of 12 months and so on and so forth, depending on the kind of choice that you're trying to make based on quarter, based on the financial year. So the simple moving average is a great way to get started with this calculation of how much money I might make versus how much money I might spend to make decisions, but it has a huge limitation, which is what if you have something totally unexpected happen? In that case, pretend that you did something really right and your sales five X in one month. And so if you had your simple moving average, that prediction would be very far off from what actually happened and that's when you have something called the exponential moving average, which accounts for that sort of potential situation. In this situation, we do a calculation that looks like that. There's a calculator on the internet. Like you don't have to do this. Like exponential moving average calculator. And this will create that smooth curve over what you're doing. So then in this case, I might say, well, you know, there's like a little bit of room for error and that gives me a number that's a little bit closer. You're still exceeding your expectations by a lot. That's what we would call an outlier, but it gives me a curve that has more room for error. And then when November, when December rolls around and I want to make the prediction for December, it actually puts a lot more weight in what happened recently, which tends to be true, especially with a new product and a new business. You telling yourself that what happened in three months ago is just as important as what happened in one month ago tends to be false. And so the exponential moving average is a formula that you can use to make that very simple calculation that you're probably doing in your head because you're telling yourself, well, if I'm making $100, like maybe next month I can make $200 and that would be great. You're telling yourself these things already, then you want to move on and use a little bit more sophisticated function that tells you or tells the numbers that the most recent month is more important than the month three months ago. And then you have this idea of decomposition. So decomposition is a look at potential trends. I mentioned this earlier. I said we could build a perfect revenue forecasting model for WordPress products, we could. And we would account for what we know of the industry, what we know of the internet in general, what we know of GDPs, what we know of all these things. And we would also account for this idea of what happens throughout the year. Some things that I've noticed with caldera forms that are probably going to be true for you if you are running or launching a WordPress product are that we are dead in the summer, people go on vacation. We don't have this idea in the United States, like vacation isn't a thing. And so in August our sales just like, they just die and it's like, what's going on? It's like, all the Germans are on vacation. It's very real. In the US we also have Black Friday and now that's a global phenomenon. We see ourselves pike up on Black Friday and we've talked to other WordPress products and doing Black Friday sales tends to be very beneficial for them because the entire world is paying attention to consumer electronics and so as a side effect they're paying attention to WordPress products. And knowing about these patterns is financial forecasting because this is how I know today in June that in November we're going to run a Black Friday campaign. We're low on time so let's skip over this. And then we have the idea of regression analysis. So this is sort of like complicated but not that scary idea behind putting numbers behind cause and effect relationships. So the most basic linear regression looks like that which if you like math, you know that one, right? That's like the line that goes like this or the line that goes like that which means if X happens then Y might happen. And in this case, we break that down into what you want to know where we're starting, how much something matters and how sure you are about how something may happen. So this would be a good, well used in trying to predict the effect of a certain action you might be taking. So for example, the average time of response of a support ticket. That is a model that we've built and we have seen the effect on not necessarily sales but on the customer satisfaction marker. I'm sure you've received an email that says like, how helpful was this answer? Good, not good, right? And so you've clicked on it every once in a while and so we look at that metric to see how people are feeling and we have seen a very related relationship with how long or short that average response time is to how many people click the thing and they click it with good. I'm going to also skip over this because we're running a little bit low on time. It's not that important. It's an idea that you could calculate your return on investment on word camps using a financial model. So this is something people ask me a lot. Like, does caldera firms really get a large return on investment on attending and sponsoring word camps? And I say, I don't know, but we could find out because we could find out how many people I speak to as a non-exact, right? Cost-benefit analysis function of what kind of effect I have at a word camp and I could look at a determinant period of time of sales to try and determine an effect. And then you have the idea of multiple regression. We're not going to go into this example because we don't use it in my examples but it's this idea that when you build a very complicated model, you can account for all of the possible things that may happen. So that's my big overarching example, right? It's like the perfect revenue model for a WordPress product. I would think about the country where it's being developed. I would think about the business model and the average cost of a purchase. I would think about all of these different factors that have to do with the product and then to build that model, we would simply put it all in a multiple regression. So what that number would look like, it would be where we're starting, right? So maybe like what amount of money we invested or what amount of money we are already making, right? And then I would put weights to each potential issue behind that model, right? So for example, I would think that if your median cost of your pricing tiers is $50, then that's going to go in there somewhere, right? And we would take all of those things and we can have an unlimited number of things that eventually build that model into like one of those scary things that you saw in math class that looked like this, right? So let's talk about what we would do with these sort of forecasts and practical ways in which we could use them. It's a simple process, right? Identifying the problem, identifying the relevant variables, deciding how to collect data, making the assumptions, choosing a method, and then verify. I showed you an example of this, the most simple ones are the revenue, right? The problem being I wanna know how much money I'm gonna make next month. Yeah, that's a common one. And so we think about what relevant pieces of information we would take into account to make that assessment, right? So the most common one is past revenue, right? But there are other things. Page visits, right? We're taking a close look at our conversion rate and identifying those things that might affect the thing you're trying to answer is called identifying your relevant variables. How do you collect your data, right? What tools do you have in place? How reliable are them? And making the assumptions. For example, when I said the example about work camps, that's a really rough model, but it's really hard to quantify this and it's not worth it to me to really try and quantify it that much. I'm just giving back to the community, I'm just here. So making an assumption that using the main variable as the number of people I talk to is probably gonna have an accuracy of close to nothing, but we're trying to build a model. And those are when we make assumptions, choosing methods and then engaging in that mathematical process that we saw. So a couple of examples, the first one I wanted to share with you, oh no, there was a cute emoji there, but we had to download the slides. How much support do we need for an increase in sales? This is a really good one that we did because we wanted to see if we could be proactive about the support text that we would need. We didn't like getting into this place where our mechanism for discovering that we needed to hire more support was that our customers were mad at us. That's terrible, that's a terrible way to run a business. And so that's exactly how we applied these concepts. We identified the problem and we identified the relevant number of variables. I could look at the number of sales per month to anticipate how many new support tickets I would get. I could also look at the number of tickets the last month and the month before and see the pattern of change over the number of support tickets. I would decide how to collect my data, so in this case, we're users of easy digital downloads and help scouts, so I have my data there and I decided that that would be okay. An alternative decision could have been that I want to make a very exact model and I want to engage in additional data collection strategies, but that wasn't worth it to me. My time was better spent actually talking to customers and continuing to make sales, but we needed to put in just the right amount of investment into this model to prevent the system that we had, which was that we would hire support tech when we were like, oh my God, all of our customers are so mad at us because we don't have enough support. Make assumptions, I just told you my assumption. I could, I'm going to have a model with moderate accuracy. And so, if we sold 10 licenses, we had 10 new support tickets. If we sold 20, we had 12, sure. An assumption is that we could have 11 support tickets per every 20 new sales. It's not gonna be true every single time, but we can pretend that it is for the effort of this model. If HelpScout says that we spent an average of 15 minutes on each ticket, we have way reduced this by now. I'm really proud of this. Then, let's make a conservative assumption and what that means is we always want to overestimate costs and underestimate income. So let's make a conservative assumption that is gonna take us way longer. We chose a method that fit, which in this case, we chose a simple moving average to work through the problem and understand that the past told us that we would need to hire a support tech every additional minutes of tech time. So this is how these problems start to work and how they become an effort in your WordPress product. And now the last example is sort of this idea of a three-year forecast. Now, we're done. Thank you, let's give Christy a round of applause. Thank you so much, Christy. So stay on stage and we'll do some questions. Cool. So how are we doing for time? I think we've time for questions. It's always nice to finish off with some questions. I got, so I hope I can got some time to sneak something as well. So where are our beautiful mic runners? One, two. Do we have some questions? Right in the middle there and then we will go to the gentleman in the pink at the back. So sorry, just before we get into questions, can I please, please, please encourage you wonderful audience to ask questions and not get too into comments? Hello, hello. Is the microphone working? Hi. Thank you for a great presentation, I really enjoyed it. I would like to ask you how do all these forecasts like in your business, do you use an app or just put it in Google Sheets or how does it work? You just put the numbers every month to the Google Sheet and you see the graphs or what do you use for that? Thank you. We use Google Sheets. And again, that's actually an important question, right? Because naturally in my past work and my education I could use Stata for this, I could use SPSS for this but why would I do that, right? We're running a plug-in business and this idea of the cos-benefit analysis is something I really wanna drill home because people think like, oh my God, I should be doing this and I'm like, should you? And if you should, if the answer is yes, do you need to take the most simple approach possible to get to your goal to get to the next place? So we use Google Sheets and it works great. The support example and any kind of year-to-year financial forecasts, spreadsheets, Excel, they're awesome. Fantastic. So the gentleman at the back there in the pink. Hi, it's a really good talk. I hope we can have more of those next year. All of these examples, right, give you a point estimate and if I can give you a short example, like let's say we started a Christmas shop, right? All of our products are Christmas related and we started a shop in November, right? We're gonna have like maybe 1K sales in November but then in December we have 10K, right? So what did I tell my CEO, right? Next month we're gonna have 20K, right? But probably in January, not really, right? So we can go with the confidence intervals but then I can tell my CEO, well, next month we can, we probably gonna go bankrupt or we gonna double our sales, right? So do you use anything else to kind of give like something around this uncertainty, you know, how likely is, you know, that we gonna make like, you know, 5K, 20K, you know, go bankrupt like Bayesian methods or something like that? That to me is also a question of cost-benefit analysis. If your example is exactly about the numbers that you used which were 10K in December but then, you know, 1K in November, I would question why your company is spending limited resources on trying to make predictions for January when you can reasonably assume it's going to be less than 10,000 and more than 1,000. If you were talking about $10 billion, that's a different story. And my idea then would be that you want to build a considerable forecast about what your organizational priorities are that accounts for calculation error and holds a strategy for the company to keep moving forward if you don't make the expected sales, if you don't make enough to cover your costs. With that said, if you are a $10 billion a month company, you probably have access to capital, you probably have cash reserves, you probably have a lot of those other things where this sort of like month to month consideration of revenue isn't so important. So to answer the question in the context of $10,000, I would tell the CEO, we're going to make less than $10,000. What have you made in November? What have you made in the month before? And how do those patterns look? Has there been an upward trend in the past numbers? I would say that if the trend looked like a line that then spikes up, then it's going to go down, but not as much, right? So like if the spike had it happen, it would probably look like that. And I would think about it in those terms and then run a exponential smoothing average, which is a calculator you can find on Google. Question answered. Happy? Yeah, great, cool. More questions, please. We have a few more minutes. I think we could take one or two. It is hard to see you anymore. Oh, good, all right. I can sneak my questions in. All right, so I wanted to ask you, Christy, do you have a different process for producing information for internal or external consumption? Because you mentioned at the beginning of your talk that the drivers for doing this might be internal or external. So I wondered if you did something different for either. That's a really good question. So right now we don't, that's something that I'm working on, because certainly when you're talking about external consumption, this idea of like, this is good enough, isn't really appropriate. And so setting up systems to actually have accurate, average order value calculations, to have accurate conversion rate calculations is important. So your e-commerce system goes a long way or a short way in calculating accurate data. Using an e-commerce system that is tailored towards this kind of sophisticated analysis is a nice idea. We are struggling with that in WordPress, but there are a lot of people that are making a lot of really good headways into making that happen. When it comes to internal decision making, again, I focus on the ticket price. If I am thinking about a project that has a small impact on my business, I'm going to assume that, eh, good enough is good enough. I'm going to do like a simple rule of thumb thing, especially if I'm talking about a smaller or a larger amount. If I'm talking about something very fundamental, right? So we're talking about do we want to release a new add-on and a new integration is going to cost this amount of money we can expect to get our return on the investment here. How much is the time of the existing team? That's going to be a bit more of an exact calculation, but I have all of that data within my existing systems. I know how much I pay people, right? So that number isn't going to be debated. I know how much my integrations in the past have made thanks to my e-commerce system. So that's not really going to be too up in the air and I can build those models in that way, but certainly the general rule is that you want to increase the sophistication when you are pushing out the information. Cool, awesome, thank you. All right, well, I think we'll leave it there. Thank you everybody for coming along. So Christy, you're going to make yourself available at the Happiness Bar. Yes, so if you have any other questions or you just want to talk or talk about WordPress or money or WordPress money, I am, again, on social media, I'm at Ceci Chirinos on all the platforms. You can go to KoderaForms and talk to me that way. You can go to my website and talk to me that way. I'll be at the Happiness Bar and I'll also be running around. So yeah, thank you so much for your attention and I hope you learned something new. Thank you. Thanks, Christy.