How Dispatch Pro Automates Dispatching

What's going on, everybody? It's Richard Kohberger here, the blue collar nerd. And in this video, we're going to be talking about Dispatch Pro, what it is and how it works. So Dispatch Pro uses AI to automatically manage your dispatch board, your schedule, and it is constantly optimizing every 10 minutes to ensure that your schedule is the best, that it can be, either the best in terms of revenue, making sure that the right technicians get to the right jobs where they're going to bring in the most money, or. Or best in terms of routing, making sure that the routing is as efficient as it can be. And it will weigh those two things differently depending on your preferences, the settings that you choose. But how exactly is it optimizing, especially for revenue, how is it determining who the right tech is for this particular job? Well, to understand that, first let's just lay the foundation of the general dispatching principles that Dispatch Pro works on. So Dispatch Pro works on the model of CSRs being a separate role from dispatchers. But in this scenario, Dispatch Pro is one of your dispatchers. So a CSR is booking the job. They're working with the client to determine what a good arrival window is going to be, what's going to work for them. So we are committing to that customer. Okay, we're going to get there sometime between noon and 4pm but what they're not doing is assigning a technician. They're leaving that field blank when they book the call. When you do that, the job goes underneath the dispatch board here in what's called the jobs tray. And then Dispatch Pro is going to pull jobs out of that bucket and stick them on the board and maybe move them around a little bit as new jobs come in, sticking, of course, within that arrival window that the CSR promised to the customer. And I want to point out that arrival windows is an optional feature in Serviceitan. So before getting started with Dispatch Pro, you should enable that in your settings and set up some arrival windows. Okay, so when Dispatch Pro is deciding who to assign to what the first thing that happens is, the list of possible technicians gets narrowed down. So we've got our full list of technicians, but you can individually enable or disable Dispatch Pro on individual technicians. So first it's looking at, okay, which technicians is Dispatch Pro even enabled on? Anybody else? We're not looking at them. Then it's going to look at zone. So what zone is this job in versus what zones are technicians assigned to? If technicians are not assigned to any zones, then Dispatch Pro assumes that that technician can work anywhere. Next Dispatch Pro is going to perform a skills match. So what skills are set up as required on this job type versus what skills are assigned to this technician? So maybe this job type requires the troubleshooting skill. And so anybody who doesn't have that, they're going to fall off the list. Then it's going to look at shift availability. So who's working during this arrival window? If you don't have technician shifts set up, then instead it's going to fall back on business hours and it's going to assume that every technician can work in every open business hour. Okay, so that took our list of all of the technicians that work for our company and narrowed it down to a few candidates. And at that point, it's going to determine which of those candidates would likely bring in the most revenue, which means that it needs to come up with a predicted revenue number for every single technician that it could potentially send on the job. So let me walk you through how it comes up with that number. So first of all, we know that when we send a technician out on a job, there's usually more than one possible way that that technician can bring in revenue. The first way is by converting the job, the job that they are on. Something's broken, they go out, they fix it, they collect money, but they might also generate revenue by turning over a lead. They go out, they see that it's going to be really expensive to fix this, maybe it doesn't make sense to fix this for the customer. And so they turn over a lead to just replace the entire system. So when Dispatch Pro is coming up with its predicted revenue numbers, it needs to take both of those scenarios into account. So you can see that with kind of our two separate threads here. So, so this top thread here is the conversion related thread. So converting the job that they're on, not turning over a lead. So Dispatch Pro is going to look at the technician's conversion rates and the technician's average ticket. And you're able to adjust in your settings how much Dispatch Pro cares about conversion rate versus the average ticket. Then down here, this lower thread is our thread related to leads. So it's looking at the lead tech conversion rate, meaning when a technician does turn over a lead, how often does that lead convert into an install and the lead tech average ticket, meaning when that tech turns over a lead and that lead sells, approximately how much does it usually sell for? And again, you can tinker in your settings to adjust how much Dispatch Pro considers one of these versus the other. And I'LL show you that later. But basically, Dispatch Pro will then blend these numbers according to your settings to come up with a technician score. Two technician scores, actually. So we have the job tech score, that's the conversion rate related score, and the lead tech score. Then it multiplies those scores by the job value model and the lead value model. So those numbers are basically averages across all your technicians for this particular job type. So for this job type, on average across the whole company, how much do we usually make if we convert the job and down here, again, across the whole company, how much do we usually make when we turn over a lead for this job type? And I'm describing these numbers as averages, but it's actually more complicated than that. These are calculated using an AI model that continually learns as you continue to use dispatch, probably. But then those numbers are multiplied together. So the overall conversion average for this job type multiplied by this specific technician's score. The technician's scores, by the way, are also specific to the job type. And then the overall value when we turn over a lead multiplied by this particular technician's score. And then we add the two numbers together. So this times this plus this times this, which gives us our predicted revenue number. So let's run an example here. Let's say we're looking at our technician, Tim. Tim is an H vac technician and he' running a no cool call. Okay, so the job value model, that's overall across the whole company, what we normally get when we run a no cool call. And let's say Dispatch Pro calculates that at 500 bucks. And let's say it calculates the job tech score at 60%. Okay? So then they get multiplied together. So job value 500 bucks multiplied by 0.6 60% expressed as a decimal, that gives us 300. Okay? And then the lead value model, let's say that comes in at 10,000. So in general, when we turn over a lead from a no cool call, we tend to get $10,000 for the replacement. And then let's say the tech score there was 50%. Okay? So $10,000 times 0.5, 50% as a decimal, that gives us $5,000. So $300 plus 5,000, that gives us 5,300. So $5,300 would be the predicted revenue number if we sent this technician on this job. By the way, there is a report where you can see those technician scores that are feeding into this algorithm. You can see that per technician. Now, something important to know is that these technician KPIs, these metrics that feed into this algorithm, they're calculated separately for historically, the past one month and the past 12 months. And those two numbers get blended together and you are able to adjust in your settings how much you want Dispatch Pro to care about that recent performance versus how much you want it to care about that further back long term performance. But by default it's set to 50. 50. That's important for a couple of reasons. It means that recent performance matters. So just because a technician was top dog for a long time doesn't mean that it's going to get skewed so much that they're just stuck there, even if they're not performing recently. But it also means that their long term performance matters too. So just because they had a bad week, it's not just going to totally tank them in the algorithm because the recent performance, it's only half of what it's looking at. And then if you have one bad week, that's only about 25% of that half. So a bad week, that's not going to mess a technician up too bad. A bad couple of weeks, a little bit more noticeable, but still not the end of the world. If they have an entire bad month though, well, now that's 50% of what dispatch Pro is looking at. Their recent performance, which is 30 days, their whole past month wasn't good. And so that is going to noticeably hurt them in the algorithm and they'll have to work to overcome that. Another important thing to understand is that Dispatch Pro makes decisions based on the context of the entire day, not just individual jobs. For example, here's a breakdown of analyses that Dispatch Pro ran when trying to decide who to assign to a particular job. And the highlighted technician is the one it chose. And you might think, what? Why look at all of these technicians that would have brought in more revenue, why would it pick that one? Ah, because there are other jobs in that same arrival window with even higher job value predictions that we want to be reserving those technicians for. Your top performing technicians are probably going to get the highest job value prediction on most jobs, but they can't go on every job. And Dispatch Pro is able to look at all of that in context. It's able to look at the entire day's worth of jobs and new jobs as they're coming in and decide, okay, we need our best performing techs on these jobs. These are going to bring in the absolute most money. Even though these best performing techs would do better on these other jobs than the ones we're actually sending Right. Your top performing techs can only perform so many jobs in a day. So it's really important that they're assigned to the ones that are going to result in the absolute most revenue possible versus using them to just squeeze a little bit of extra juice out of that worst call at the expense of burning a job that would have resulted in a lead that resulted in a $25,000 ticket had we sent the best technician to that one. Okay, so that's the basics of how Dispatch Pro makes decisions. Now let's get into the settings screen and I'll show you some of the more nitty gritty details. All right, so here we are on the settings screen and you can see we've got a few different tabs here. So right now we are on the routing tab. So the first thing that we need to tell Dispatch Pro is how important is getting the most revenue to us versus how important is making the most efficient route. So sliding more towards more revenue, we're really focusing more on right tech, right job, versus when we slide over towards routing, we're really focusing on just driving efficiency. Do note though, even if you slide it all the way over towards mostly routing, Dispatch Pro never just totally throws out the predicted revenue portion of it. It's just a matter of how much is weighed. So more routing side is better for things like probably pest control. But for most people watching this video, I would say that more revenue is a good place to start. I wouldn't start all the way over at mostly revenue, I would start at more. Just so that we're still keeping the routes fairly reasonable, but we are focusing on getting that right tech on that right job. And keep in mind, this is editable so you can tweak things later. Okay, next we can tell Dispatch Pro what our maximum drive time is. So we can tell Dispatch Pro, hey, never make a technician drive more than, for example, one hour. And you have options all the way down from 30 minutes up to 70 minutes. And you can see here that you do actually get recommendations based on other companies that are like you. So same trade and like similar size. Okay, next, we're also able to set up maximum drive time exceptions for specific zones. So we can say, all right, never send a technician more than 60 minutes away. But if they're in this particular zone, in this particular cluster of zip codes, then never send them on more than a 20 minute drive or 70 minutes or whatever you want. This is just a free type form, so you don't have to make any edits there. You can just Leave every zone at default. But if for some reason you do want to have a custom setting for particular zones, you can do that there. All right, and then finally on this page, we're going to choose how important is it for us to get the technicians close to home at the end of the day. So how much do we care if the technician's last job of the day is somewhere close to their house? Okay, moving on to the next tab here, which is locking jobs. So there's only two settings on this tab. The first one is should Dispatch Pro optimize appointment scheduling until the very end of the arrival window. So basically, okay, remember that our CSRs are assigning an arrival window when they book the job. So let's say we've got a job and we told the customer that their arrival window would be noon to 4pm so dispatch pro is free to move that job around as long as it stays within that window of noon to 4pm but with this setting, you can dictate, okay, if we're getting close to the end of that arrival window, we're. We can set a threshold of like, all right, just cut it off. Let's stop messing with it. Let's not reassign it, let's not move it. Let's keep it where it is. So if we left that setting at zero minutes in our example where we gave an arrival window of noon to 4pm, it's free to move that job around, reassign it, whatever it decides is best, all the way up until 4pm so if at 3:50, a better tech becomes available, it's gonna move it over to that tech. If you don't want that, then you can set a threshold here, anywhere from zero to 60 minutes. So I can say 15 minutes. All right, at 3:45, we're going to stop messing with this. We're going to lock it in. Okay. And the second question on this page is how much notice do you give to the next customer before the technician is dispatched? So with this setting, you can tell Dispatch Pro to lock in an appointment to stop moving it around if we're getting close to the job scheduled start time. So again, in our example, where we've got a job with an arrival window of noon to 4pm, let's say that Dispatch Pro has decided it's going to throw that job on the schedule at 2:30. If you leave this setting at 0 minutes, then dispatch Pro is still free to move or reassign that appointment all the way up to 2:29, as long as the technician hasn't already dispatched. Once they've dispatched, it's locked in no matter what. But if we put the setting at 20 minutes and dispatch Pro has that job on the board for 2:30, then at 2:10 it's not allowed to change it anymore. It's going to be locked in. So this first setting is when do we want to lock it in based on how close we're getting to the end of the arrival window. And the second setting is how soon do we want to lock it in based on how close we're getting to the scheduled appointment start time. Okay, next we've got the prioritization tab. So here we're able to do a couple of things. First of all, you can entirely disable Dispatch Pro on certain job types if you want to do that. So you can see over here we've got this column, Dispatch Pro status. They're all currently set to on, but we could change one of them to off if we wanted to. So the recommendation is to start by optimizing all of your service and maintenance job types. And I would say once you're comfortable with that, I would move sales in there as well. And from there on, it's kind of up to you what you want to hand off to Dispatch Pro. Okay, so under that we can set up which job types have higher priority and we've got priority levels one through eight. Okay, so if a job is higher priority, then Dispatch Pro is going to work harder to make sure that it books that job in the earlier part of its arrival window. So again we've got that arrival window of noon to 4pm if this is a priority one job, a P1, then it's going to work really hard to make sure we're booking as close to 12 as possible. But if it's like a P6, P7, you know, 3O', clock, 3:30, that's probably going to be fine. And this is just a drag and drop experience. So let's say an inspection that's low priority for us. A filter change that's low priority. A system check for 12 plus years, that's going to be higher priority. System check for 8 +. Still high, but not as high. Heating tune up. Meh, kind of middle of the road estimate, we could say that's P1. We want to get there as soon as possible. You get the idea, I'm just kind of making this up. But you would just go through and prioritize your jobs. Okay, so that assigns the priority of a job just based purely on its job type. But you can also use tags as modifiers, so you can assign tags so that if that tag gets applied to a job, it's automatically going to make that job the highest priority. It's automatically going to make that job the lowest priority, or it's going to move the priority up or down by one level. So let's say the potential member tag, that's going to bring it up one level. I've got a tag in here, member demand call. So maybe we promise our members that as a perk of being on the membership, they get a priority service whenever something's down for them, whenever something's broken. So we've got a member with something broken automatically. Highest priority. Maybe if this is a tenant we're going out to and not an owner, we want to bring that down one priority. New client, maybe you want to go up one priority. And by the way, the ones that move it up or down, those are cumulative, meaning if you assign more than one tag to the job, like if this one had new client and potential member on a single job, it would move up two, not just one. All right, so that's how that works. Now our next tab here is Technician Settings. This is where we're able to enable or disable Dispatch Pro for certain technicians. So if we just didn't want Dispatch Pro to consider this particular tech, we could turn it off. And you can also override the defaults that you set up for things like max drive time, how important the routing is versus the revenue, and how important it is to get them close to their house at the end of the day. So if you want most people on the default, but you've got one or two technicians that need specific rules, you can do that. Okay, and finally our last tab here is Technician Ranking. Now this tab is newer, it came out with ST69. So if you notice that it just kind of popped up appeared in my screen recording here. That's why it wasn't there when I first started making this video. I'm recording this on a different day. A peek behind the curtain of movie magic. So the settings on this tab mainly have to do with that tech score that we talked about when we were going over the calculation. So we've got that conversion rate related job tech score and then that lead tech score. And I mentioned there were some things that we were able to tinker with as far as how that's calculated. This is where that's done. So the first question here is how would you prefer to weigh short term and long term performance when ranking Technicians. So by default it's set to balanced. That's 50. 50. That's the way I described it working earlier. But you can tweak it to lean more towards long term or more towards short term, or to just only care about short term or only care about long term. The next setting is which metric do you want to use to track technician performance? So again, back to our calculation, this is that conversion rate and that average ticket. Again, by default it's kind of split down the middle, but you could lean it more towards average ticket. Lean it more towards conversion rate only. Average ticket only conversion rate. Next setting, how many jobs per job type do you want technicians to have to start tracking their individual performance? I think this one's going to make more sense if we at the same time talk about the setting underneath it, which is how would you like to rank technicians that don't have enough data? Okay, so in a scenario where we've just hired a technician, or a scenario where we've just created a new job type, we're starting from zero. Dispatch Pro doesn't have any historical context in order to make its calculations. So these two questions are basically asking, how do you want Dispatch Pro to deal with that? So when we don't have enough data to rank a technician, we have to kind of just default that technician to something. How do you want to do that? Do you want to rank them as below average by default? As average by default, or just at the bottom of the bucket? And then this question above it is basically asking, what is enough data? When do you want to throw out this default that we've set here and start actually using the calculation? Do you want to wait for them to have three jobs, four jobs, five jobs, all the way up until 10 jobs? Now, of course, the higher number that you select here, the stronger that your data is going to be when you start. But the trade off, of course, is then you have to wait longer for the Dispatch Pro algorithm to actually start kicking in for that technician. Anyways, that's all I got for today. Hey, if you're interested in Dispatch Pro, you don't already have it, then I've got a link in the description box down below where you can click to sign up. Be sure to hit like, if you like this video and found it valuable, be sure to subscribe to service Titan's YouTube channel. If you haven't done that already, hit that bell icon so that YouTube notifies you anytime we upload a new video. Video. Appreciate it. Peace.