Photo courtesy of the The U.S. Library of Congress

Want to download your 2019 energy budget forecast from Snapmeter? Tim Canavan and Tom Arnold share budget forecast best practice tips in this recorded and transcribed webinar.

Tim: Hi everyone, and thanks for joining us today as we review Best Practice Tips for 2019 Budgets in Snapmeter. My name is Tim Canavan and I run operations here at Gridium.

Tom: And this is Tom Arnold, I’m CEO of Gridium. Excited to show you everything new on budgeting and to get everyone set for the 2018 budgeting season.

Tim: Great! So Tom, let’s start with the general overview of how Snapmeter works.

Tom: Yeah, so many attendees use Snapmeter for weekly energy management, but only really dig in on their budget once a year. So it’s worthwhile just reminding everyone what’s going on in the background.

Snapmeter tries to make your budget really easy and it does a lot of the heavy lifting for you. So, our little web engines are going to go into your utility account and grab all the latest gas and electric bills. They’re also going to get the interval data from either the utility-provided meter or on-site meters. They grab the tariffs as well, so we know how much that energy is going to cost you, and they also use weather data to try and normalize what a range of outcomes could be on a set of weather. That’s going to be delivered to you all on an online experience, which we’ll go through in a second.

Tim: Great! And how is Snapmeter budgeting different than more traditional approaches to budgeting?

Tom: Yeah, for many of you, this is your first year using Snapmeter as your budgeting tool. And it’s worth reminding yourself of the orientation here. You know, before Snapmeter or any kind of automated tool, you had to kind of like go get all the historical bills, lay them out, try to think about consumption trends, rate trends… you know, you’re essentially the analyst. Here, it’s completely different. The computer is doing all the hard work and it’s going to prepare a budget for you automatically. Your job is to think about what the computer is doing and say, “Do I believe that?” You’re much more a supervisor and you’re using your instincts and asking questions about what the computer is doing to make sure that it’s pointed in the right direction.

Tim: Great! So, with that background, here’s what we’re going to cover today. First up, is budgeting best practices. These are the things you need to know and consider before running your budget.

Next, we’ll show you how to review and modify forecast assumptions, and download the budget itself.

And finally, we’ll wrap up with common questions and then answer questions that anyone enters in the chat box during the webinar, as we go along. If we don’t get to your question from the chat box during the webinar, don’t worry; we’ll make sure to follow-up directly afterwards.

So first Tom, how would I find my budget in Snapmeter?

Tom: Well, first place is go to That’s where the new application is. You can reset your password here if you’ve forgotten it; there’s a “forgot my password” link. If you were forwarded this invite for the webinar and don’t have an account yet or have any other issue, just email us as and someone’ll get to you right away.

Once you’re actually in the account, you can find your budget forecast under the “energy cost” tab. So, navigate first to “energy cost” and then underneath it, you’ll see “budget forecast”. Budgets display for meters, for buildings, and for groups, but you always will find the information under the “energy cost” tab, “budget forecast”.

Tim: Great. So, how is a budget created in Snapmeter?

Tom: Well, we could have a very long conversation about that, but all you really need to know is that the computer is doing a lot of work in the background. We take in three sources of data. So, principally we’re using that meter data, we’re also using weather data. The combination of those two build the statistical model so that we know whether last year’s September bill was particularly high because of the Labor Day heat storm, or because of underlying trends. We also use billing data; that gets the tariff for us and that shows you how much your energy is actually going to cost.

One note here, there’s a lot of machine-learning technology in Snapmeter, and Snapmeter’s always trying to get the best estimate of your building. And so in terms of what the computer interprets, it’s always going to think that the last few months of energy use are really kind of what should define the budget for the next year.

Tim: Right, that makes a lot of sense. So, Snapmeter is generating your budget automatically. The first step in the budgeting process within Snapmeter is to review what Snapmeter determined your budget to be. How would someone go about doing this?

Tom: Exactly. So, the first thing I like to do is I like to pull up what the computer forecast and why there was a change. And these two tables that we’re showing here are the two principle tables. So, let’s just take this example: here on the previous 12 months, the building spent 1.19 million on electricity. For the 12 months ahead, the forecast is 1.23. So, that’s a change of 42 thousand dollars.

My first instinct is, “Okay. The forecast is being lifted by the computer. Why?” And for the why, you go to the “sources of change” table, and basically you’re trying to head-check these assumptions.

So the first one, well this makes total sense to me. This building is in San Francisco. We had a crazy weather year last year. The computer is saying that’s not likely to continue and instead, using a ten-year weather average to drive the forecast, so it’s lowering it by a little—about a half a percent. The other two are operations and rates, and these are the two ones that you want to have your own perspective on. Here, the computer is saying “lift operations” and that’s based on what it sees in the interval data and the energy model. And it’s also lifting rates, but you can change those and you probably should take a closer look at those assumptions.

Tim: Yeah, so there are two key assumptions that you may want to change on your automatically generated Snapmeter forecast: the first is rate and the second is use. And you can edit those numbers directly—you can see them circled here in orange. You get them directly here in the Snapmeter budget forecast, and then if you click on that blue checkmark to the right, your budget will automatically update with those changes.

So Tom, why would someone want to enter a rate assumption?

Tom: Well, this is the big one.

You know, rates are difficult. It is a process of kind of reading the tea leaves. Sometimes utility will actually tell you. So in California for instance, some of you have already received formal guidance from your utility rep about what to budget in terms of increases for next year. Sometimes it’s triggered by a tariff change, so Gridium or someone else might’ve suggested that you switch tariffs. They might have shown you savings for those tariffs. You might actually want to lower the rate increase based on a tariff change.

You might enrol in a program; so, you might enrol in Peak Day Pricing or Critical Peak Pricing…something that might earn you money over the summer, and you might want to put that into your rate forecast. And finally, especially in California, we now have 18 operational CCAs where you can buy power from a municipal organization instead of the utility, and some of those CCAs come with strong rate savings. So, those are some of the factors you might want to take a look at.

In terms of bundled IOU rates, for each of the territories that we serve, we will publish formal analysis and recommendations in late June, so that should be in time for everybody’s budgets, but these issues: incredibly complex. So one number might not be sufficient, especially if you’re doing fuel cells or solar or storage. So please, if you’ve got a complex situation, reach out to us; make sure that we’re talking to you and having a discussion about what the appropriate rate escalator should be for 2019.

Tim: And then why would someone want to enter a use assumption?

Tom: Yeah, so remember that Snapmeter is going try and guess what it thinks energy use will be for next year. That’s based on a statistical model that is essentially backward-looking.

You have all the forward-looking information.

So, for instance, you might know about a big tenant moving in or moving out. You might know about an energy-efficiency project that is going to come online in 2019. You might know about TIs that are going to trigger code and that space is going to become more efficient. You might know about major changes to equipment and then some of you are very active in installing batteries and distributed generation such as fuel cells or solar, and obviously that’s going to change your use quite dramatically.

So, those are the principle reasons that you see. If it’s just kind of business as usual, Snapmeter’s going to be your best tool 90% of the time. Just pick up exactly what it thinks is going to be in terms of use, and use that as a go-forward assumption.

Tim: Yeah. Could you talk a little bit more about occupancy? That’s one of the most common questions that we get, about modifying the use number in a budget.

Tom: Yeah. Many people will express it as, “Hey, my building is 60% fixed, 40% variable.” And when you start asking about where that number came from, there’s no good answer. There might have been a study that was conducted decades ago.

What we do, is if you want us to, we’ll actually study that occupancy and we use recent energy data, our energy models and we build a custom regression for you that examines the relationship between occupancy and energy use.

Often times that yields surprising results.

I think everyone that’s mechanically-inclined will understand that if a space can’t really be turned down, if it’s unoccupied because of the configuration of the HVAC system or the lighting control system, you’re not really going to see the change that you expect.

Similarly, if you condition to show because it’s a hot market, and brokers want to be able to see the space no matter what, then you’re not really going to see the change there as well. And so, often times when we do these studies we find actually a much lower correlation between occupancy and energy use than the building planned for.

By the way, it only really makes sense to do these studies if you’re having a 10% or more change in occupancy. You know, if you’re going from 75 to 78% occupancy, it’s not really going to pick it up. If you’re going from 75 to 95, it’s probably worth taking a look at this.

Tim: And how would someone do that?

Tom: Yeah, so if you want to do this, all you do is fill out this form. It’s under the “building info”. You can just pop the percentage occupancies in for each of the months. Save it and send an email to and that’ll kick off the process of doing the study for you.

Tim: Yeah, it’s really important that you put in both the historical occupancy and then your forecast occupancy.

Tom: Yeah.

Tim: Okay. So, you’ve logged into Snapmeter, you’ve reviewed the Snapmeter-generated budget, compared it to the sources of change. If needed, you’ve modified your rate and use assumptions. Now how do you get this data out of Snapmeter and into your normal budgeting process?

Tom: Yeah, well energy is just one component of your budget, as everyone knows. And so the best way to integrate this is Excel.

Here is just a single-meter budget. If you just press the download button, you get a very nicely formatted Excel file. The format of that Excel file will not change, which means you can wire this directly into your budget and if you want to come back to this a few months later and see if anything has changed or, you know, update your assumptions, you can just refresh the data fields. You’ll notice here that there’s a summary for monthly, quarterly, annual and on a meter basis. If you’re downloading at a higher level of aggregation, say at a building level or a group level, then the file will aggregate everything automatically for you.

Tim: Yeah. So, go a little bit more into that aggregation. What if I have a portfolio of five buildings, each with one or two electric meters and a gas meter. How do I get all of this out of Snapmeter together?

Tom: Yeah, so everyone I think knows about groups. Groups are user-defined aggregations of meters. So, the example that you mentioned Tim is very simple. Create a group for all those buildings. The budget will show in exactly the same place for that group. Download it and you’ll get summaries right away.

The other thing that you might find is that your accounting department wants this split in different ways. So, you might create groups for different ownership entities, different regions in your portfolio, different GL codes on your finance team. And you can edit those groups, get it all right, and then just download it at once. This also allows different people to work on the budget. So, if you’ve got a very large organization with lots of property managers, each property manager can you know, refine the budget to their tastes and then someone can centrally download this and integrate it into the other budgeting items.

Tim: Yeah, and what if you want to do… you know, you’re not really sure what rates are going to be next year, you’ve got sort of a window of changes there or you’re not sure what to do for use. You’ve got maybe a large tenant that’s moving in with a lot of networking equipment. How would you do some scenario planning within Snapmeter on budgeting?

Tom: I think scenario planning is a great tool to help people understand what the range of outcomes would be.

So, for instance, you can—as we’ve shown—easily model different use assumptions. You could create a fully-leased scenario, a “as normal” scenario and understand what the range of outcomes would be on that. Similarly you could say “high rate case”, “low rate case”; energy-efficiency project approved/energy-efficiency project not approved.

Just download them and you know, maybe name them for convenience, and you can go back to that as a record of “Hey, here’s what we were thinking as we were developing the budget.” This is especially true because a lot of you, especially in the most recent years, have kicked off budgets really early—like right now, your first budgets are probably being due shortly and will continue that budget process well into September.

And so, what you want to make sure is that you’re updating things as you get more information about leasing and projects and other factors that might affect your budget.

Tim: Alright, well that’s a high-level overview of the Snapmeter budgeting process. Some of these are common questions that we always receive, so why don’t we go over a few of those.

So, the first sort of new question and is new for this year, is gas. So, how does gas analytics and gas forecasting work in Snapmeter? What’s different about gas than electric?

Tom: Yeah, so not everyone has gas, but most of the California buildings now do have gas in the tool, and you can budget just like electric.

What are the differences?

First, gas is tremendously more weather-sensitive than electric in commercial office. So, it feels like summer has barely started in San Francisco and you’ll notice that your gas use is actually higher this year, as an example. The big technical difference is we see a lot more variety of energy data on gas. So, some of you only have monthly gas and Snapmeter will try to pull a budget out of monthly data. A lot of you have daily gas and Snapmeter will do a good job with the daily gas meters. And some of you are very lucky and have hourly gas data, and there, you’re getting really, really good estimates on what that gas usage is.

The rates are a lot less complex for gas. I think the one thing that people should think on on their gas supply is if they’ve signed a third-party supplier, some of those suppliers bill on PG&E data and we get the data. Some of those suppliers bill separately.

So, if you’re seeing your gas rates in Snapmeter are roughly 50 cents a therm, you’re missing the commodity charges and you need to either add those back in or have them loaded into Snapmeter—we can do that for you. If you’re seeing about a buck a therm, then you’re good. Everything is going through as it should.

Tim: Yeah, and actually there’s a follow-up question on gas from someone that looks to be logged into Snapmeter right now asking about… saying that they don’t see gas data available for their specific building. I guess a similar question to that is, you know, what if there are other buildings that you may have in your portfolio that are not enrolled in Snapmeter or even, you know, really small meters. Do you want to run through the process? How would someone go about doing that?

Tom: So, if it’s a known gas meter, you should’ve already gotten a reach out about “Hey, it’s live!” or, “We’re working on it.” But many of these are actually unknown. So, if you know you have a gas meter and don’t see it in Snapmeter, please reach out. It’s totally included in your subscription and we can load it pretty quickly and make sure that you get that into your account so you can use it for budgeting.

In terms of other buildings, yes. I mean, I think as long as we’re all reasonable here, if you’ve got like an outlying, triple-net building, but you want to use this for budgeting, just let us know. We’ll add it to your account and you can quickly use the same software for the budgeting process on those buildings.

Tim: Okay. And we have a question about portfolios, so looking at multiple buildings or meters at once. Can you change those assumptions broadly, across the portfolio? Or do you have to do that meter by meter? Why?

Tom: Yeah, you do have to do that meter by meter. So, if you’ve got a particularly large portfolio and you’re a tier 1 client with Gridium, we can do the heavy lifting on that, so let us know. But yes, it is meter-specific at this moment.

Tim: And there are a handful of questions about occupancy studies and how that works. Can you just go a little bit more… give a little bit more detail on the occupancy study and what information that gets you?

Tom: Yeah. The goal of the occupancy study is to determine the elasticity of occupancy, and this is a fancy term for the percentage change in energy that you should expect from a percentage change in occupancy.

Basically what we do is we take the energy model of your building, the power of Snapmeter, and we add occupancy to it—both historical and forecast. And we try to tease out what the relationship is. In some buildings, this will be a very strong relationship. In some buildings, you know, even if you have a vacant floor it’s either conditioned to show or you can’t really end-cap off those zones, therefore there’s not really the relationship that you see.

What you’ll get is a number, and you’ll get a fancy study with graphs, but what you really care about is the number so you know how much to lift the use if you were going to have a 10 or 20 or 30% increase in occupancy.

Tim: And there’s a couple of questions here about sort of tracking a budget or sort of seeing your performance in 2019 towards your budget. So I guess, how would you recommend someone track their performance in Snapmeter against the budget?

Tom: Yup. There are two things to say there: first, over years and years of experience, the best way to do variance reporting is year-on-year.

What you want to examine is how energy is changing year on year and why. You do not want to continue to beat yourself up about a budget that was piled high or padded really excessively or cut by finance at the last minute. You want to look at year on year energy.

That being said, some of it you do actually have to report out budget variance, and what my recommendation there is always as long as all of the numbers are small, the percentage variances in Snapmeter can be applied to a historical budget.

So, if in June 2019, you notice that operations is down 3% in Snapmeter, you could apply that to your same budget numbers and say, “Hey, the budget variance here is a 3% drop in operations.”

Tim: There’s one question here about adding other users or getting other people access, so if there’s a finance department that may not currently be using Snapmeter. How would someone go about doing that?

Tom: Yup, just hit, every Snapmeter subscription is unlimited users. A lot of people have external users on it as well, so if you have a consultant helping with the budget, we can put them on there. It just takes a few minutes to get a new user administered and set up.

Tim: There’s a question about water data here as well. We had electricity and we’ve added gas. What about water data?

Tom: Yeah we’re tracking water very closely. We do want to do it, it is on the roadmap. As you might know, for those of you that have large portfolios, there are thousands and thousands of water companies in the United States. But right now, for 2018, there’s no support for water budgeting.

Tim: Alright. Thank you everyone for attending. This webinar is going to be sent out via an email in video format for everyone that registered. So, feel free to forward that along to anyone in your organization who you think would be interested.

Don’t hesitate to reach out to us at Or the phone number here with any questions you have about Snapmeter or the budgeting process, we’re here and happy to help. You can also chat with us directly in the Snapmeter application—once you login, you’ll see a chat window pop up.

Have a great day everyone!

Tom: Thanks everyone. Good luck with the budgets and we look forward to helping everyone with the budget process in 2018. Have a great day!

About Tom Arnold

Tom Arnold is co-founder and CEO of Gridium. Prior to Gridium, Tom Arnold was the Vice President of Energy Efficiency at EnerNOC, and cofounder at TerraPass. Tom has an MBA from the Wharton School of Business at the University of Pennsylvania and a BA in Economics from Dartmouth College. When he isn't thinking about the future of buildings, he enjoys riding his bike and chasing after his two daughters.

0 replies on “2019 Budget Download Tips–Webinar Recording”

You may also be interested in...