Capex Planning & Execution at All Business Levels | Michael Tremor

Speaker 1: ...my name is Michael Tremor. I'm a senior analyst on the corporate applications team at Vistra. I started there about three years ago during our Planful implementation. And I've been serving as the primary support for both our finance and accounting teams since. I'll jump into my favorite Vistra- related anecdote. So, back in 2010, I graduated from Virginia Tech with an engineering degree. My first internship while I was there was at Bechtel, an engineering procurement and construction firm. And I worked on the design of a lignite- fired, coal power plant that had just been awarded. Fast forward about 12 years, and on my first day at Vistra, they are actually the parent of the company that commissioned that power plant. So I found out on that day that they had finished design, finished construction, gone into operations and decommissioned that power plant. So that's the oldest I felt to date, and it's a great reminder of how fast life can move. So, coming into Vistra, I had a limited background on the finance side of things. Having that construction background, that's really where I came from. But I found Planful relatively easy to pick up. And I've been growing every day, along the way. As a SaaS platform, I appreciate Planful on the IT side, because it takes away a lot of my headaches. I'm not worried about application patches, I'm not worried about server vulnerabilities, I'm not worried about network or server issues or critical down times during a crucial close. These all get taken care of for me, so life is fairly easy. Another of the benefits I've seen with Planful is related to the pandemic. With the transition from working in an office to working from home, there was a lot of concern about impact to everybody's ability to continue doing their jobs. With Planful, no internal network connection is required. It's a safe and secure access through our single sign- on portal. So users were able to continue their work seamlessly, whether they were working from home or working in the office. And that brought a period of uncertainty and made it a lot simpler. So Planful has definitely been a valuable tool for us at Vistra. Speaking of Vistra, we are a integrated power, or integrated retail electricity and power generation company. We're a Fortune 275 company, based just outside of Dallas in Irving, Texas. So what is an integrated retail electricity and power generation company? Other than a mouthful, it is a company that, on the retail side, operates in competitive markets around the country. We have about four million customers between residential, commercial and industrial. Then on the generation side, we have 39, 000 megawatts of capacity, and we operate in wholesale markets, selling our power there. So, for a lot of you, you might just get a bill from the electric company, but we operate where consumers have a right to choose their power provider. And we offer different rates, different packages you can select, such as a green energy- only package. So there's a lot of benefits to being able to pick and shop for your electric provider. To give some context around our Planful configuration, we have about 200 active users, which is quite a lot, and a lot of them are also operating in Spotlight. So it's a very heavy implementation. I'd consider us one of the biggest customers here. But like I said, not a lot of headaches on my end. We have been using the tool for about three years, both on accounting and finance side. Our consolidations are run out of Planful, but just a little different than everybody else. We run our consolidations in modeling, with a model feeding some of our equitization and reclassification data. So, we're not strangers to nonstandard solutions at Vistra. In fact, another one of our quirks is that instead of using the typical Boomi integration, feeding your data in from your ERP, we're a little different and consume Planful's web service API offering, and feed our data over, on a nightly basis, via those API web service data loads. So, definitely not your standard implementation, but it's one that's been successful for three years. So why are you here today listening to an IT guy, with an engineering degree, talk about CapEx? Well, I hope to cover a few things of value for you here today. First, I'd like to talk about CapEx planning at our corporate level, and how a large corporation with employees all over the country can leverage Planful to simplify tasks and empower the business. Next, I'd like to build from that context and talk about our IT planning and some of the flaws in the process we were having there. Finally, I'll jump into the solutions we found on those problems and how we can implement them with Planful. So how is Vistra, at a corporate level, using Planful to meet its day to day CapEx planning needs? Like I mentioned, we're an integrated power provider, so we have plants and locations all over the country. It's typically at these sites that we're doing our capital projects, plan improvements in different projects there, so we need the users at those sites to be able to drive our CapEx planning process. So we have templates corresponding to different budget entities for all of these different locations. And users at those sites are able to go in, enter their plan data and feed it up through the workflow and approval process, so that the home office has their say in the numbers. Just to give some more context, we have over a thousand budget entities. Obviously, not all of those are CapEx, but it gives you an idea of the scale of how much time is saved through using a tool like Planful. Without it, it would be a very intensive process getting all that data from sites all over the country. And that is something I would not want to be involved with. With these sites, we have similar reporting needs across pretty much all of them. You want to compare your plan and actuals, for that given site, for all the different projects. So how do we do that quickly and efficiently? Planful gives us a few opportunities for time savings here. If you've been in a bunch of the other sessions, like I have, you've probably heard of a few of these. Report collections is probably the number one feature. We're able to create a set of reports, run them with the same inputs and generate an output to a distribution list of relevant personnel and have those reports delivered to them on an automated or a scheduled basis. Combining that with substitution variables, we're able to save a lot of time on the upkeep of reports. So instead of manually updating the time and scenario for 10, 15, 20 different reports, we can go to one centralized location, update the values there, and it propagates out to all the different reports. That's an example of using smart report design to save time. With smart hierarchy design, we're able to maximize the benefits from our smart report design. So, all of our plants and locations are operating in a regional capacity. So having those regions in the hierarchy, above the plants, allows us to report against not just the plants, but at a regional level. So, those managers sitting in the office back in Texas can execute a report set and get a set of reports for all of their relevant plants, a huge time savings for everyone. Outside of the initial setup, it's a minimal time investment to upkeep these reports as well, instead of countless man hours being spent individually building each of these reports at each different site, and then finding out at the end that nobody had the same methodology to doing it and it's not really comparing apples to apples. We're able to keep our finger on the pulse by looking at the relevant data in the same context. And at the same time, we're saving time and allowing more analysis to occur rather than report building, letting people focus on their jobs rather than the manual repetitive tasks. The best part about all of this is that a lot of it is managed by the business. As the IT guy, I can say that my hand in Planful is fairly minimal compared to a lot of other applications. We've designed Planful carefully with the role in security setup, so that appropriate business users can perform functions that they need to, to do their jobs. Take, for example, a what- if scenario. If someone wants to compare a plan scenario that exists against" What if they did something else?" They can go and copy that scenario quickly and easily. They're not waiting for me, they're not waiting for a DBA. They're just performing that task. It's done five minutes later and they're reporting against the differences. So, allowing the business to take control of their work and empowering them to do what they need to, to accomplish their jobs, is a big part of what we do with Planful. Just to give some more example around this. And I apologize for the blur, just trying to limit the data shared, but I can promise you that at least half of those have the word CapEx in them. This is a sample of our cloud scheduler, showing the report collections run over about a three- day period. So you can see quite a lot of use out of that reports collection feature. And each of these contain upwards of a dozen reports. So this screenshot here represents dozens of hours worth of time savings just in reporting alone. And that really benefits the business as a whole. So now, jumping into issues around IT planning. So, why is our technology group still using Excel for planning? A very good question. But first, I guess I'll touch on within IT, where does CapEx come in? So as a department, we're planning against OPEX spend, CapEx spend, cost to achieve, and corporate development. And then operating as a cost center, planning is a very crucial component of IT. Without the ability to drive extra revenue, to offset any spend increases or variations, we need to be able to respond to changing business demands. So, if our leadership team comes to us and asks for us to trim a few million dollars off the budget, how are we going to do that in an Excel world? It's a fairly tall task. So, how can we move forward with that task if our actuals are coming in, we're spending time processing and building reports. And by the time that data's through to us, we aren't in a position to act on it. So that delay really hamstrings any effort to find those cost- saving opportunities and act on them in a timely manner. So, they say profits are a dream and cash is a reality. How are we spending that cash? In the IT group, hardware and software are our fixed assets of the company, with those costs being capitalized and depreciated over the life cycle of the software. We have another major component, our costs in the transition from physical data centers to cloud hosting. That is a major CapEx effort, and those costs are considered prepaid assets of the company, so they are placed in prepaids and amortized over the period of the term of the agreement. Sorry. So, in our IT, our planning process was nowhere near as modernized as the corporate process, leveraging Excel reports out of our ERP, we were manually building reports and creating our planning deliverables. We had a lot of manual work as part of this process, including some core components. One of those being our hardware and software cost tracking. That's a huge part of IT. And to be entering those man numbers manually in an Excel sheet is far from the best way to be doing it. And it introduced data quality issues, because that was not always a reliable or accurate way to enter the data. Those data quality issues propagated to the plan, and you're not comparing good plan data versus good actuals data. More quality issues came in from our PO expense process. Employees within our organization are taking their expenses, applying improper GL coding, as they like to do, and submitting those expenses. That led to a robust process to identify those improperly categorized expenses and reclassify them. That's all times doing work that isn't really of a value to the business. It shouldn't be that way. Are these issues things that people here can relate to? I'm sure it's not just IT- centric that people are doing those bad data practices. So how did we, in technology, end up as the least modern group doing planning? It's very easy to say that we should just use automation and modern tools, but not as simple in practice. There's a classic scene in the sitcom, Malcolm in the Middle. The dad, Hal, he comes home and he finds a light bulb out. That's an easy fix. He goes to his closet, he finds the pack of replacement light bulbs, and he grabs one. But the shelf that the light bulbs are on, just a little bit wobbly, so he decides he can fix that. He goes, grabs a screwdriver, and when he closes that screwdriver drawer, it just sticks a little bit, nothing a little WD 40 can't fix. So he goes to the garage, grabs a can, and empty can. Easy fix for that. He gets his car keys and gets ready to head to the store to buy some more WD 40. Car won't start. So, cut to him with the engine out of the car, working on it. And Lois, his wife, comes home asking why hasn't he replaced the light bulb yet? And he responds, " What does it look like I'm doing?" So, as we evaluated our IT process, this is a great representation of what we found. A lot of issues were discovered along the way, those data quality issues being a major component of it. And then on the IT side, I'll say, what we care about in terms of planning differs greatly from the corporate side of things. We're interested in software, vendors and assets, none of which exist in our ERP. We also have our own project management tool, with projects that also don't exist in our ERP. So we have a lot of data that we're trying to plan against, but it's not got a route into Planful. Another major component of the IT planning is comparing those projects in the different stages of their life cycle. So a project exists as an LRP project, then it exists as a proposal, then it exists in execution, then it exists in service. So how do we compare how the project evolved through those different stages and justify why the numbers might've changed between those different phases? That way we can better understand and make more accurate LRP plans in the future, if we are armed with better data. We also book our hours in the project management tool, and that's a major component of capital recovery against those CapEx projects. People doing the work is a big part of those IT CapEx projects. So, having those hours accurately counted for is how we get our recovery. But there's another wrinkle there, of course, not every hour spent against a capital project is considered CapEx by our group. So, if you're in the planning or the requirements gathering and analysis phases of a project, we consider that operations work. If you're in the post- go- live support phase, that's operations again. So it's only during those build phases of the project that we're actually capitalizing the hours and accruing against the project. So, that necessitates that we need to accurately track our time for each phase of the project, so that the actuals can be fed with accurate recovery amounts. Presented with all these challenges, how do we move forward? These problems can't all be solved at once. So we implemented a continuous improvement plan. Our goal was to minimize manual touchpoints so that we can clean up some of those data quality issues. We wanted to further solve those issues by cleaning upstream processes so that the data coming into our system was a good foundation to plan and report from. We started leveraging some BI tools and automation around our ERP reporting into planning deliverables so that less turnaround time was necessary to report on actuals. These steps got us closer and improved our process, but how do we start leveraging Planful? The synergy of having our plan data native to Planful is a worthwhile goal, but that disconnect in dimensions makes it a very tough one to solve. So, when trying to solve the problem of nonaligned dimensions with Planful, there's a variety of options available. This is something we started in the discussion with our customer support manager, just going through those different options, weighing out the pros and cons of each, and just getting a better understanding of what our possibilities were. I would always recommend leveraging this Planful support team. There's a lot of helpful people there. In conjunction with our partners at WG Consulting, we were armed with the knowledge we needed to help us make an informed decision on how to tackle this problem. I would add that none of these is a single, correct answer to this problem. It's much more nuanced than that. So I'd recommend definitely reaching out, if you find yourself in a similar situation, and consulting either Planful or a partner in getting a better understanding of the situation. These three options were the ones we looked at. That does not necessarily mean these are the only options available. Our use case was a little narrow. Speaking of our implementation, we're currently maxed out on dimensions in our corporate tenant. So adding another dimension to track vendor or asset or software, just wasn't a possibility. So, let me run down through these different options for you. First, is the structured planning solution. So leveraging attributes and structured planning. This is a fairly simple solution. You're just going in and you're mapping one to one between what you want to plan against and an existing dimension in your hierarchy. For us, this was not exactly feasible, just based on the amount of different dimensions we were looking to add. Another approach would be to use modeling and dynamic planning. It would be the most obvious solution, just building an external source model, with a custom hierarchy, and then leveraging maps to feed data in between. And that would provide us everything we needed to plan our own way. There are some limitations on the planning side, just between key value mapping and overall model size, but none of that eliminates modeling as an option. And the benefit of it all is that at the end of the day, we could take that model plan data that we generated for the IT team and feed it straight into the corporate plan. No more manually entering the data into the planning templates. Finally, the out- of- the- box approach would be, and the presentation literally doesn't have a box, building a separate structured planning tenant. So this is one we didn't really even consider at first, until we started talking with Planful. For this one, we would be able to build a separate, structured planning with our own dimensions that we care about in IT. This would give us all the benefits of structured planning, that simplicity of use, but we'd have our own hierarchy and there'd be no impact on our corporate plan, which they do not want us touching too much. Additionally, we'd be able to have our own dynamic planning environment attached to that separate tenant. So, if we had any need in the future, we could build models related to our IT planning. Exporting data out of Planful and loading it into the corporate tenant via DLR, that's a simple process and one I could probably have up and running in a day or two. So, which did we decide to use? Of course, the building a separate structured planning tenant. Now, let me tell you the reasons why we did that. To us, this was the best option for non- finance users to get in and pick up easily and quickly. We're able to manage our own hierarchy, building DLRs to populate everything that we need to and getting that data set up quickly. Like I said earlier, we have our own dynamic planning environment accessible and ready to go. So we are starting to brainstorm different use cases that we could leverage modeling with, and that's a great upside for us. All that functionality I talked about in our corporate world, in terms of report collections, substitution variables, that all applies here, too. So we're getting the same time savings they are. And we're able to cut down on the monthly reporting cycle, delivering managers, individualized reports, relevant to their cost centers. And finally, like I said earlier, we can take that data and get it into the corporate plan, just with a quick export and feedback into a DLR. So that's sort of an IT- centric approach, but those solutions, those options, are relevant to any department out there. I'm sure a lot of people have experienced that redheaded stepchild, typically IT, but not always the case, that has to do planning their own way. Maybe they care about different dimensions for good reasons. Maybe you think they're crazy, but I'll help run down through these options a little bit more, give a little more context around them and why we did not select the other two. So, for the attribution structured planning, it would be definitely the easiest implementation, but there's that requirement of a one to one between what you're trying to plan against and a dimension in your existing structure planning. And with five dimensions, that's a very tough game to play, to try to match those up. For us, it was not feasible, but if it were, the flexibility offered by DLRs, we could reload our tree with those attributes added and we could be up and building reports by the end of the week. Modeling is always an attractive solution. We have quite a few different models built for our use, and it's a great tool. You could probably talk to every person in this room that's using modeling and they'd have a different use case for it. It's powerful and it's capable of supporting that variety of different requirements. And it definitely was a potential solution for us. It was an actual decision we had to make between modeling and a separate structured planning tenant. So, if you're on the fence, I would definitely recommend talking to someone, because there's a lot of information that they can provide. Planful itself has paid a lot of attention to the modeling world recently. So with their enhancements on the interconnectivity between structured and dynamic planning, tying those user groups together so that you're managing users from one place rather than in a disconnected state, along with their improvements around model aggregation, modeling is in a very healthy space right now. For us, though, like I mentioned, we do our consolidations there. We have several other important models working there. We just didn't want to overcrowd that environment. So, finally, we're happy with the flexibility and the scalability offered by a dedicated IT tenant. It's our own environment. We can make of it as we want. So there's limitless opportunities for us to go and build all around all the functionality Planful offers. So, how are we bringing this vision into reality? To start, we worked with Planful. Sally, in the back, helped us get our tenant set up. That was a relatively simple, easy process. From there, we worked with our implementation partners, WG. They got us set up through the initial configuration of the tenant. Once that was complete, we began setting up our hierarchy, building DLRs to load in members, plan data and actual data. That overall process was targeted for about a month after the Planful part of the project began. Like I said, this is part of a much bigger, overall project, with a lot of upstream work that needed to be done to get us to this point. The project's actually still underway, as happens to all projects, at least in my experience at big corporations. There's always those roadblocks and obstacles, resource availability, things that come up that throw off your original plan. We're dealing with those now, but we're hoping to be still wrapped up by the end of the year. And hopefully, I'll be back at Perform next year to talk about what a success this has been. So, next for us would be starting with the manual load process. We really want to de deliver a MVP, a minimally viable product, to the users. So instead of spending extra time getting the integrations working, getting everything automated, starting with the manual process to get the stakeholders into the system and realizing the benefits as soon as possible. Our stakeholders were very excited about being able to compare LRP data versus the quarterly forecast versus the actuals, all in one place, rather than having this Excel sheet, this Excel sheet, and trying to compare numbers in between them, building all the different functions you need to do that, just a hassle. So the last thing we want to do as an organization is to delay or cancel a project because we outspend our plan in other areas. We're hoping to get our team into the software and acknowledging those problems faster and having less impact on what work we try to accomplish in a year. We're hoping for about a three- month turnaround time on getting the manual process up and running. And that includes some design around the reporting and the budget templates for data input. Excuse me. Once we are confident with that design, it's been tested out, all the stakeholders are happy, everybody's contributed to the process and made whatever small tweaks they need to, then we would begin our automation. So building integrations and getting everything flowing in between our different systems. So, for us, we're going to pool our data from all our different sources into Snowflake, a data warehousing tool, and then load it into Planful via those same web service API loads we use on the corporate side. We're also hoping to develop a link between our IT tenant and our corporate tenant, so that with the push of the button, that IT data is going straight from our plan into the corporate plan. And that's something we're very excited about. So, what kind of benefits are we looking to see from doing all of this? Our stakeholders are excited to begin using the tool and having, like I said, those numbers available for quarterly versus actuals versus LRP is our finance manager's number one goal. Every time I talk to him, it's the only thing he can talk about. We also hope to deliver some time savings around key areas, mostly around the monthly reporting process. With an investment of about 30 or so hours, we're hoping to rebuild the 15 to 20 reports that they manually create on a monthly basis and get them set up in Planful, with complete end-to- end data, without having to do any extra processing. So for that, we hope to save over a hundred hours on a yearly basis, just from the time it takes to gather all the data and produce those reports. Then on the feed into the corporate plan side, we're hoping to save at least 50 hours a year, which it may seem like a lot. We're just taking the numbers and we're putting them into the corporate plan via templates, but there's a lot of time spent rolling up the numbers on the IT side, massaging them and getting them ready, then entering them over into the corporate plan. Then we figure out we need to change those numbers and we need to re- roll them up. Then we need to go and re- change them on the corporate plan. And then we do about that about three or four more times, and then we're finally happy. So each process is a little more intensive than you would hope. Finally, we're hoping to empower our IT managers with detailed analysis of their planning and actuals data, giving them a granular look at where their dollars are going, based on vendors, assets, and software, the stuff they care about, the stuff that they can quantify quickly and easily, and allowing them to make informed decisions around their spending. Instead of having a shared bucket, each manager would have their own, with several lines in it to plan against. And that's just not the case if we were to try to plan in the corporate environment. In a corporate environment, IT dollars would go into maybe three or four buckets, and then we'd have multiple managers planning against a single bucket. And let me tell you, I've seen children share candy more civilly than I've seen IT managers share money. So that's just a messy case that we don't want to dip our hands into. So, what did we cover here today? We touched on how we can leverage Planful at a corporate level, using functionalities, such as report collections, substitution variables, and other reporting benefits to save time, for our CapEx planning, for our corporate team. Then we dove into our IT planning process and the issues we identified there. Finally, we walked through a variety of different solutions, none of them perfect, but all of them applicable, into how we can use Planful to modernize our IT planning process. In closing, I hope I've shown how a company with diverse departments, with diverse requirements, can come together with Planful and build solutions that boost our efficiency and our effectiveness. So thank you all. And I think we're going to open the floor to some questions, if anyone has any.

Speaker 2: We have time for a really quick question. We're actually right on time. Oh, yep, yep.

Speaker 3: How do you manage your approval process for your CapEX? inaudible tool-

Speaker 1: Yeah, the built in workflow functionality. So each budget entity corresponds through a single template and that has an approval workflow tied to it, that someone in the corporate office is looking at that and hitting approve.

Speaker 3: Are you attaching detail inaudible, stuff like that?

Speaker 1: Not in most cases, it's more on a person- by- person basis how much they want to put into the tool.

Speaker 3: inaudible-

Speaker 1: Yeah.

Speaker 3: inaudible. Thank you.

Speaker 1: Definite.

Speaker 2: Okay guys, we're at time. Thank you so much, Michael. Please give him a round of applause-