May 11, 2022 | 8 min read Seven steps for building a maintenance analytics program by starting small By: Marc Cousineau Back to blog “Start small.” These two words are supposed to make big, scary projects vanish into thin air. Need to set up condition-based maintenance? Or create a new work request system? It’s a piece of cake. Just start small. But, uh, what does that mean? Vague advice is useless advice. That’s why we’re exploring all the tiny details of starting small. And there’s no better case study than building a maintenance analytics program from scratch. It’s a daunting project. But managing maintenance with hunches just won’t cut it. This article is your owner’s manual for creating a maintenance analytics program in small doses and growing it piece by piece. Table of contents What is a maintenance analytics program? How to start small with any maintenance project Seven steps for building a maintenance analytics program by starting small Start small and win big with a maintenance analytics program What is a maintenance analytics program? A maintenance analytics program is a set of processes and standards for using data to guide the actions of the maintenance team and measure the impact of those actions. A maintenance analytics program is made up of four elements: These four elements are combined into three critical steps: See how to use maintenance data to increase uptime and reduce waste What does a maintenance analytics program look like in action? A maintenance analytics program doesn’t need to be complex. You just need to regularly collect and analyze data, and make decisions with it. Let’s look at what that means for planning maintenance on a piece of equipment. You can use the three phases of an analytics program and three central questions to create a maintenance schedule. Collect data Analyze data Make decisions based on the data What’s happened before? Find unplanned downtime metrics (ie. how much, why, how often) Spot trends in the downtime metrics (ie. Bearing failure was the most common cause of failure) Schedule bearing inspections more frequently than regular maintenance on the machine What’s currently happening? Track production data (ie. throughput, production time) Flag above-average usage on the machine (ie. throughput is going to be 10% higher than usual) Schedule an inspection on failure-prone components to balance production demand with risk What should happen next? Measure team workload (ie. costs, size of backlog) Watch for indicators of overwork or inefficiency (ie. a growing backlog, higher contractor costs Hire an extra team member to keep up with backlog and instead of paying contractors Build your perfect maintenance schedule with this free template How to start small with any maintenance project What does it mean to start small? Starting small means implementing an idea or project in small steps instead of all at once. Here are some examples of starting small in maintenance: Project #1 – Overhauling your maintenance work request process. Start small by changing one element at a time, like setting up work request software, but keeping the rest of the process the same. Project #2 – Moving to just-in-time parts purchasing: Start small by testing a new purchasing process for one part. As you get more confident forecasting lead times, you can expand to different parts. Project #3 – Improving team communication. Start small by planning a shift change meeting three times a week to develop routines you can build on slowly. Three benefits of starting small? It can be tempting to go all-in on a project and get big results as fast as possible. Unfortunately, that rarely happens. In fact, 70% of all projects fail. Starting small eliminates a lot of the problems that plague larger projects, which helps you: Catch problems while they’re small and reduce risk: If something goes wrong or isn’t working, the consequences are contained. You can also identify and plan for risks when expanding the project Get adoption and buy-in faster: Change is always hard. But it’s easier when it’s smaller. This gets people on board quicker, which gets you results faster. Quick wins fuel even more buy-in. Implement projects with fewer delays: When things go wrong on a large project, they take longer to solve and the project stalls. Starting small allows you to bypass a lot of red tape, make decisions faster, and get more done. Eight ways to start small with your next maintenance project One of the biggest choices you’ll make when starting small is where to begin. Here are a few areas of your operation to test out a new idea: How to go from small to big There are three questions to answer if you’ve been running your project for a while: Is it working? Create clear benchmarks and ways to track progress. Search for qualitative feedback as well. Talk to as many people as possible who are related to the project. Try to understand what’s going right, what’s going wrong, and if those things can be replicated or solved. Is it ready to grow? First, make sure you’ve tested your idea for long enough (at least three to six months) to get accurate results. You should also be confident that any major problems and risks have been identified and addressed. Lastly, figure out if a larger group of people will embrace the change. If it is, how? There are two ways to grow a small project—expand slowly or implement it broadly. The direction you choose depends on the team’s appetite for change. If widespread buy-in is unlikely, expand the project one shift, line, or facility at a time. If you decide to grow quickly, recruit major stakeholders from your pilot project to identify risks, train others, and be project champions. Check out this guide to getting buy-in and adoption for a big maintenance project Seven steps for building a maintenance analytics program by starting small Step #1: Ask the right questions Collecting data without a purpose leads you nowhere. Finding that purpose starts with asking the right questions. These three questions can help you build a foundation for an analytics program: What are your business’s biggest goals and how can maintenance impact them? What challenges is your company facing and how can maintenance solve them? What opportunities can your organization pursue and how can maintenance help? These are big questions. But they’ll help you get a clear view of where your team can make the biggest impact with the fewest obstacles. Here’s an example of how to answer these questions: What are your business’s biggest goals? Ex. Reduce waste across all business units How can maintenance impact those goals? Ex. Increase preventive maintenance to reduce downtime and maintenance costs What challenges is your company facing? Ex. Inflation is driving up the price of materials, which is shrinking profit margins How can maintenance solve those challenges? Ex. Improve asset performance to increase production capacity and quality What opportunities can your organization pursue? Ex. Demand is exceeding capacity, so the company is looking to expand How can maintenance help? Ex. Create, optimize, and standardize processes that can accelerate expansion Step #2: Connect data to your goals, challenges, and opportunities It’s time to figure out how to measure the impact of maintenance on your business. Take the answers from the right-hand column above. List the different ways to quantify maintenance for each. Don’t limit yourself to maintenance metrics you currently track. If you have a lot of questions and answers, pick two or three with the strongest connection to maintenance. Here’s an example of what you’ll end up with: What are your business’s biggest goals? Ex. Reduce waste across all its operations How can maintenance impact those goals? Ex. Increase preventive maintenance to reduce downtime and maintenance costs What data can help you track progress? Unplanned downtime (YoY) Mean time to repair Wrench time Labor costs Parts costs What challenges is your company facing? Ex. Inflation is driving up the price of materials, which is shrinking profit margins How can maintenance solve those challenges? Ex. Improve asset performance to increase production capacity and quality What data can help you track progress? Clean startups after maintenance Found failure rate on PMs Recurring incidents (YoY) Scheduled maintenance backlog (hrs) PM compliance What opportunities can your organization pursue? Ex. Demand is exceeding capacity, so the company is looking to expand How can maintenance help pursue those opportunities? Ex. Create, optimize, and standardize processes that can accelerate expansion What data can help you track progress? H&S work orders completed on time Completion times (ie. sign out parts) Process coverage (ie. all processes are documented) PM compliance Planned maintenance percentage Get this goal-setting template and align your maintenance strategy with company goals Step #3: Choose your north star A north star is the one activity or metric you anchor all your efforts on. If your project is growing too large and unwieldy, your north star will bring focus and clarity back to your efforts. The first north star to pick is a goal, challenge, or opportunity. This is what you’ll build a maintenance analytics program around. For example, it might be improving asset performance to increase production capacity and quality. The next step is to narrow your list of data points. To figure out which metrics to begin with, use the matrix below. It grades data on usage (how much it can guide your decision-making) and accessibility (how easy it is to collect and analyze): Here’s what the matrix looks like using the challenge and corresponding metrics from step #2: In this situation, you can build your maintenance analytics program using two metrics—recurring incidents and clean startups after maintenance: Step #4: Find your starting point Remember, starting small means choosing one line, shift, site, or task for tracking data. Here’s a template for picking a starting point, using clean startups after maintenance as the metric: Questions to ask Outcome Starting point Where’s the greatest opportunity? Assets with the highest rates of post-maintenance failure Site #1 Shift #1 Assets A, B, and C Where can I get the most data? Assets with the most maintenance Site #2 Shift #1 Assets A, C, and E What will make the biggest impact? Assets with the highest target throughput Site #1 Shift #2 Assets A, B, E Where is this strategy the easiest to implement? Sites and shifts where this data is already being collected Site #1 All shifts Assets A, C, E Using this chart, we can see the most opportunity and impact (with the least effort) can be had on site #1 and assets A and E. Step #5: Assemble your people, processes, and systems Answer the following questions about your people, processes, and systems to make sure you have a strong foundation for your maintenance analytics program: People Processes Systems Who is accountable for this metric? Who is responsible for collecting, analyzing, and making decisions with data? Who is consulted and informed about the data and results? How often is the data collected, analyzed, shared, and used? What benchmarks and goals are associated with the data? What unit of measurement is being used? Where is the data stored? What happens if data improves or regresses? What systems are needed to collect, store, and access data? How is data being collected and organized in the system? What tools are required to analyze, share, and act on data? Step #6: Plan, do, check, act (PDCA) The PDCA model helps you make use of your data and fill the gaps in your maintenance analytics program. Here’s how it works: Plan Put all the details about what data you’re collecting, how you’re collecting it, and what you’re doing with it into a standard operating procedure. Everyone can refer to this document when learning the new process. Do Follow through with the plan. Consistency is key. Collect data at the same frequency, in the same way, for as long as it takes to get reliable results. Check Analyze the data and find out if your plan is working. On the data side, make observations and find insights. For example, the observation might be that clean startups increased 25% last month. The insights could be that the specs were recently updated on those machines so repairs were more in line with the operating contexts. On the planning side, audit your processes, people, and systems. If you see roadblocks, like missing or inaccurate data, find the root cause. Act Use the insights from the previous step to make decisions. If spec changes led to more clean startups, audit and update specs on all assets. The same goes for your processes, people, and systems. If a technician is lacking information, for example, consider giving them extra training. Step #7: Scale Expanding your maintenance analytics program can be straightforward from an operational perspective—take the formula that worked well and apply it to more shifts, sites, or assets. But that’s not all you have to consider. A larger program requires more of everything, like: Training Standard operating procedures System functionality Buy-in Reporting A key piece of scaling your analytics program is convincing people to give you more time and money for all these things. This article goes deep into using data and storytelling to highlight the value of your project, but here are some major takeaways: Start small and win big with a maintenance analytics program Analyzing data can be like stepping into a giant maze. It’s disorienting, especially when you’re staring at a spreadsheet full of numbers. Starting small shrinks that maze into a manageable size. It does this by focusing your efforts on one part of your operation. You can find trends in your data much faster, take action immediately, and build on that success with greater support. Once you’ve mastered the art of starting small, you can replicate this formula for all projects, eliminating red tape and second-guessing along the way. When you get small wins on a big scale, your maintenance team can bring massive value to the company. (opens in new tab) (opens in new tab)