FM benchmarking is a valuable tool for measuring and improving the performance of your facilities and services. But what if you don't have enough data to compare your results with industry standards or best practices? How do you handle data gaps in FM benchmarking?
Understanding why data gaps occur is the first step in addressing them. Data gaps can be caused by a lack of data collection or reporting systems, inconsistent or inaccurate data entry or validation, different definitions or metrics used by different sources, missing or outdated data from external benchmarks or peers, and changes in scope, scale, or quality of services or facilities. Identifying the source of data gaps can help you prioritize and address the most critical issues and avoid inaccurate analysis.
If we have gaps in benchmarking we would benchmark against like minded organsation, or use historical information as a guide.
The paramount of successful and efficient asset or facility management process is to have strong data set for future diagnostics and analysis. The platforms like Building automation System and FM platforms like Maximo or IVIVA host very useful data like when the maintenance is carryout, how frequent, details of spare parts and BMS will provide valuable information about performance of each asset. Initial setup and ongoing use of these platforms are very important to retrieve meaningful diagnostics in the future. This approach could eliminate garbage-in garbage-out scenario.
While data compilation& reporting contributes greatly in the overall objective,it is to be noted that data has to be effective & accurate and must capture the riders and dependencies clearly. Normally, when reporting is done across different geographies a lot of local factors get into the game. Uniform & flexible templates with clear methodology of capturing &calculating efficiencies must be established to draw out trends, shortcomings & areas of improvements. Another, typical mistake committed while benchmarking FM data is comparison of companies which differ in span, scope & nature of business (the classic mistake comparing apples with oranges),though, best practices can be derived based on its suitability, resource and budget required.
It's very important in Data collection that one must ensure that all Data is collected completely in accordance to the Benchmark. Where there is a gap in the Data collection the Benchmarking cannot be accurate . However one must also make sure that the computing of the Data is done properly to enable you get the exact results.
To fully understand data gaps, the first action is to ensure all stakeholders understand the desired outcome, review the current data collection, validate the metrics/benchmarks being used, then ensure these all align to any revised scope and quality of service. All can be achieved through open collaboration.
The second step is to fill the data gaps with estimation methods that are appropriate for your situation and objectives. These methods can be simple or complex, depending on the type and amount of data available, the level of accuracy required, and the resources and time available. Examples include extrapolation, which uses existing data to project future or missing data based on trends or patterns; interpolation, which uses existing data to estimate data between two known points; imputation, which uses statistical techniques to replace missing data with plausible values based on other variables or assumptions; proxy, which uses data from a similar or related source to substitute for missing data; and expert judgment, which uses the knowledge and experience of subject matter experts to estimate data based on their opinions or expectations. It is important to document and justify your choices and assumptions when using these estimation methods as each one has its advantages and limitations.
We have to recognise that data gaps (predominantely asset related) will always exist in any commercial outsourced arrangement - the key is how they are managed. An appropriate, risk based, model is key as well as acceptance on both customer and supplier that this is the case but it will be priced accordingly. But always remember if you outsource a mess then expect a mess back + a margin whilst it is tidied up. Post contract, ensuring a focus by supplier teams on identifying all compliance related assets (rather than every single one) will significantly reduce the risk for both parties without a significant obligation to identify every single radiator out there.
Sadly I spent most of my 42 years in the maintenance industry putting right other peoples mess. Usually caused by price undercutting by people doing just this. Ie estimating (usually poorly) the cost. Then complaining when the client dumps them because the job wasn’t done right. Or the maintenance company pulls out because there’s not enough money in the job. DONT GUESS OR ESTIMATE IF YOU DONT HAVE GOOD WORKING EXPERIENCE OF THE CONSEQUENCES.
Attempt to gather missing data through various sources. This might involve reaching out to facility managers, collecting historical records, or utilizing third-party data providers.
Documentations is key in data gathering and this must be done consistently. It requires adequate training for persons saddled with such task to work on comparitive analysis of events over time .
The third step is to validate and verify your data and estimates to ensure they are reliable and consistent. Validation and verification are processes that check the quality and accuracy of your data and estimates against predefined criteria or standards. To do this, you can use methods such as data cleaning, data reconciliation, data auditing, sensitivity analysis, and peer review. These techniques can help you identify and address potential errors, biases, or uncertainties in your data and estimates. Additionally, they can ensure that your results are accurate and reliable.
certain data points are impossible to obtain, consider using estimation methods. Statistical techniques like regression analysis or interpolation can help fill in gaps with reasonable approximations based on existing data.
Measurement and Verification is key process. For that a strong level instrument shall be available and investment to this is vital. For example understanding if the motor is over heating is important. Adding a IOT based sensor in the motor chassis could be very useful. The next process is comparing the data from the field with common sense and the equipment datasheets. This will help to set a bench marking and set the algorithm to decide if the equipment is doing the right thing or not. End of the opinion from subject matter expert mean a lot get the help from mechanical technicians, electricians, plumbers etc. who can provide very valuable insight on the data from their years of experience and these all information can built a baseline.
In my experience this is often not done. Monitoring systems are rarely validated, poorly configured and often simply not working. If you are connecting to any existing system proceed with caution. Assume no data is accurate until you have done your own verification and validation process.
The final step is to communicate your data and estimates clearly and transparently to your intended audience. Communication is essential for sharing and explaining your findings, insights, and recommendations from your FM benchmarking. When doing so, it is important to use appropriate formats and visuals to present your data and estimates, such as tables, charts, graphs, dashboards, or reports. Additionally, use clear and consistent language and terminology to describe your data and estimates, as well as provide context and background information for the audience. Furthermore, highlight the key messages and implications of your data and estimates, such as trends, comparisons, gaps, or opportunities. Lastly, disclose any uncertainties or assumptions in your data and estimates to build trust with the audience. Communication can support decision-making and improvement actions.
If direct data collection is not possible, consider substituting missing data with proxy data. For instance, if energy consumption data is missing for a certain period, you might use temperature data as a proxy since energy use often correlates with temperature.
Understanding why data gaps occur is the first step in addressing them. Data gaps can be caused by a lack of data collection or reporting systems, inconsistent or inaccurate data entry or validation, different definitions or metrics used by different sources, missing or outdated data from external benchmarks or peers, and changes in scope, scale, or quality of services or facilities. Identifying the source of data gaps can help you prioritize and address the most critical issues and avoid inaccurate analysis.
In my experience, one of the reasons for missing data is due to the system that collects the data. ?If there is even the slightest inconvenience to the people responsible for collecting and storing data, it is often neglected. ?In addition, if people do not understand the importance of collecting data, then there is not enough passion.
One thing that can contribute to data gaps is the software. Be very articulate with what you want the program to do. Will each piece of equipment get its own asset tag? How far back can the team track the information? Will the software be able to grow with the company? The right software will prove worth its price while the wrong software will not help you collect important data.
No two organisations are the same so they rarely produce/provide all the data that you need for total like for like comparison purposes - so you must first identify what like for like data you have and then try to fill in the gaps - asking the executive team and comparing the answers with those you get from people at the workplace might get you on the right track - but remember that, to be effective, bench marking must be organised as a collaborative exercise so you might be teaching as much as you are learning - and going forward two brains, perspectives and experiences will always be better than one.
Always remember, while addressing data gaps is important, the ultimate goal is to achieve meaningful insights for better decision-making in Facilities Management. In essence, handling data gaps empowers Facilities Managers to make well-informed decisions, optimize operations, and drive continuous improvement in their facilities. #fm #facilitiesmanagement #facilitymanager #fmpulse
Back filling and extrapolations are not always accurate. It is key to have past data for future scenario modelling but manipulating the data can topple down the model or the assumptions. So always invest in backup or redundant data management system. Also it is very important to get the people on the ground to use the FM planforms to fetch the data which doesn't happen every time.
Perform sensitivity analysis to understand the impact of data gaps on benchmarking results. This will help you assess how sensitive your conclusions are to missing data and whether they significantly affect your overall findings.