Developing a cost database for accurate cost estimates – 5 core elements
As a cost estimator you most probably use historical data from previous projects to create accurate cost estimates. It would be especially handy to have a structured, ready-to-use cost database where you can access all that information quickly and easily.
At Cost Engineering we believe having a structured database is necessary, as it will save time, reduce the number of errors, and maximize the quality of your estimates. Our experts have, therefore, created a 5-step approach to developing such a database as demonstrated below.
Step 1 – Classify the cost data by discipline or activity
As a cost estimator, you may already have several historical projects with useful data on your drive, which you keep browsing in search of the right information for your current estimate. The first step to creating an efficient cost database is to gather all those reference projects and classify them.
You would need to know from which disciplines or activities the data comes. Do you have some installation norms that you wish to classify between piping or equipment, or whether the cables are instrument or electrical?
Step 2 – Classify the cost data by type of costs
After classifying the data by discipline and/or activity, you will need to sort that data by the types of costs. Are those “material”, “supply” or “labor” costs? Do you have the rates per hour?
Knowing the type of costs is key to making an accurate analysis of the costs in the upcoming steps of data development.
Step 3 – Consolidate the cost data
The data you receive will most likely come from various projects with different attributes: currencies, years, units of measure, etc. Since the goal of the database is to have a reliable set of data to use in future estimates, it is important to consolidate the data in a way that makes it possible to compare and analyze.
You must, therefore, index similar cost components to a single year, and align any additional information for each data point based on what is included or excluded in the costs.
Step 4 – Analyze data to build structures or dynamic components
Once the data sets are classified and consolidated, your next step is to analyze them. The goal is to find trends and cost drivers that will lead to standardization of the data. Based on the quality and the type of data it will be possible to say what estimates these data sets can support (Detail estimates, high-level estimates, etc.)
During the analysis, you will generate additional cost points based on factors/judgement, and based on inclusions/exclusions.
At the end of the analysis phase, you will create a standard set of data (static or dynamic) with clear cost attributes and specifications (inclusions and exclusions of the costs).
Step 5 – Group the cost data in the format of the cost management software
Lastly, you should prepare the data for future use. The data needs to be in a format that makes it easily accessible for estimators: assembled in a recognizable matter, for example, via a cost structure.
Additionally all attributes of each component must be linked to it both technical (Weight, Size etc) and Commercial (Incoterms, Cost, Reference date) apart from generic properties such as unit of measure, or description.
It is also important to consider how the data would be reported. Create all the required reporting structures (breakdown structures) and assign each cost item to the correct key.
It is necessary to see in what format an estimator is expecting the data. For example, if the goal of an estimator is to create a detailed estimating, maybe having unit rates with just cost per unit of measurement is sufficient. On the other hand, for a higher level estimation, it is logical to have assemblies that group the costs (for example, material and labor costs together). All of these should be clear in the structure the data is stored in.
After all the processes are followed, the cost dataset is ready to be migrated into a digital software solution.
Do you have any questions on how to develop an efficient cost database, or do you wish to learn more about CESK Data? Please feel free to contact us.