In the previous article on measuring the quality of your estimates we looked at how can make use of the different versions of the bid to win ratio to gather useful information. However there is an inherent problem with it. The bid to win ratio is only proxy measure of quality of estimates as a winning bid is not necessarily an accurate one.
Important note
When I talk about both estimated and actual costs in this article I am assuming of course that you include your profit margin in both of them.
Difference in estimated and actual cost
To overcome this one needs to look at the actual costs incurred and compare them to the estimated costs. The simplest way would be to look at the difference between the actual costs and the estimated costs. Anything different from zero and your estimating wasn’t fully accurate.
Ratio of estimated and actual cost
However simply calculating the difference will not tell you anything, especially when you want to aggregate different jobs and arrive at a single number of your overall quality of estimating. You need to calculate a ratio so that the size of the job is taken out of the equation. To do that, you divide the estimated cost by the actual cost and multiply the figure by 100 to express it in terms of percentage.
Aggregating ratios for different jobs
A common mistake while aggregating ratio similar to the one given above is that people tend to simply average out the different ratios of all the jobs. They simply add all the ratio (in percentage terms) and divide the sum by the number of ratios (number of jobs). This is completely wrong!
To aggregate the ratios and arrive at a single measure of quality one needs to look at differences from 100 in absolute terms. This is because a ratio that is over 100% is also just as inaccurate as a ratio below 100%. So the first step is to calculate the difference from 100 for each of the ratios. Then you should treat all the differences as positive numbers, so that if you add them up negative errors and positive errors do not cancel each other out. Divide the sum of all these numbers by the number of terms you’ve added (the number of jobs). The resulting figure gives you in percentage terms the amount of error in your estimating. You can subtract the number from 100 to give in percentage terms how accurate your estimates are.
Different types of the ratio
The above ratio can be used in many ways by combining only specific types of jobs as well as looking at a specific area of all the jobs.
Calculating ratio for a specific subcategory of all estimates
Quality of estimating material quantities
By only looking at the ratio of estimated material quantities to the actual material quantities used in the job you can find out specifically the quality of your estimates in estimating material quantities.
Quality of estimating labor hours etc.
Similarly you can look at the ratio of estimated labor hours to the actual labor hours used. I don’t think I need to mention anymore examples. You can look at the quality of whatever specific part of your estimating you want.
Calculating ratio for specific kinds of estimates
You can look at the article on how to measure the quality of your estimates – bid to win ratio for a detailed description on how you can categorize your jobs and then calculate separately for them. You can look at the quality of your estimates for big jobs only and compare them to that of the quality of small jobs. You can look at which clients you estimate better for or which type of work (eg: low voltage) are you better at estimating and so on.