I think we firstly need to define “critical success factor of health and safety performance benchmarking”. I understand  what the individual words means but not the phrase as a whole.   

Everyone working their stats out the same so its actually meaningful.

Using the same formula and including all the same data. Unless its a level playing field another companies data is pretty much useless to benchmark against.

Ali

Thunderchild has pointed to the problem.

It is very difficult to compare like with like EVEN when different organisations are playing by roughly the same rules.

You cannot reasonably compare the accident performance of a business that makes bread in vast quantities with an artisan baker.

One has almost everything done by automated machinery where the workers press buttons remotely, the other has the workers at the sharp end.

So, you need to find organisations of comparable size and activity to benchmark against and that is not nearly as easy as it sounds.

Suppose you have two major construction contractors with similar number of staff and turnover, both doing new buildings.

One might subcontract just about all the more dangerous tasks and in effect operates as a project management company. One might want to do those tasks in house partly as they want a better handle on what is happening.

All else being equal, the second of these contractors SHOULD have higher incidence numbers and rates. Higher does not equal Worse!

Ok

When I was involved in bench marking ( involving half a dozen government agencies involved in scientific work) we first all agreed a set of performance indicators which were not necessarily the ones that we used internally.

We also made sure that the numbers took into account the differences in agency size. So, we tended to use figures that were per FTE(Full Time Equivalent)  employees.

We also had to take into account that  some agencies had a lot of  interaction with the public while others did most of their activities behind 20 foot electrified fences.

Establishing leading indicators was difficult as, for example,some agencies wanted to measure how many risk assessments they had done/reviewed  while others regarded that as an opportunity to create more activity for the sake of the stats without contributing to overall improved H&S.

Finally, the benchmarking was not an end in itself: we did  not  create a league table where we could say that agency  A was much better than Agency B because of a higher score. Instead, we used this information as a starting point  for a discussion as to why some agencies had “better” scores  and how we could share each others best practice.

Originally Posted by: A Kurdziel

Ok

When I was involved in bench marking ( involving half a dozen government agencies involved in scientific work) we first all agreed a set of performance indicators which were not necessarily the ones that we used internally.

We also made sure that the numbers took into account the differences in agency size. So, we tended to use figures that were per FTE(Full Time Equivalent)  employees.

We also had to take into account that  some agencies had a lot of  interaction with the public while others did most of their activities behind 20 foot electrified fences.

Establishing leading indicators was difficult as, for example,some agencies wanted to measure how many risk assessments they had done/reviewed  while others regarded that as an opportunity to create more activity for the sake of the stats without contributing to overall improved H&S.

Finally, the benchmarking was not an end in itself: we did  not  create a league table where we could say that agency  A was much better than Agency B because of a higher score. Instead, we used this information as a starting point  for a discussion as to why some agencies had “better” scores  and how we could share each others best practice.

is it possible to share the benchmarking report you prepare?

In a previous company we used to benchmark against others in a similar industry (20 of us). The problem was we worked out from the anonymised info that we were below average. We asked the organisation that collated the info if they would ask the one with the best stats, if we could meet, to ask if they would tell us what they were doing we were not.

The answer turned out to be nothing, we were in fact doing more than them, the reason they had better stats, was they didn’t pay sick pay and they had killed someone a few years before and that was still in everyone’s mind.

So benchmarking, in itself does not give you anything that useful. You are either going to be better or worse than the average. If you are better, are you just going to pat yourself on the back, and if you are worse how will you find out what the relevant issues are. 

Chris