This column suggests some guidelines that will help you get the facts about measuring cleaning performance.
Who”s Measurement Standard?
Many quality measurement systems are founded on the wisdom and experience of the inspector, who is often a former or current cleaning supervisor or manager.
While it is better than no measurement program, this approach has several cautions, not the least of which is that the inspector has substituted their personal standard for that of the customer.
So, the first guideline is that any performance measurement program — including software applications — must be founded on the customer”s definition of performance, not the inspector”s.
The cleaning operation does not exist to satisfy the inspector; it exists to satisfy the customer.
The Right Rating Scale
How many different survey scales are used for rating customer perceptions about cleaning performance?
Fortunately, we don”t need to argue our opinions, as actual research has been done on this question.
Dr. Rensis Likert conducted extensive research to identify which scales delivered the most accurate survey information.
His book on the subject is clear: Using a five-point scale or a seven-point scale delivers the most accurate information.
So, when establishing a performance measurement program — paper-based or software-based — use a five-point or seven-point scale.
The Right Metric
There are a number of inspection programs that support a five-point rating of appearance, which relies on the subjective opinion of the person doing the observation.
When two or three or a dozen people look at the same space, you can often end up with different opinions and performance ratings.
Also, subjective-based metrics lack the ability to generalize your ratings between buildings or portfolios that have different inspectors.
So, which one is the best rating system? The answer lies in the metric selected.
The best metric is an attribute-based measure of performance.
That is, measuring the description of the condition that is visible and provable.
When establishing your cleaning performance measurement system, keep the metric objective and verifiable by using an attribute-based system of measures.
Where do you find these attributes? Check my previous column on this topic.
Measure The Right Thing
Many cleaning performance measurement programs will rate the appearance of a room as a whole.
I”ve seen inspectors rate an office a “6” or “good,” apply a weight and declare a “score” for the room.
They go from room to room pronouncing rooms as good, fair or excellent.
This type of measurement process is quick and you do get a rating — by room type and overall.
However, opinions vary and so do the ratings, even if you”re using an attribute-based metric, which won”t work at a room-level inspection.
What”s the alternative? Well, use the way you clean as a guide.
You don”t clean a room; you clean the things in the room — more specifically, the surfaces and items.
And, that”s also the best strategy for measuring performance — at the item level, not the room level.
Look for a cleaning performance measurement system that is based in rating attributes on an item and surface level.
Room Sampling Calculation
One common error we see in calculating a building”s cleanliness rating is to somehow use the total square foot of each type of room as a factor.
With this strategy, if an inspector measures three offices and gets a 100 percent rating in each one — and offices are 80 percent of all space — is it accurate to say that 80 percent of the space is 100 percent clean?
Or, in the total building rating, is it accurate to say that offices now rate 80 percent instead of 100 percent in our calculation of the overall building?
No, they rated 100 percent, not 80 percent. Measurement is about accuracy, and simplicity is a virtue.
If you rate three rooms, you have accurate information about those three rooms, and anything else you do with that data is based on assumptions, approximations and guesswork.
Reporting on the specific rooms and items measured is accurate.
The rest have no statistical validity or objective reliability.
So, when designing your calculation spreadsheet, only use the information collected and avoid overreaching to tell a performance story that does not exist.
Check back next month for the conclusion of this topic.
