The feedback rate is used as if it were a positive indicator: the less feedback, the better the job. This reading turns what the indicator measures upside down.
Silence has two meanings
The text may have met the expectation, or it may not have been read at all. From the outside the two situations produce the same number, and no interpretation can be made without telling them apart.
The busiest period looks like the best period
When the other side is busy it cannot find the time to examine the texts, and the feedback decreases. While the rate is improving, no information at all has been produced about the real quality.
The moment of reading may be delayed
A text is read at the moment it is to be used, months later, and the finding emerges then. Periodic measurement has to take this delay into account.
Making feedback easier raises the rate
If the channel is unclear and the contact person is not defined, the feedback is lost inside the organisation. A low rate may be an indicator not of quality but of the channel.
When the question is asked narrowly, an answer arrives
The question "Are you satisfied" is brushed aside politely. The question "Is there a rendering in the terminology list that you would like to change" produces a concrete answer and makes the rate genuinely measurable.
The type of feedback is separated
When error feedback and preference feedback are added up in the same number, a rise in the rate is taken to be bad news. Yet an increase in preference feedback is more often than not a good sign.
It is expected to be high in the first jobs
Because the preferences are not yet in writing, there is a great deal of feedback in the first period. A low rate in this period may show that the learning has not started at all.
Counter-view: asking for feedback tires the client
It is true that the party receiving the service should not be expected to do additional work. But what is asked for is not a detailed report, it is a single narrow question; and the return on that question comes back as fewer corrections in later jobs.
Usage information is added
If it is known whether the delivered text was really used or not, the meaning of the silence becomes clear.
The rate changes from person to person
Some contact persons give detailed feedback, and some give none at all. When the person changes the rate changes too, and this may be taken for a change in quality.
Positive feedback is counted too
In a system in which only the corrections are recorded, the choices that were liked remain off the record and may be changed in the next job without anyone noticing.
If the findings are not entered in the guide, the rate is meaningless
If the feedback that arrives is not recorded, the same finding repeats in every period and the rate marks time.
The trend matters
A fall in the rate over time may show that the preferences have settled into the guide. This fall carries a different meaning from a fall in the first months.
It is read together with the other indicators
A feedback rate that is not assessed together with the number of correction rounds and the density of findings is only a measure of communication.
Takeaway
The feedback rate measures not quality but communication. Asking for the reason for the silence, defining the channel and the contact person, separating error feedback from preference feedback and entering the findings in the guide: unless these four are done, a low rate means a text that has not been read just as much as it means a good job.