Not every file can be measured. For a scanned document, a locked file or a paper of which only a photograph has arrived, no count can be taken and the volume is estimated. An estimate is a legitimate method, but it is not a measurement; when the margin of deviation is not stated at the outset, it leads to a dispute on delivery.
Definition
A volume estimate is the forecasting of the size of the job in files that cannot be measured. The method rests on assumptions. The result is approximate.
When it is needed
It is needed for scanned, locked or low-quality files. An automatic count is not possible. The method becomes compulsory.
Average per page
An assumed average number of characters per page is used. This average varies with the document type. The assumption is stated.
The effect of the document type
There is a large difference between a sparsely written form and a dense contract. A single average does not fit every document. The type is taken into account.
The sampling method
An average can be worked out by actually counting a few pages. This increases the accuracy of the estimate. The method is preferred.
Choosing the sample
The sample pages are chosen to be representative. Looking at dense pages only inflates the estimate. The selection is made in a balanced way.
Communicating the margin of deviation
The margin of deviation of the estimate is communicated to the client. An estimate presented as if it were an exact figure leads to a dispute later. Transparency is preserved.
Moving to an exact measurement
Once the text has been extracted, an exact count is taken. The estimate is updated with this figure. The process is explained at the outset.
Managing the update
If the exact count differs markedly from the estimate, the client is informed. A silent update damages trust. Notice is given.
Defining the deviation limit
The rate of deviation up to which the quote remains valid can be put in writing. This protects both sides. The limit is defined in the contract.
Estimating visual content
The text in images is estimated separately. These areas do not show up automatically. An additional item arises.
Handwritten areas
Handwritten sections cannot be counted and are also processed more slowly. This effort is assessed separately. The estimate is made accordingly.
Density of tables and forms
In tables the characters are few but the effort is high. The estimate is not made on characters alone. An additional assessment is included.
The effort of text extraction
The time for text extraction is added to the estimate as well. This is a separate item of work. It is taken into account.
The alternative for the client
If there is an editable source file, no estimate is needed. For this reason the file is asked for insistently. The gain is explained.
Recording the estimate
The assumptions on which the estimate was made are recorded. The record makes later comparison possible. Data accumulates.
Use of past data
The actual counts from similar documents improve the estimate. This is why the archive is valuable. The data is used.
Putting the estimate in writing
The estimate is given in writing, not orally, and is recorded together with its assumptions. An estimate given orally is later remembered as if it were a firm commitment. A written record protects both sides.
What it is confused with
A volume estimate is frequently presented like a measurement and perceived as if it were an exact figure. Yet it rests on assumption; when the assumption on which it was made and the margin of deviation are not stated, a gap in expectations emerges on delivery.