What is digital when we get right down to it?
Automation holds two possibilities when it comes to information and technology – it can get things right at scale, or it can get things wrong at scale.
One of the things that people are naturally very good at, is sensing context errors.
We know that there’s a problem if we’re looking at John Smith’s file and we’ve got records in it for Jill Jones and Fred Bloggs.
Automation though, isn’t necessarily sensitive to this.
If you’re creating a client portal to expose all of a client’s records to them based on file content, the system is going to very happily automate the process of creating a privacy breach by showing Jill Jones and Fred Bloggs records to John Smith.
If you’re using object metadata to serve up information to people making decisions, and you get it wrong – automated decision support will also very happily give the wrong information to as many people as it can.
The two examples above are related to the accuracy of file content, but that’s a very small sliver of something much larger.
If we start to define quality as the accessibility of information from within a record, we have a whole new spectrum, and can serve a whole new set of use cases.
Forms area a meaningful example of the quality spectrum. If we’re accepting the information in any unstructured format at all (ie. email), reliable automation is next to impossible – you must have a human in the loop or make a heavy investment in technology that can understand and process natural language.
If you move to an electronic form, the quality of the information presented and the automation possibility rises dramatically. Digital forms are higher quality again because we remove OCR errors.
Before our records become fit for digital, we’re going to have to work out how to dramatically raise the level of quality so that the right information is reliably accessible for automation technologies. If you want a seat at the digital table, you have to have a view on quality, and how you can raise it in your organisation because automation can add scale really quickly – so a small privacy or confidentiality problem can get much larger, very quickly.