now that we have our first product to its first release, its worthwhile to spend some time describing what's behind this process, and why we develop the set of products that we have, are, and will develop. This will probably cover a few posts so this first one is really the "setup"
there are a couple of principles that we've (perhaps largely unconsciously) been following.
Firstly, it's about trying to bring some better order and structure to unstructured and semi-structured data that is found in most government, corporate and community organisations, and is now starting to find its way into domestic situations as well.
I have worked in many environments over the years and never cease to be amazed at the level of disconnected-ness that exists between the various items of information that are used in most knowledge-intensive or information-intensive situations. The development and proliferation of "office" products like spreadsheets, word processing tools and presentation platforms like powerpoint together with the internet has meant that there are more and more ways of creating and distributing more and more information to more and more people.
And all these "containers" of information are largely uncontrolled in either design or in the data with which they are populated. This becomes much more extreme when we add to the mix a growing group of temporary workers (contractors, consultants and the like) who work across multiple organisations.
Given that email and other messaging is such a large part of many knowledge-workers interactions with other people, these "containers" of information get distributed around and we process them in a continuous stream of information "transactions": receiving a document or spreadsheet or even an extract of one of these in an email, doing something with it, and then issuing our own communications out to others as a result (including responding to the originator).
But I've noticed that a larger proportion of time is involved in one or more micro-steps in each of these "information transactions" that occur in our working lives: these micro-steps are: conversion and validation.
"Conversion" is getting the data in the originating container into a format that we can use for whatever our function is at a particular time. Sometimes its as simple as just digesting an email, but often it goes further. Perhaps we have a table in a Powerpoint pack that originated in an spreadsheet in some other person's computer. Formatting is different (wrapping font etc) and perhaps there have been some extra rows/columns thrown in for added meaning (subtotals or even just category grouping or spacing).
So at some level, whether its' just unconsciously cognitively decoding and absorbing this, or even physically converting it to another format, it seems to be taking more time and energy on average to access the information in the container.
"Validation" is making sure that the information is actually what you want, and is accurate for the purposes that you wish to perform with the information. So now I'm thinking that there are two kinds of validation that we do: content validation (say for accuracy) and validation of "completeness".
there are plenty of other "micro-steps" too: analysis, manipulation (e.g. merging), summarisation etc but I'll come to those later. between the outside world and these other steps are the two I've just mentioned above.
An example: Say you receive a table copied and pasted into the body of an email, say its a set of projects with some statuses, or territories with sales figures or whatever. If the data in the table is the full set of a known and small population of items (i.e. it's all the projects or all the territories) then there's little completeness validation - you've either got all your sales districts or not. But if the data in the table represents a small and / or varying subset or a larger population (say a small set of projects out of a large portfolio) then there is the issue of completeness. Usually this subset has been extracted from the larger set and given that the contents of the container are largely uncontrolled, how do you know that the set of data you are looking at is actually the complete and current set that you need to have in order to perform your function?
do we have another list that we compare the first list too? Mentally in many cases we do and we either assume that all is ok, or we rely on our mental comparison of one list against another, or we get out a pen and paper and check off one against another.
whatever we do, it seems to me that the amount of effort involved in these two micro-steps in my end-to-end treatment of each information transaction is getting larger, the probability of error is going up, and the downside of making a mistake grows.
So all this is a long-winded way of describing a scenario that we want to try and solve: how do we put simple power in the hands of individual knowledge workers to address the growing burden of conversion and validation. So what we want to do in the example above is make it really easy for someone to get the information out of whatever "container" they receive, convert it to a format that is manageable, and do the validation that they need to do as quickly as they can.
the problem with not handling these micro-steps well are all too common and all too manifest in many work environments today: assumptions, errors, missed communications and sometimes inadvertent disclosures. And of course sometimes information that gets completely lost.
so providing tools to assist with these microsteps (and others) is what we set out to do in the long product development journey that is starting to bear fruit, initially with OS4XL.
Have we succeeded in this with OS4XL - in part yes, but there's still some way to go (watch this space for new enhancements soon). And, there's all these other containers as well (watch this space for new product announcements along those lines as well).
Next post will be about what we do with the data once we've unpacked it and validated it.
cheers.
as always, welcome your feedback.
