A few days back, an email was sent to our ‘allstaff’ mailing list inviting us to join in a bocce tournament. This took me a bit of time to digest, not least because I felt impelled to look up what ‘bocce’ means (it’s an Italian variant of pétanque, if you are interested). I guess this took a couple of minutes of my time in total. And then I realized I was probably not alone in this – that over a thousand people had also been reading it and, perhaps, wondering the same thing. So I started thinking about how we measure costs.
A single allstaff email at Athabasca will likely be read by about 1200 people, give or take. If such an email takes one minute to read, that’s 1200 minutes – 20 hours – of the institution’s time being taken up with a single message. This is not, however, counting the disruption costs of interrupting someone’s train of thought, which may be quite substantial. For example, this study from 2002 reckons that, not counting the time taken to read email, it takes an average of 64 seconds to return to previous levels of productivity after reading one. Other estimates based on different studies are much higher – some studies suggest the real recovery time from interruptions to tasks could be as high as 15-20 minutes. Conservatively, though, it is probably safe to assume that, taking interruption costs into account, an average allstaff email that is read but not acted upon consumes an average of two minutes of a person’s time: in total, that’s about 40 hours of the institution’s time, for every message sent. Put another way, we could hire another member of staff for a week for the time taken to deal with a single allstaff message, not counting the work entailed by those that do act on the message, nor the effort of writing it. It would therefore take roughly 48 such messages to account for a whole year of staff time. We get hundreds of such messages each year.
But it’s not just about such tangible interruptions. Accessing emails can take a lot of time before we even get so far as reading them. Page rendering just to view a list of messages on our web front end for our email system is an admirably efficient 2 seconds (i.e. 40 minutes of the organization’s time for everyone to be able to see a page of emails, not even to read their titles). Let’s say we all did that an average of 12 times a day – that’s 8 hours, or more than a day of the institution’s time, taken up with waiting for that page to render each day. Put another way, as we measure such things, if it took four seconds, we would have to fire someone to pay for it. As it happens, for another university for which I have an account, using MS Exchange, simply getting to the login screen of its web front end takes 4 seconds. Once logged in (a further few seconds thanks to Exchange’s insistence on forcing you to tell it that your computer is not shared even though you have told it that a thousand times before), loading the page containing the list of emails takes a further 17 seconds. If AU were using the same system, using the same metric of 12 visits each day, that could equate to around 68 hours of the institution’s time every single day, simply to view a list of emails, not including a myriad of other delays and inefficiencies when it comes to reading, responding to and organizing such messages. Of course, we could just teach people to use a proper email client and reduce the delay to one that is imperceptible, because it occurs in the background – webmail is a truly terrible idea for daily use – or simply remind them not to close their web browsers so often, or to read their emails less regularly. There are many solutions to this problem. Like all technologies, especially softer ones that can be used in millions of ways, it ain’t what you do it’s the way that you do it.
Email is just a small part of the problem, though: we use a lot of other websites each day. Let’s conservatively assume that, on average, everyone at AU visits, say, 24 pages in a working day (for me that figure is always vastly much higher) and that each page averages out at about 5 seconds to load. That’s two minutes per person. Multiplied by 1200, it’s another week of the institution’s time ‘gone’ every day simply waiting to read a page.
And then there are the madly inefficient bureaucratized processes that are dictated and mediated by poorly tailored software. When I need to log into our CRM system I reckon that simply reading my tasks takes a good five minutes. Our leave reporting system typically eats 15 minutes of my time each time I request leave (it replaces one that took 2-3 minutes). Our finance system used to take me about half an hour to add in expenses for a conference but, since downgrading to a baseline version, now takes me several hours, and it takes even more time from others that have to give approvals along the way. Ironically, the main intent behind implementing this was to save us money spent on staffing.
I could go on, but I think you see where this is heading. Bear in mind, though, that I am just scratching the surface.
My point in writing this is not to ask for more efficient computer and admin systems, though that would indeed likely be beneficial. Much more to the point, I hope that you are feeling uncomfortable or even highly sceptical about how I am measuring this. Not with the figures: it doesn’t much matter whether I am wrong with the detailed timings or even the math. It is indisputable that we spend a lot of time dealing with computer systems and the processes that surround them every day, and small inefficiencies add up. There’s nothing particularly peculiar to ICTs about this either – for instance, think of the time taken to walk from one office to another, to visit the mailroom, to read a noticeboard, to chat with a colleague, and so on. But is that actually time lost or does it even equate precisely to time spent? I hope you are wondering about the complex issues with equating time and dollars, how we learn, why and how we account for project costs in time, the nature of technologies, the cost vs value of ICTs, the true value of bocce tournament messages to people that have no conceivable chance of participating in them (much greater than you might at first imagine), and a whole lot more. I know I am. If there is even a shred of truth in my analysis, it does not automatically lead to the conclusion that the solution is simply more efficient computer systems and organizational procedures. It certainly does bring into question how we account for such things, though, and, more interestingly, it highlights even bigger intangibles: the nature and value of work itself, the nature and value of communities of practice, the role of computers in distributed intelligence, and the meaning, identity and purpose of organizations. I will get to that in another post, because it demands more time than I have to spend right now (perhaps because I receive around 100 emails a day, on average).