2009 Prediction: There Will Be Pronouncement That Unnecessary Interruptions and Information Overload Tops $1 Trillion ($1,000,000,000,000)February 17, 2009 at 5:05 pm | Posted in Attention Management, etiquette, Information Work, interruption science, knowledge management | 10 Comments
Commentators and average folk alike were aghast as the amount of the financial bailout crept towards the $1 trillion mark. But as Congress backed away into sub-$800bn territory (for now), another cost is likely to be announced that beats them to this lofty mark: the cost of information overload.
These Basex figures get quoted a lot in the press and, while I do believe that many people and organizations do suffer from information overload, I’m not buying into attempts to quantify it and certainly not at a price tag of $1,000,000,000,000. In fact, I think there’s long term harm from trying to get people to act by shocking them with inflated numbers. Just look at knowledge management. KM was a real issue and worthy cause too, before it was done in by money-losing attempts to recover the huge dollar estimates of its inefficiencies.
How do I know this pronouncement is coming? I’ve been following their stats on “unnecessary interruptions” for some time. They went from $588 billion in 2005 (for interruptions without the “unnecessary” tag) to $650 billion in 2007 (you’d think the number would decrease when just the unnecessary ones are counted, but it jumped up instead). In December, they posted a blog entry saying
According to our latest research Information Overload costs the U.S. economy a minimum of $900 billion per year in lowered employee productivity and reduced innovation. Despite its heft, this is a fairly conservative number and reflects the loss of 25% of the knowledge worker’s day to the problem. The total could be as high as $1 trillion.
I’ll have to examine that new research more. How do the interruption and information overload numbers intersect? Are they separate (totalling $1.6 trillion?) or are interruptions part of information overload (which makes sense, but then why is the umbrella number smaller than the 28% of worker’s day previously quoted for interruptions?).
If the $1 trillion figure is anything like the $650bn number I’m not going to buy it. I haven’t seen a full disclosure on their methodology for workplace interruptions, but from what I could glean there were potentially several techniques used to generate a large number:
1. Lumping in social interactions and distractions with interruptions
Just lump all time wasting annoyances, distractions, and socializing in with the more scientific-sounding “interruptions” and you’ll get a pretty big number. Or better yet, don’t define interruption in any strict sense and survey takers will do the lumping for you. You’ll be able to lump 28% of the average information worker’s day into this category.
2. By counting all costs and no benefits (quote “total cost” instead of “net cost”)
How do you lose $10,000 at blackjack while walking out with the same $100 you went in with? Simple, just tally up all the losing hands and ignore the winning hands. Play 200 hands at $100 per hand, win half and lose half, and you’ll come out even. But that means you lost $10,000 (the total of the 100 losing hands)!
If you don’t like the blackjack analogy, then plug in your own one-side-of-the-coin analogy. How about totaling up just the expenses on a large company’s income statement without subtracting it from revenue and being shocked at how big the number is and the potential that even a small amount of improvement in that number could make?
One non-Basex study I saw asked how many interruptions people had, then assumed 50% of them were unnecessary based on other research. Fine, but then interruptions as a whole average out, don’t they? You can still optimize – a company that’s at break-even can always reduce costs, but the size of the total cost pool is not the issue then. It counted all the losing hands (calling them “unnecessary”) and ignored the winners. The implication is that you can keep all the necessary ones and chip away at the unnecessary ones, but who is involved in judging an interruption as “‘unnecessary?”
3. By ignoring closed-loop analysis
Here’s a surefire way to double the $10,000 in losses I quoted in #2. Just interview everyone at the table (you and the dealer) and add up all their losses. Since the half I lost was $10,000 and the half the dealer lost was $10,000, that’s $20,000 in total losses at that table. But we both came away even!
Basex went to some effort to quantify “unnecessary” as not urgent, not important, could have been done another way, etc. But if you ask individuals this instead of both sides of each transaction, you’re just interviewing the dealer and the player about their blackjack losses and forgetting that quite often one wins when the other loses. Almost every possible model I can think of for interruptions (see interruption patterns) results in one of the parties involved losing on the deal, so pretty much every interruption will be counted as unnecessary by someone and without closed-loop analysis almost every interruption will get incorrectly totaled.
You need to do closed loop analysis – treating each interruption as an interaction between the interrupter and those interrupted and determining, as a whole, if it was useful to the organization. Most interruptions are useful to someone, or why would they do it (I propose only a small proportion are careless etiquette transgressions)? If it’s a matter of self-important timing on the part of the interrupter, consider if there is ever really a “good” time you could push these interruptions to.
4. By playing loose with the definition of “unnecessary”
Reversing a question can help validate it. In this case, ask the question from the other side to see if you get the same answer. Ask each survey taker how many times they interrupted someone else that day and how many of those were unnecessary. If the interrupter thinks it was necessary, shouldn’t a conservative estimate give them the benefit of the doubt? I predict the difference in results between the question that yields $900bn and this one would be enormous. Only a small portion would be because the interrupter forgets they interrupted someone – the rest is the inaccuracy of the methodology.
In common parlance, any unnecessary activity interrupts a necessary one you’re working on. Have to stop working on your coding to go to a stupid meeting? That meeting interrupted your coding. That’s 1 hour of interruption plus 15 minutes to get back to what I was doing. If I decide to take a break and look at email, and then get sidetracked by a dumb one? The email “interrupted” me unnecessarily. If you want to let survey takers count all unnecessary activities as “unnecessary interruptions” that’s fine, but throwing interruption technology and etiquette solutions against the general problem of business inefficiency is like throwing a pebble at a wall to knock it down. The survey definitions and the solutions have to use the same definition of “unnecessary”.
5. By comparing against perfect short-term productivity instead of long-term sustainable productivity
Yes, people take breaks and, being social creatures, they often interrupt others to do it with them. People need breaks. Even the best runners have to pace themselves for a marathon. I calculated that optimal performance for the best marathon runner is obtained by running at only about half speed. What if you spend a bunch of time and effort getting people to eliminate certain time-wasting habits, and they just re-fill that time with other habits because they need or want that time? It may be worth figuring that out before throwing a lot of time and money away.
So you’ve figured out by now that I don’t buy the $900 billion number and I certainly won’t when it hits a trillion. Maybe the surprising part if you don’t regularly read my blog is that I’m very much a believer that attention management is a very useful approach and that organizations and individuals can take real steps to manage their attention better (for enterprises see my Enterprise Attention Management conceptual architecture; for individuals my Personal Attention Management tips). But I also believe in having an accurate picture of costs and benefits.
Another techno-cultural topic I believe in is knowledge management. KM’s basis tenets were sound- that knowledge (or at least “information” if you don’t want to sound too pompous about it) is an asset just like a factory or an employee and needs to be managed as such. But KM became a dirty word after a few years of consultants exaggerating the size of the problem and what could be done about it. It’s taken about a decade for KM to get back on its feet, and only now under new names so as not to arouse those burnt on KM before. I don’t want this to happen to attention management and information overload too. It’s a real problem, but a complex one that is impossible to pin a real number on. And it has real solutions too that can help when recognized problems exist – if you don’t promise too much.
Note: This version has been updated due to a helpful comment from Mark Worth pointing out the shift from quoting “unnecessary interruptions” to “information overload”.