One consequence of working from home is that mistakes are made. People miss messages; spinning plates fall to the ground; the falcon cannot hear the falconer; a New Yorker journalist broadcasts his genitals to everyone else on a Zoom call.
Much has rightly been written about the emotional challenges of remote working, the difficulty of spotting colleagues in distress and the inadequacy of email as a way of providing support and kindness.
What gets less attention is the simple communication error. It is remarkable that activities for which face-to-face co-ordination seemed essential can be conducted from kitchen tables across the world. Yet eventually the cracks start to show: everything from missed emails causing missed meetings to yawning chasms of understanding about fundamental goals.
I have been searching for solutions. One source of inspiration is the humble credit card. I learnt recently from A Podcast of Unnecessary Detail that what looks like a 16-digit credit card number is in fact a 15-digit number plus a check digit. The check digit is calculated using a formula that involves each of the other digits. If any single digit is mistyped, the check digit will not match; most transpositions, such as typing 47 instead of 74, will also be detected.
Check digits are a specific instance of the general principle that redundant information can reduce errors. Redundant information comes with a cost. It takes up space (if you are a digital message such as a credit-card number) or time and attention (if you are a human). But the price may be worth paying.
We can use a phonetic alphabet, which takes much more time, to give a postcode over a patchy phone line. Do you live in east Bristol (BS5) or between Shrewsbury and Stoke-on-Trent (TF9)? “Tango Foxtrot Niner” is quite distinct from “Bravo Sierra Five”. Languages such as English already include considerable redundancy. Tht s why y cn prbbly rd ths sntnc, vn wth th vwls rmvd.
For meetings, I specify both the date and the day of the week. A mismatch is an opportunity to notice the problem.
Automatic error correction is even better. I used to be irritated when people sent me electronic calendar invitations to events I’d already noted down. Now I realise it’s a valuable opportunity to make sure everyone has the same understanding of what is going on. A well-crafted invitation includes the correct link for a meeting and the contact details of everyone involved, and automatically ensures that everyone has the same timings, no matter their timezone. Manually assembling such information from email chains, phone calls or, worst of all, PDFs offers opportunities for error.
Brevity helps if it focuses attention on critical information. Too often I receive emails with a subject such as “URGENT PLEASE READ”, when the entire message could be within the subject line: “The 4pm meeting is postponed to 4.15pm.”
But brevity can also be a trap. Tom Standage’s book The Victorian Internet describes the woes of Philadelphia wool broker Frank Primrose, who in June 1887 sent a telegram that used a compressed code. (Telegrams were expensive and charged by the letter, so such codes were common.)
The message read BUY ALL KINDS QUO rather than BAY ALL KINDS QUO, leading Primrose’s agent to erroneously buy half a million pounds of wool. A little more redundancy in the telegraph message would have saved Primrose a small fortune. These days, most messages with such high stakes would come with error detection built in, as with the check digit on a credit card.
Catastrophic typos are easily spotted, but a more insidious threat is the slow accumulation of what the psychologist James Reason calls “latent conditions”. In an industrial context, latent conditions might include an untidy pile of flammable oily rags in the corner of a workshop, a fire alarm that is out of batteries and a blocked fire escape. None of these latent conditions will cause trouble, until a stray cigarette causes a fatal fire.
For an organisation with an ambitious project to deliver, the flat batteries and the blocked fire escape are metaphorical. Colleagues working together on a project can find themselves labouring under very different assumptions about deliverables, budget, priorities and even who is handling which tasks. The mismatch of beliefs can persist for a long time until a sudden and perhaps catastrophic moment of realisation. As the saying goes, it’s not the things you don’t know that cause trouble — it’s the things you do know that aren’t true.
A related issue is when some participants fear that a project is badly flawed, but there is no easy, informal way for them to pull a superior to one side and discreetly express their disquiet. One can’t help wondering if this summer’s bungling of grades for cancelled exam results in the UK would have been quite so shambolic if everyone concerned had regularly seen each other face to face.
There is no easy solution to this problem. Error detection is an endless process. One possibility is to reduce the level of complexity by pursuing projects that involve fewer people, simpler technology and more slack. Another is to embrace brief, frequent conversations to ensure everyone is on track. Are we sure this is a good idea? Without those frequent informal checks, things fall apart.
Written for and first published in the Financial Times on 7 November 2020.