A Very Silly KPI

Shop in a typical UK supermarket and you’ll see how little priority is given to customer service. Checkout staff are measured on how many items they scan per minute. This doesn’t speed up the checkout process. The goods pile up in a random heap at the end of the conveyor as the hapless customer tries to re-sort them into a sensible order for packing. The end result is an irate customer, damaged goods, and a queue which is processed more slowly than if the checkout operator had been allowed to deliver the simple customer service of matching scanning speed to the customer’s packing speed. Similar examples can be found in many other areas, where easily measurable targets are used as a substitute for good process. The DeliveryDemon is not impressed!

Delivering Spurious Accuracy, Demanding Constant Attention

The DeliveryDemon did a double-take. The hospital receptionist had actually offered a follow-up appointment time of 8.48 a.m. Not 8.45, not 8.50, 8.48 precisely!

Great efforts were made to get the patient to hospital for the prescribed minute. As the seconds clicked by, the DeliveryDemon gazed at the white blocks on the bilious yellow screen, idly wondering whether the developers were aware that the inability to distinguish between yellow and white is a common form of colour blindness. The clock ticked over to 8.48…… Nothing happened! At 8.55 the appointment pinged up for its allotted 3 seconds then disappeared into the ether, never to re-appear.

There is a huge disconnect between the design of this system and practical reality. The underlying driver may well be a target of fitting 5 x 12 minute appointments into the hour, but that 12 minutes is an average. Pinning appointments to an exact minute means that overruns delay subsequent appointment, but nothing is gained if an appointment finishes early – an approach which increases the likelihood of targets being missed. It also invites mockery.

The mechanism for summoning patients is equally poorly conceived. It relies on the assumption that patients gaze non-stop at the single screen, waiting for their numbers to flash up for those 3 short seconds. In reality, pillars obscure the screen from some seats, and passers-by may obscure it from any position. The area, lacking sound absorption, is noisy, so any audio cue is lost. A patient with poor eyesight may need to move closer to the high-mounted screen to read it, and age or infirmity would make those 3 seconds of visibility completely inadequate. And of course real patients are chatting, reading newpapers, watching the world go by, as the clock edges beyond the allotted minutes of their appointments. That sickly yellow screen is by no means the cynosure of all eyes.

Part of the DeliveryDemon wanted to laugh at the absence of basic common sense in the design. The reason she is not still in giggling thrall to those ridiculous flaws is the context. This was an NHS hospital. Huge sums of taxpayers’ hard-earned money went into the creation of the system. The appalling design is unlikely to result in patient fatality, but the blatant absence of commonsense in a patient-facing system must call into question the quality of other systems which are life-critical.