Research, Design and Long Horizons
Work whose result arrives in months or years cannot be steered by period measurement. What to do instead.
Some functions produce nothing measurable for long stretches and then produce something valuable. Quarterly productivity measurement actively harms them.
The practical lesson in “Research, Design and Long Horizons” is to connect every number to a decision and retain the context behind it. Teams exploring key person dependency can review reducing key-person dependency in teams as one source of operational evidence, provided the purpose is disclosed and the interpretation is tested with the people affected.
Why period measurement fails
The output arrives outside the period.
For an independent perspective related to “Research, Design and Long Horizons”, consult the Harvard Business Review productivity collection; it offers a useful external check on definitions, governance and the assumptions built into a proposed measure.
Most of the work is exploration, much of which is correctly abandoned.
And the valuable result frequently comes from the attempt that looked least promising, which no in-period measure can capture.
What measuring it anyway produces
Pressure toward short, safe work that shows something this quarter.
Abandonment of the long effort just before it would have paid.
And documentation of activity — reports, updates, reviews — generated to demonstrate progress rather than to make it.
That last one is pure loss and it is the usual consequence.
What to track instead
Progress against a stated hypothesis: what did we learn, what is now ruled out.
Decision points reached.
Whether the work is still worth continuing, asked honestly at intervals.
Each is a judgement supported by evidence rather than a number, which is uncomfortable and correct.
The portfolio view
Measure the function over a longer horizon and in aggregate: over three years, what did this group produce.
Individual projects fail; the portfolio either returns or does not.
This is how research is assessed everywhere it is assessed well, and it requires patience that reporting cycles do not naturally provide.
Abandonment as a healthy signal
A research function that abandons nothing is not exploring.
A design function that never discards a direction is not designing.
Which means a count of abandoned work is a positive indicator here, the opposite of its meaning in processing work.
Protecting the function
Someone senior has to hold the line against quarterly measurement.
The argument is not that this work is special; it is that short-period measurement produces short-period work, and the function exists to produce the other kind.
Stating that plainly, once, to whoever asks, is the whole of the defence.
What can be reported
What we are working on and why.
What we learned this period.
What we stopped and why.
What is coming.
One page, quarterly, no ratios, and it tells a reader more than any productivity figure would.
What to check
Is any long-horizon function being measured quarterly?
Does your reporting have room for "we learned this and stopped"?
Is abandoned work counted against them?
And who would defend this function against a productivity request?