Psychologic
Fix the emotion behind the metric, not the metric itself
- Difficulty
- Moderate
- Time to result
- ~weeks to results
- Steps
- 4
- Confidence
- 82%
Psychologic is Rory Sutherland's principle that most experiences are hurt by an unmeasured human emotion rather than the measurable metric teams fixate on. The classic case is the Uber map: it does not make your car arrive faster, but by showing the car approaching it removes the pain of uncertainty while waiting. Because organizations have no metric for uncertainty, anxiety or resentment, they default to optimizing speed, cost and distance, which correlate only loosely with satisfaction. The method is to identify the emotional problem behind a metric, then solve that emotion with a cheap perceptual intervention rather than an expensive engineering one. HS2 spends billions cutting minutes off a train when better yield management and a countdown-style anticipation could deliver more felt value for a fraction of the cost.
Origin
Rory Sutherland, Vice Chairman of Ogilvy UK, developed 'psychologic' across his advertising career and books; he articulates it here using Uber's arrival map and Britain's HS2 rail project as contrasting cases.
Core principles
- 01The number you can measure is rarely the thing that actually hurts.
- 02Reducing perceived pain often beats improving the underlying metric.
- 03We optimize speed, cost and distance because they are measurable, not because they matter most.
- 04There is no metric for uncertainty, so engineers ignore it.
- 05Cheap perceptual fixes can outperform expensive engineering.
How to run it
- 1
Name the real emotion
Identify the human feeling the experience actually produces, such as uncertainty while waiting or resentment at being upsold. This is usually not the thing anyone is measuring.
Pro tip If you cannot name the emotion, you are almost certainly optimizing a proxy metric instead.
- 2
Expose the proxy metric
Write down the engineering metric the team is trying to improve (speed, time, cost, distance) and ask honestly whether it correlates with the emotion you just named.
Watch out Do not assume speed and satisfaction are the same variable; they rarely are.
- 3
Design a cheap perceptual fix
Brainstorm interventions that reduce the felt pain without changing the underlying metric, such as visualizing progress or removing uncertainty.
Pro tip The Uber map costs almost nothing yet transforms the wait; look for that kind of leverage.
- 4
Test perception before funding engineering
Run the perceptual fix first. If it moves the emotional outcome, you may not need the expensive engineering project at all.
Watch out Committing billions to an engineering fix before testing perception risks solving the wrong problem entirely.
In the wild
Rory notes the Uber map is ingenious precisely because it does not increase the speed at which your car turns up. Instead, by showing the vehicle moving toward you, it massively reduces the pain of uncertainty while you wait. The felt experience improves even though the physical wait is unchanged, because the real problem was never the minutes, it was the not-knowing.
→ A near-free interface feature solves an emotional problem that faster dispatch could not.
Rory argues Britain's HS2 spends around 60 billion pounds shaving minutes off the London-to-Manchester train, optimizing an engineering metric. He proposes instead spending roughly 500,000 pounds on an app letting pre-booked passengers jump onto earlier trains with empty seats. That reduces real end-to-end journey time and dead waiting, improves load balancing, and earns Virgin more revenue, because it targets the felt journey rather than raw speed.
→ A cheap software fix could beat a multi-billion-pound engineering project on the metric that actually matters to travelers.
Common mistakes
Optimizing the measurable proxy
Teams improve speed, cost or distance because those are easy to measure, while the emotion driving satisfaction goes untouched.
Funding engineering before testing perception
Committing huge budgets to a physical fix before checking whether a cheap perceptual change would do risks solving a problem customers do not actually have.
Is it for you?
Best for
Product, service-design and marketing teams whose real problem is a human emotion, not a technical spec.
Not ideal for
Genuine engineering-limited problems where the felt pain and the physical metric are truly the same thing.
From the transcript
“it doesn't increase the speed at which your car turns up but it massively reduces the pain we experience in uncertainty while waiting for it…”
“we don't have metrics for the human emotions and so we're trying to improve human emotional state by optimizing things which don't really correlate neatly…”
“they're spending 60 billion on the sonic trains because they're measuring the wrong bloody things”
From the episode
The Psychology Of Irrationality - Rory Sutherland - #255
Rory Sutherland