Satisfaction scores have a credibility problem that the industry has been politely ignoring for about fifteen years.
They are collected from a self-selecting minority, immediately after an interaction, by the organisation being assessed, using a question most respondents answer on autopilot. They correlate weakly with retention. They move for reasons unconnected to service quality — seasonality, question wording, where the survey appears. And they are reliably high in operations that customers describe, in their own words, as exhausting.
Effort is the better measure, and it has one property that satisfaction does not: it describes something the organisation can actually change.
Why satisfaction misleads
Three defects, each survivable alone and fatal in combination — and specialists who assess customer experience across outsourced operations will usually find all three sitting in the same monthly report.
Response bias
The customers who complete surveys are not the customers who leave. People who have given up rarely stay on the line to rate the experience, which means the most important population is systematically absent from the data.
Recency framing
A survey delivered immediately after a resolution captures relief rather than judgement. Ask the same customer a fortnight later, after the refund failed to arrive, and the answer differs considerably.
Averaging across incompatible populations
Automated resolutions produce a bimodal distribution — strong when they work, poor when they do not, with little middle. Human interactions produce a fat middle. Averaging the two conceals movement in both.
“Every operation I have ever reviewed had a satisfaction score it was comfortable with,” said Ralf Ellspermann, founder and chief strategy officer at PITON-Global. “Comfortable scores and comfortable customers are different populations and nobody notices until churn moves.”
What effort measures instead
Effort asks how hard the customer had to work to get what they needed. It is the closest available proxy for the experience people actually describe when they complain, and it predicts disloyalty considerably better than satisfaction predicts loyalty.
“Ask an operation to show you how often customers repeat themselves and you learn two things,” said John Maczynski, chief executive of PITON-Global. “What the number is, and whether they are able to find out. The second answer is usually the more revealing.”
Crucially, it decomposes. A high-effort interaction is high-effort for identifiable reasons, each of which is fixable — which is why effort travels better than satisfaction does across a reporting boundary, whether that boundary sits with a domestic supplier or with the Philippine call centre industry.
Channel switching. The customer tried self-service, then chat, then telephone. Each transition is effort and each is visible in the data.
Repetition. The customer explained the problem more than once — to an automated system, then to an agent, then to a second agent after transfer. Repetition at handoff is the single most diagnostic effort signal in a modern operation.
Contact count. One issue, several contacts. The measure that matters is per issue, not per contact, which requires linking records across channels and is the reason many operations cannot produce it.
Delay with uncertainty. Waiting is tolerable when the customer knows what they are waiting for. Waiting without information is where effort concentrates.
Measuring it without a survey
The most useful effort measurement in a UK operation is not a survey question at all. It is behavioural, and it is already in the data.
Repeat contact rate per issue within a defined window. Channel transitions per resolved issue. Transfer rate, and specifically transfers after the first minute, which indicate misrouting rather than appropriate escalation. Proportion of interactions where the customer restated information the organisation already held.
That last one requires transcript analysis rather than metadata, which is why it is rarely measured and why it is worth measuring. It is the closest thing to a direct observation of wasted customer time that an operation can produce.
Why this matters more in outsourced operations
Because effort accumulates at the seams, and an outsourced operation has more seams.
Between the machine tier and the human one. Between the offshore floor and whatever is retained in Britain. Between the front office and whichever back-office queue actually completes the work. Each junction is a place where context can be lost and the customer asked to supply it again.
None of these seams appears in a conventional scorecard. Each team measures its own performance and reports it accurately, and the customer’s experience of the whole remains unmeasured by anybody — which is how an operation ends up with four green dashboards and a rising complaint rate.
Start by instrumenting the junctions. They are where the organisation’s view and the customer’s diverge most sharply, and they are the one part of the operation nobody has been asked to report on.
Making the transition
Nobody should abandon satisfaction measurement abruptly. Finance and operations have years of trend data and comparability has value.
Three things make that transition succeed. Agreeing definitions in writing with the provider before the first report, because issue-level contact counts can be computed several defensible ways. Instrumenting the handoff deliberately rather than inferring it from existing logs. And giving one named person ownership of the metric set, because a scorecard owned by everyone is revised by nobody.
The workable approach runs both for several quarters. Keep reporting satisfaction, clearly labelled for what it is. Stand up the effort measures alongside it, instrument the seams, and build enough history to make decisions from. Then let the old measure recede to a secondary position rather than retiring it with an announcement.
See also: The Future of Tech-Driven Businesses
Accurate, or comfortable?
What a buyer actually wants to know is whether customers find the organisation easy to deal with.
Satisfaction answers a different question — whether the people who bothered to respond felt reasonably treated at the moment they were asked. That is worth knowing and it is not the same thing.
Effort answers the question directly, in terms the operation can act on, using data it mostly already holds. The reason it remains the minority practice is not that it is harder to measure. It is that the numbers it produces are less flattering, and an organisation has to decide it would rather be accurate than comfortable.
