What it means
The index exists because raw survey data is unwieldy. A 40-question staff survey produces 40 separate averages, and nobody in a board meeting can hold that in their head, so a handful of core questions are combined into one number instead.
A common construction uses three questions answered on a 1 to 5 scale: how satisfied you are overall, how well the job matches your expectations, and how close it is to your ideal job. The three averages are combined, then rescaled so the worst possible answer is 0 and the best is 100.
It matters commercially because satisfaction leads other numbers. Teams with low scores tend to show higher absence, higher voluntary leaving and slower delivery a few months later, which makes the index an early warning signal rather than a soft nicety.
The index is used comparatively, not absolutely. A score of 68 means very little on its own, but 68 this quarter against 74 last quarter, or 68 in operations against 81 in finance, is a clear signal about where attention is needed.
Two cautions are worth holding onto. Response rates below roughly half the workforce make the number unreliable, and an index only stays meaningful if the questions and scale stay identical between rounds, so resist the temptation to reword the survey.
The index is also only as good as what happens after it. Companies that publish results and then act visibly on one or two specific issues tend to keep response rates high, while those that survey repeatedly and change nothing find participation and honesty both drifting downwards within a couple of rounds.
In practice
Real-world examples.
Example
A retail bank runs its index quarterly and finds branch staff at 61 against head office at 79. The gap prompts a review of branch staffing levels, which had been cut without adjusting opening hours.
Example
A fast-growing marketing agency ties 20% of each department head's bonus to their team's index score. Scores rise, but the finance team also introduces an anonymity threshold so no manager can identify individual responses.
Example
A hospital trust compares its index before and after introducing self-rostering for nurses. The score moves from 58 to 71 over two survey rounds, giving the project sponsor concrete evidence to extend the scheme.
Think of it
“Satisfaction index shows how happy employees are with their jobs and workplace.
Formula
Calculation
Employee Satisfaction Index = ((Average score across the index questions - 1) / (Maximum score - 1)) x 100, using a 1 to 5 answer scale.
A manufacturing firm surveys 400 staff and gets 310 responses. The three index questions score 4.2, 3.8 and 4.0 on average, so the combined average is (4.2 + 3.8 + 4.0) / 3 = 4.0.
Applying the formula: (4.0 - 1) / (5 - 1) = 3.0 / 4.0 = 0.75, and 0.75 x 100 gives an index of 75. If the same three questions averaged 3.6 next quarter, the index would fall to (3.6 - 1) / 4 x 100 = 65, a six-point drop that is worth investigating even though the raw average moved by only 0.4.Case study
Seen in the real world.
Harbourline Foods is an invented company used here as an illustrative case. Its overall satisfaction index sat at a comfortable 74, and the executive team treated staff sentiment as a solved problem. A new HR director insisted on splitting the score by site.
The split showed three sites in the high 70s and one distribution centre at 52. That site also accounted for a disproportionate share of the company's overtime bill and short-notice absence, which nobody had previously connected to sentiment. Follow-up interviews pointed to a rota system that gave staff less than a week of notice.
Harbourline moved that site to four-week published rotas and rebuilt its shift-swap process. At the next survey the site's index had risen to 66, absence had fallen, and the company had learned that the aggregate number had been quietly averaging away its most expensive problem.
Watch out
Common mistakes.
- Changing the wording or the answer scale between survey rounds, which breaks the comparison that gives the index its entire value.
- Reading the absolute score as if it had objective meaning, rather than treating it as a trend line and a comparison between teams.
- Publishing scores for teams so small that individuals can be identified, which destroys trust and makes future responses less honest.
Questions
People also ask.
What response rate is good enough?
Aim for at least 60% of the workforce, and be openly cautious about any team-level score built on fewer than about ten responses.
Is this the same as employee net promoter score?
No, because that measure asks a single recommendation question and produces a score from -100 to +100, while a satisfaction index blends several questions on a 0 to 100 scale.
How often should the survey run?
Quarterly or twice yearly works for most companies, since running it monthly usually produces survey fatigue and falling response rates without adding much insight.
From the founder's library
Take it further with the book.
Build your financial confidence beyond this definition. Shihan's full-length guide, Accounting Fundamentals, takes the same plain-English approach and turns it into a complete, practical playbook for non-finance managers, business owners and students - with chapter-end quiz answers and presentation slides included.
25% off with code MMHQ25, applied at checkout. Priced in USD - checkout may show the equivalent in your local currency.
View the book and save 25%