About
Why this exists.
Most advice about work comes in two flavours. Academic papers that are careful but unreadable, and motivational posts that are readable but unsourced. Managers end up running teams on folklore.
The Return on People sits in the gap. Each post takes the newest research on what AI does to people at work, reads it through the lens of work and organisational psychology, and turns it into a real number, a named source, an honest caveat, and one thing to do differently on Monday.
What it is for
AI and new technology arrive in every workplace now. Too often they are rolled out like software and land on people as extra pressure, fear or rework. The Return on People exists to help HR people and team leads do it differently: bring in technology so it helps the workplace and the humans in it, instead of breaking them. Every guide, tool and post is built for that.
Who writes it
The Return on People is written by one person, a master's student in work and organisational psychology. The posts grow out of the reading for that degree: the studies that get taught to future organisational psychologists, translated for the people who actually run teams.
Why I started this
I study work and organisational psychology. In my reading I kept finding careful research on what new technology does to people at work, and very little of it reached the people who actually roll AI out. So I started translating it: one study at a time, with the number, the catch and one thing to try.
I have no consulting background and nothing to sell. What I can offer is careful reading, and showing my work so you can check every step yourself.
The rules every post follows
- The primary source. The study itself, not a blog that mentions it.
- The number is the number. No rounding up to make the point land harder.
- Caveats go in the post. If the sample was small, old, or all students, it says so.
- One practical takeaway. Something a team lead can try this week without a budget.
- Corrections are public. If a claim turns out wrong, the correction gets posted, not buried. See Corrections below.
How every number is checked
Before anything is published, it is read three separate times:
- Facts. Every number, percentage, sample size, author and year is checked against the original study, not against a summary or a news article.
- Claim strength. Does the post say more than the study supports? A survey shows a link, not a cause, and the post has to say so.
- The final text. The finished post or page is read once more as a reader sees it: spelling, the honest limit, and the source line.
If a number cannot be confirmed in the original source, it is not published.
Corrections
None so far. Last checked 12 November 2026. If you spot a mistake, write to [email protected] and the correction will be listed here with the date.
What this is, and what it is not
- It is a free library of checked research on AI at work, written for HR people and team leads.
- It is not a consultancy. There is nothing for sale, no sponsor and no ads. If that ever changes, it will say so on this page first.
- It is not legal advice. Pages about the law say when they were last checked.
Get in touch
Questions, corrections, a study or a case you think deserves a post, collaboration or press: [email protected]. More on what to write about.