How this study is run

What this is

Where Would You Go? is a pair of short interactive studies run by Alltoogether, a UK workplace health, employee benefits platform and employee benefits broking business. Alltoogether commissioned and paid for the research, and there is no third-party funder.

Both editions ask about the same five everyday workplace situations. The employee edition asks where you personally would go first. The employer edition asks where you think employees at your organisation would go first.

What this is not

• It is not nationally representative. Participants opt in, mostly from paid social advertising, and self-selected samples are not a random sample of anybody.
• It is not a matched study. The two editions are recruited separately. No employer response is connected to any employee response, and no comparison between them describes a gap inside one workplace.
• It is not a clinical or mental health assessment, and it is not a validated psychometric measure.
• It is not employment, legal, medical or financial advice.
• It is not an audit of any participant's employer, and it does not verify what any employer provides.
• It cannot show that navigation causes anything. Nothing here demonstrates an effect on retention, absence, turnover, performance, productivity or health.

What is measured

Employee edition: stated first route in five scenarios, six statements about whether people are asked how they are doing and whether they see anything change, and optional questions about the kind of workplace.

Employer edition: expected first route in the same five scenarios, six parallel statements about the organisation's own practice, and the same optional workplace questions.

One employee question is not hypothetical: it asks whether money worries have affected how you feel at work in the last 12 months and, if so, whether you looked for help and what happened. It is the only item in either edition that records behaviour rather than intention, and it is reported separately for that reason.

Question order and randomisation

Within each scenario the substantive answer options are shuffled for every participant, so no route benefits from always appearing first. Three options are always pinned to the bottom, in this order: that the employer does not offer anything for this, that the participant would not know where to start, and something else. The order each participant saw is recorded, so order effects can be checked rather than assumed away.

The five point agreement scale is never shuffled.

Scoring, and what we refuse to do with it

No composite score is calculated or shown. Participants see a plain description of the pattern in their own answers, chosen by a fixed published rule, not a number, a grade or a benchmark.

For reporting, items one to three are grouped as "asked" and items four to five as "acted". That grouping is ad hoc. It is not a validated scale and it will never be described as one. The sixth item, about whether somebody would feel comfortable saying they were struggling, stands alone and is never rolled into either group.

Anonymity and your email

Research answers are anonymous. Employee participants are asked, optionally, which of four broad age groups they are in, so that answers can be compared across generations; it can be skipped and it is a band, never a date of birth or an exact age. Beyond that we do not collect names, employer names, precise location, exact age, gender, ethnicity, disability, job title, income, or any health information or diagnosis. We do not store your IP address.

If you ask us to send you the findings, your email address is stored in a different place from the answers, with no reference connecting the two. It is stored with the date only and not the time, and with the campaign only and not the specific advert. That is deliberate: a precise timestamp on both sides would let somebody match an email to an answer even with no shared identifier, which would make the promise on the page untrue.

Consent to receive the findings is separate from consent to marketing. Neither is pre-ticked, and research answers are never used to target anybody.

Recruitment

Participants are recruited through paid advertising on Meta platforms, organic social posts and the founder's own channels. Every session records which advert brought it, and at least one advert per edition is a neutral reference creative that asks the study question without suggesting an answer. Findings can therefore be reported for the whole sample and for the reference creative subsample separately, and no headline figure will rest only on a leading advert.

Data quality and exclusions

A valid completion answers all five scenario situations and all six statements. Permitted skips remain valid answers. Completions that are implausibly fast, or that select the same point on every statement, are flagged rather than silently deleted. The exact thresholds are set from the pilot data before any figure is published, and the rules in force are stated with the results.

Reporting thresholds

No numerical subgroup finding is published below 150 valid completions in that subgroup, and no subgroup figure goes into press outreach below 250. Where a figure is reported, the sample size and the self-selected nature of the sample appear in the same sentence, not in a footnote.

Retention

Structured anonymous answers are kept so the study can be repeated and compared over time. Optional free text is kept in its original form for a maximum of twelve months and then deleted, with only de-identified coded themes retained. Emails given for the findings are deleted after the findings are sent, unless separate marketing consent was given.

Version

Instrument version 1, methodology version 1, August 2026. If the wording changes, the version changes with it, and every response records the version it was collected under.