Question #282

Can a city explain its own decisions?

REALITY EXPERIMENT · PROTOTYPE

Project Inside

What is this workplace actually like — before you give it years of your life?

Glassdoor tells you what people wrote. Inside shows you what repeatedly happened.

A star rating averages opinions. It does not tell a candidate whether the advertised salary was honoured, how long the process took, or whether the role matched the advert.

Inside collects small, structured observations at each stage of a job journey — application, interview, offer, onboarding, six months, exit — and publishes only the patterns that repeat.

No reviewer profiles. No comment threads. No paid removal of criticism.

Workplace reality map

Enter an employer and a role.

Northgate Logistics

Logistics · 1,200 staff · Warehouse Operative

148 observations · Updated 18 days ago

  • Recruitment clarity

    Usually clear

    Most contributors reported knowing the stages, the pay band and who they would meet before the first interview.

  • Advertised salary honoured

    84%

    84% of recorded offers landed inside the advertised band. Reductions clustered around one shift pattern.

  • Interview respect

    Strong

    Interviews began on time and ran to the stated length in most recorded cases.

  • Role matched advert

    Mixed

    Planning roles frequently absorbed unadvertised weekend cover within the first 90 days.

  • Workload predictability

    Low

    Hours varied week to week for the majority of contributors in operational roles.

  • Management consistency

    Varies by team

    Two sites reached the response threshold and produced clearly different results.

  • Feedback after rejection

    Rarely given

    Most candidates recorded no response after a final-stage interview.

Employer response · evidence supplied

We have published our shift-pattern pay bands and introduced a 10-day feedback commitment. Evidence submitted March 2026.

Project Inside is a Quinn Labs research prototype. The employers, signals and percentages shown are illustrative sample data created to demonstrate the interface. They do not describe any real organisation. Nothing you record here is transmitted or published; it is stored only in your own browser.

Contribute · private to this browser

Record what happened to you, one stage at a time.

Six short observations. No free text, no name, no profile. In the prototype nothing leaves your device.

  1. 01 · Application

    Was the advertised pay range stated before you applied?

  2. 02 · Interview

    Did the process run as described to you?

  3. 03 · Offer

    Did the offer match the advertised range?

  4. 04 · First 90 days

    Did the work match the job description?

  5. 05 · Six months

    Were your hours and flexibility respected in practice?

  6. 06 · Exit

    Would you choose the same manager or team again?

0 of 6 stages recorded

How it behaves

No followers. No threads. No paid removal.

  • No public reviewer profiles

    Contributions carry no name, photo, follower count or history. There is nobody to pile on to.

  • Structured observations only

    Contributors answer defined questions about what happened. There is no free-text allegation box.

  • Response thresholds

    Nothing publishes until enough independent observations exist. Team-level results require a higher threshold again.

  • Employers may respond with evidence

    An organisation can publish a documented response. It cannot pay to remove or soften a result.

  • AI summarises, never invents

    Pattern summaries are generated only from recorded answers, and always show the number of observations behind them.

  • Serious allegations leave the product

    Anything alleging unlawful conduct is routed to a separate moderated process, not the public reality map.

The difference

  • An overall star rating

    Signals for each stage of the job journey

  • Opinions written once, years ago

    Evidence freshness shown on every signal

  • Anonymous free text

    Structured, comparable observations

  • Reviewer profiles and replies

    No profiles, no threads, no likes

  • Paid reputation management

    Employers pay for private diagnostics only

How it could pay for itself

  • Free for candidates

    The essential reality view is free and requires no account to read.

  • Private diagnostics

    Employers may purchase their own detailed breakdown, including team-level patterns held below the public threshold.

  • Benchmarking

    Sector and region comparisons against anonymised aggregate data.

  • Verified improvement

    An organisation can record a change and let later observations show whether it held. It cannot buy a better public score.

Open questions

  1. 01What is the minimum number of observations before a signal is honest rather than noisy?
  2. 02How do you keep a contribution unidentifiable inside a five-person team?
  3. 03Can an employer response be published without becoming a right of reply that drowns the evidence?
  4. 04How should a signal decay when nobody has contributed for a year?
  5. 05What prevents an organisation from seeding its own favourable observations?
  6. 06Should a candidate be able to see a result before they apply, or only after?
  7. 07Does showing 84% change behaviour more than showing a star rating?

Part of The Reality Layer — promise to experience.