Week 5 Autopsy: The reviews nobody believed
Week 5 — plotting review sentiment over time to find the point the decline actually started, versus the point people started talking about it.
24 August 2026
Before the session#
Pull every review you can date and plot rating (or your own rough sentiment call) against time. You’re hunting for the point the line bends, which is very rarely the same week people online started saying “I heard this place is struggling.”
In the session#
Desmond’s argument: a five-star review from a regular and a one-star review from someone who waited forty minutes carry different evidentiary weight, and treating them as equally informative is how a lot of amateur post-mortems go wrong. You practise weighting your own specimen’s reviews and defending the weights out loud.
Afterwards#
An inflection point with a date attached, and a first honest answer to whether the reviews saw the ending coming before the owners did — or whether, just as often happens, they didn’t.
The spec
What you bring to the session. Written so a reader can tell whether it has been met without asking you.
- a timeline of review ratings or sentiment across the specimen's lifetime
- an identified inflection point, distinct from the closure announcement
- at least one review explicitly discounted as noise, with a stated reason
Teaching team
- Desmond Achebe — tutor