Intake
A short call and a written brief. We name the product, the audience, and the one question the reading must answer. If the product has no living visitors, we stop here.
The method page
How a reading is done
A usage reading is a sequence of paper acts. We take a question, collect traces, sit with the people who leave those traces, draw the paths, and brief the team that can change the product.
What we treat as a trace
A trace is anything that shows a person met the product: a recorded visit, a support ticket, a session you are lawfully allowed to replay, a form that was started and left, a search typed inside the product, a paper queue at a counter that the digital desk was meant to replace. Volume without a path is not enough. A path without a person we can speak with is a sketch, not a reading.
What we will not treat as a finding
A spike after a television mention. A comparison with an unnamed “benchmark.” A colour preference collected from a corridor. A sentence that begins with “users want” and ends with a feature the team already liked. Findings in our reports are tied to a screen, a week, and a set of people.
Languages and place
Interviews are in Georgian or English. Batumi is not a backdrop; it is where the paper is drawn and where the briefing chairs are. When a product serves the Adjara coast — ferries, rooms, clinics, municipal pages — we can sit some conversations in town rather than only by video.
After the briefing
The report is yours. We do not publish it. We do not keep your raw records longer than the period named in the privacy note. If you want the maps redrawn after the product changes, that is a new commission, lighter than the first if the question is the same.
When you are ready, request a study with the product and the question you cannot settle.
Order of work
How a reading proceeds
The sample
You send an anonymised slice of usage records. We say whether it can carry a path. Tamar will send it back if names or passwords are still in the file.
Walkthrough
Someone who knows the product walks every main room with us, in the order a new visitor meets them, then in the order a returning one does.
Conversations
Eight to twelve people who actually use the product. We ask what they came to do, where they stalled, and what they opened by accident. We do not ask them to rate the product.
Drawing
Paths become sheets. Empty rooms get a margin note. A stall is drawn as a stall, not as a percentage. Giorgi will refuse a line the records cannot support.
Briefing
The report is sat with the people who can change the product, at the Batumi desk or by video. A recap letter follows. Two weeks of written questions are included on a full study.