Marbling · B-ONLY

Testing names and voices in the tray

Ink swirls on the tray never repeat — yet each sheet must belong to the same book.

How are naming and voice options tested once a positioning line is signed? Like marbling — multiple variants floated, disturbed with deliberate combs, then read against real sentences until one pattern clears the bench.

  1. Fill the tray. We collect names or voice variants from structured workshops — never brainstorm chaos without criteria. Each variant links back to the positioning line. Orphans are folded away immediately.
  2. Comb the surface. Variants are stressed: spoken aloud by non-marketers, typed into headline-length spaces, embedded in an investor paragraph and a customer-support reply. Weak options reveal themselves quickly.
  3. Read against real sentences. Voice work uses your actual copy — error messages, onboarding steps, recruitment posts — not lorem ipsum. If a voice makes your API documentation sound absurd, it fails here, not after launch.
  4. Reject with reasons. Discarded names and voices are logged with plain-language rationale so stakeholders do not revisit them silently. Surviving options are fewer than five — usually two.
  5. Fix the pattern. The chosen name or voice is written into a guide with examples your team will reuse. Marbling ends when the guide survives a week of internal use without emergency edits.
  6. Archive the tray. Rejected variants and test sentences are filed in an appendix — useful when a stakeholder asks why a favourite name is absent six months later.
Swirled marbled paper sheets drying flat against a white wall
Marbled sheets — each unique, each part of one volume
A handwritten letter signed at the foot resting on a desk
Signed approval — decisions locked on paper

We do not sell naming as a standalone trophy hunt. Names serve positioning. Voices serve narrative. If your brief skips the bonefold, we return you to it — marbling on unstable paper produces patterns that crack when pressed.

Marbling sessions include people who will live with the name daily — support leads, product owners, sometimes board observers. Regional teams hear names spoken in Singapore English first, then test translations where needed. Voice variants are read at normal speaking pace because names must survive Monday stand-ups. Combs, in our metaphor, are editorial tests — headline length, investor paragraph, support reply — not physical tools.

No name proceeds to domain registration without a signed shortlist and documented rationale. See Casebound for anonymised sketches of how this sequence played out in past collaborations.

Rejected names and voice variants are filed with reasons so future hires understand what was tried. Surviving options move to message hierarchy work described under Gatherings.

Naming trays hold variants side by side so comparisons stay honest. We read candidates aloud in a quiet room — if a name stumbles in speech, it rarely survives on a homepage.

Combs in the tray metaphor stand for structured tests: headline patterns, error messages, recruitment paragraphs, investor one-liners. A voice that works in one context but fails in another is rejected even when it sounds elegant in isolation.

Domain registration waits until leadership initials the shortlist — we do not recommend purchasing names that have not survived the tray tests described above.

Marbling page · studio map