[kausora]

A design preview — nothing on this screen is saved

Performance ranking and revenue attribution is built in Sprint 09. The figures below are illustrative, the controls change nothing, and none of it comes from your workspace.

Measured against what each one was for

An awareness post and a sales ad are not comparable, so they are never ranked together. Each creative sits in a cohort of work with the same objective, channel type and window.

Awareness

Ranked on reach

  • Spring Trails — cold open

    TikTok

    412,900

    1 of 9 in cohort

    Updated 4 hours ago

Engagement

Ranked on saves per 1k reach

  • Spring Trails — kit layout

    Instagram

    18.4

    2 of 11 in cohort

    Updated 4 hours ago

Sale

Ranked on cost per order

  • Spring Trails — bundle offer

    Meta Ads

    £14.20

    3 of 7 in cohort

    Updated 26 hours ago · still estimated by the platform, not final

Traffic

Ranked on outbound clicks

  • Tarn Fleece — restock story

    Instagram Stories

    2,180

    4 of 6 in cohort

    Updated 4 hours ago

Outcomes, and how well we can actually know

Every figure carries the standard of proof behind it. A strong number with weak evidence is still a weak number.

£18,420

Spring Trails — GB

order_placed
directA code or identifier ties the touchpoint to the outcome.
Model
Direct identifier v1.3
Window
30 days from touchpoint
Data cutoff
25 July 2026, 06:00 UTC
Completeness
98%

Counts orders carrying the COLDSPR25 code only. Orders from the same visit without the code are not included here.

€9,180

Spring Trails — DE

order_placed
platform_attributedThe platform reported it under its own attribution settings.
Model
Platform reported v1.3
Window
7-day click, 1-day view
Data cutoff
25 July 2026, 04:00 UTC
Completeness
86%

Reported by the advertising platform under its own attribution settings. Not reconciled against the Shopify order record.

£3,940

Tarn Fleece — restock

order_placed
journey_influencedAn authorised sequence touched these customers first. Influence is inferred.
Model
Position based v1.1
Window
14 days
Data cutoff
25 July 2026, 06:00 UTC
Completeness
74%

An authorised Connect sequence touched these customers before the order. Influence is inferred from sequence membership, not from a persistent identifier.

Spring Trails — organic social

order_placed
correlatedAggregate association only. Nothing links these posts to any individual order.
Model
Correlated aggregate v1.0
Window
Weekly
Data cutoff
25 July 2026, 06:00 UTC
Completeness
41%

Aggregate association between posting weeks and order volume. No identifier links these posts to any order. This figure cannot be read as revenue caused by the posts.

What the pattern memory suggests

Learned from this workspace only. Approving one drafts a variant. It never posts anything.

Reuse hookmoderate confidence · 9 comparable posts

Open on the weather, not the product, for TikTok awareness cuts.

Because
Ranked first for reach in a cohort of 9 comparable organic TikTok awareness posts over 90 days.
Would be wrong if
The next two cuts using this hook landing outside the top third of their cohort.
Fatigue risk
Low — used twice in the last 90 days.
Refresh fatiguemoderate confidence · 4 comparable posts

Retire the kit-layout still on Instagram.

Because
Saves per 1k reach fell across four consecutive posts while reach held steady.
Would be wrong if
Saves recovering on the next post without changing the format.
Fatigue risk
High — six uses in 60 days.
Localise marketlow confidence · 2 comparable posts

Shoot German creative in German locations rather than subtitling the British cut.

Because
Sample is too small to rank. Raised because the German cut currently reuses British footage with translated overlays.
Would be wrong if
The subtitled cut performing within the cohort's middle range in October.
Fatigue risk
Not established.