The Weekly Marketing Report, No. 34. The window is 17–23 August 2026. The issue was backfilled — written after the week closed, because it was missed at the time. Companion to 1-research/weekly-reports/2026-W34-trend-report.md.
What backfilling costs, stated plainly. Every item here carries a publication date
inside the window and anything undatable was dropped. But no real-time observation is
possible for a week that had already closed — no search movement, no view counts as
of the week itself. Nothing in this issue rises above level 4, and the single level-4
item is a retailer channel ranking reported by a third party.
Three items were removed on dating grounds and do not appear here, although No. 37
placed them in this window. Bonobos' The Weekday Warrior Show, Cadbury's There's
Magic in More and Amazon's ChatGPT advertising pilot are all 10–11 September
stories. All three errors ran the same way: undated front-page items swept on 11 September
were given dates in the middle of a four-week window instead of at the sweep date. The
correct default for an undated item found on a front page today is today.
Sources swept: the core eleven, this issue's rotation (W34 mod 3 = 2 → Beauty
Independent and Ad Age), the social and creative set, Reklamombudsmannen, and
CEW's Beauty's Top Headlines, which is dated daily and carries its own sources —
the highest-yield instrument for reconstructing a beauty week.
Type. Changed with the 2026-09-19 pass, toward the register of a fashion title. Bodoni Moda — a Didone, with hairline strokes, flat unbracketed serifs and vertical stress — carries every heading: the masthead, the department headings, the card titles, the pull quotes, the large figures and the lead, italic where a word is being pointed at. It is a variable face with an optical-size axis, so it is set at the size of the place it appears rather than scaled: opsz 72 on the masthead, the lead title and the pull quotes, 48 on the department headings, 24 on the card titles and 14 in the lead’s running text, where a display cut would go spindly. Archivo, a neutral grotesque, carries every section’s text, the tracked labels and the evidence lines; Menlo appears only on code. Both faces load from Google Fonts, with Bodoni 72, Didot and Playfair Display behind the serif and Helvetica Neue behind the sans. They replaced Cormorant Garamond and Jost across the whole series.
Colour. The page is a whisper off white: #FBF9F6 for the ground and the card face alike (a beige laid over white at eight per cent: warm enough to see, far too light to read as beige) — rules do the separating, not a fill — with #F6F3EE, one step down in the same warmth, for the secondary blocks and near-black #121110 for the pull quotes and this colophon. One accent — bronze #85643A, with champagne #B89A66 for the hairline marks — for the pointed word, the chart and the focus ring; the masthead does not take it and is set in plain ink. Nothing is rounded and nothing casts a shadow. Strong colour appears only on the finds, in four places: each card’s art panel, a 3px spine along the top of the card, a 7% wash of the same colour behind the chapter heading and its entry in the key (8% in the dark setting), and the sector chip — beauty rosewood #BE7268, fashion plum #755980, FMCG sage #446E67, adjacent stone #918779, and the record uncoloured, its spine and rule in ink. Market stays a plain chip. A dark setting inverts the neutrals and keeps the panels.
Contrast, computed not eyeballed — in the light setting. The page and the card are the same near-white, so each tone has one figure on the page and one on the grey block. Ink #111010 measures 18.08:1 on the ground and 17.16:1 on the grey; secondary #4B4845 8.64:1 and 8.21:1; tertiary #6B6660 5.41:1 and 5.14:1 — darkened for the warm ground, because it darkens the chapter bands with it and the old #6F6A64 fell to 4.42:1 on the record’s band. The bronze reads 5.15:1 on the ground. The chapter bands are the tightest place on the page, and the tertiary clears the floor on every one of them — the record's neutral band, the darkest, at 4.69:1. On the near-black blocks, #F2F0EC reads 16.57:1 and the secondary #B4B0A9 8.73:1. Text on the sector chips: ink on rosewood 5.24:1, white on plum 6.01:1, white on sage 5.72:1, ink on stone 5.38:1 — every one clears 4.5:1 for small text. The drawings sit on the same panels, and the second tone inside each — white on rosewood 3.62:1, ink on plum 3.16:1, ink on sage 3.32:1, white on stone 3.53:1 — clears 3:1 as a graphic. The plum was set by this constraint: a deeper one reads better on the chip but takes the drawings’ second tone under 3:1. Spines and department rules are graphics too and clear 3:1 against the ground on their own: rosewood 3.45:1, plum 5.71:1, sage 5.44:1, stone 3.36:1. Chart: bronze 5.15:1 and ink 18.08:1 on the ground, far enough apart in lightness to separate under every form of colour-vision deficiency.