Skip to content
Get Ranked On TopIllustrative case studyContent Engine plan

A fictional business. The figures are a modelled scenario, not a client result.

ALABASTRA PERFUMERY · Niche perfume house, atelier and workshops · Grasse

How a Grasse perfume house became the answer to questions about notes and longevity

+340% non-brand organic traffic, cited in 41 of 120 AI answers

18 months · Europe · Content Engine plan · Query & prompt research · Content pyramid plan · Expert-led guides · AI-answer optimisation

THE CONTENT FORMULA
NotesContentPieces
Top notesQuick answers30
Heart notesGuides22
Base notesDeep expertise12

01THE CHALLENGE

The questions were about scent. The answers came from someone else.

Search and AI answers about notes, longevity and ‘perfume like X’ went to databases and forums; the house's site had product pages and nothing showing its expertise.

Alabastra composes 14 fragrances in a Grasse atelier and teaches small-group workshops upstairs. People search for exactly what the house knows best: what a heart note is, why iris lasts, which perfume smells like a rainy chypre. But the site had product pages and a booking form, nothing else. Search results and AI answers sent those questions to fragrance databases and forums, which then sent buyers on to the big houses.

WHO HELD THE TOP TEN

Sample of 1,900 queries on notes, longevity and ‘perfume like X’

  • Fragrance databases 54%
  • Forums 33%
  • Retailers 9%
  • Brand sites 4%

THE SITE, BEFORE

14 fragrance pages, 6 workshop pages and 0 pages that answered a question.

02DIAGNOSIS

What the audit picked up on the strip.

CONTENT

0

pages that answered a question

Twenty pages in all: 14 fragrance pages written as mood copy and 6 workshop dates. Not one page explained a note, a concentration or how long a scent lasts, so a question had nowhere to land.

Finding 1 of 4

SEARCH RESULTS

87%

of the top ten held by databases and forums

Across 1,900 queries on notes, longevity and ‘perfume like X’, databases and forums held 87% of the top ten. Brand sites held 4%, mostly the glossaries of the big houses.

Finding 2 of 4

AI ANSWERS

3 / 120

tracked prompts whose answer cited Alabastra

We tracked 120 prompts in AI assistants and AI search, from ‘does iris perfume last’ to ‘perfume workshop in Provence’. Alabastra appeared in 3, each time as a shop link, never as the source.

Finding 3 of 4

EXPERTISE

22

years of the house perfumer's craft, nowhere on the site

No author pages, no named perfumer, no Organization or Person schema, and the atelier's own blotter tests kept in a paper notebook. The expertise that should anchor every answer was invisible to search.

Finding 4 of 4

03THE STRATEGY

A content plan composed like a perfume.

Built on our Content Engine plan: research, writing and AI-search work for brands whose expertise should be the answer.

  1. 01Query & prompt research

    1,900 queries · 120 prompts

    Every question people ask about notes, concentrations, longevity and ‘smells like’, clustered by intent, plus 120 AI prompts tracked every month. Each cluster was given its layer of the pyramid before a word was written.

  2. 02Content pyramid plan

    30 · 22 · 12

    Quick answers as top notes: short, precise, published fast. Guides as heart notes: the body of the plan. Deep expertise as base notes: few, slow to make, built to last. Trend pieces capped at a test of six.

  3. 03Expert-led guides

    12 base-note pieces

    Every base-note piece written with the house perfumer: a named author, atelier photographs described in alt text, the house's own blotter and skin longevity tests on all 14 fragrances, and sources cited.

  4. 04AI-answer optimisation

    120 prompts tracked

    Answer-first paragraphs, definitions an assistant can quote whole, FAQ markup where it fits, Organization, Person and Product schema, and links from every answer up to its guide and down to the fragrance.

04THE PYRAMID

Top notes catch the nose. Base notes stay on the skin.

In perfume, top notes are what you smell first and lose first, heart notes are the character, and base notes are what is still there the next morning. The content plan follows the same structure, so every quick answer leads down to a guide and every guide rests on real expertise.

30

TOP NOTES · 30 pieces

Quick answers

Short, exact answers to one question each: what eau de parfum means, how many sprays, what a heart note is. Quick to win, quick to fade.

Example piece: Eau de parfum or eau de toilette: what actually changes?

22

HEART NOTES · 22 pieces

Guides

How to make a scent last, how to choose by note family, ‘if you love this accord, try…’ guides built on materials, never on other brands' names.

Example piece: How to make perfume last longer on skin

12

BASE NOTES · 12 pieces

Deep expertise

Pieces only a Grasse perfumer could write: why orris takes three years, the May rose and jasmine harvests, the house's own longevity tests. Slow to make, still read a year later.

Example piece: Orris: why iris root rests three years before it becomes perfume

6

TREND NOTES · 6 pieces

Trend pieces

Seasonal ‘scents of the summer’ pieces: a burst of traffic, then gone. Kept to a test of six.

05THE BLOTTER TEST

Dip a strip into each kind of content. Watch it dry for a year.

Perfumers judge a scent on a paper blotter over hours. We judged content the same way, over months: one typical piece of each type, from publishing to month 12. The colour left on each strip is its traffic, as a share of its launch month.

Choose a strip and dip it, or dip all four, then let them dry.

Base notes

Deep expertise

Not dipped yet

The dry-down: a deep piece starts quietly and keeps deepening. A year later it brings twice its launch traffic and is what AI answers quote most.

Traffic of a typical piece · launch month = 100
0100200036912
  • Trend pieces
  • Quick answers
  • Guides
  • Deep expertise

06MONTH BY MONTH

18 months, from first sniff to dry-down.

  1. Months 1–2Smell the market

    1,900 queries and 120 AI prompts mapped, a baseline of who ranks and who gets cited, and long interviews with the perfumer and the workshop team.

  2. Months 3–4Write the formula

    64 briefs placed on the pyramid, author pages and schema built, and a longevity test designed for all 14 fragrances, on blotter and on skin.

  3. Months 5–10Top and heart notes

    30 quick answers and 22 guides published in weekly batches, plus six trend pieces as a test; every answer linked to its guide and its fragrance.

  4. Months 8–16Base notes

    12 deep pieces written with the perfumer, one every three weeks, each with test data or atelier process that no database has.

  5. Months 15–18Dry-down

    Pieces that had drifted refreshed, four trend pieces folded into guides, and workshops linked from the base-note pieces that sent the most bookings.

07RESULTS

Eighteen months on, the base notes are still on the strip.

Non-brand organic sessions a month
+340%
AI answers citing the house (of 120 tracked prompts)
3 → 41
Workshop bookings from search a month
11 → 46
Month-18 traffic from evergreen guides
62%
NON-BRAND ORGANIC SESSIONS A MONTH2,600 → 11,440 sessions a month
04,0008,00012,000M1M6M12M18
AI ANSWERS CITING ALABASTRA, OF 120 TRACKED PROMPTS

Month 1: 3Month 18: 41

WHERE MONTH-18 TRAFFIC COMES FROM
  • Deep expertise4,180
  • Guides2,860
  • Fragrance & workshop pages3,020
  • Quick answers1,240
  • Trend pieces140

62% comes from guides and deep expertise pieces, the layers that last.

WORKSHOP BOOKINGS FROM SEARCH11 → 46 a month+318%

Questions the house now answers

QUERY, AS SEARCHEDLAYERBEFOREMONTH 18AI ANSWERS
eau de parfum vs eau de toiletteTop notes100+4Cited
what are base notes in perfumeTop notes100+2Cited
how to make perfume last longer on skinHeart notes713Cited
perfume like a vintage chypreHeart notes100+5Not yet
parfum iris qui tient longtempsHeart notes482Cited
orris butter vs iris absoluteBase notes100+1Cited
récolte du jasmin à grasseBase notes633Cited
atelier parfum grasseBase notes142Not yet

Modelled scenario for a fictional business — not a client result. Rankings, traffic and AI citations are never guaranteed; see our Terms.

Get a plan like this

Your expertise should be the answer, not someone else's database. Start with the Content Engine plan, or tell us what your customers keep asking.

← All case studies