Skip to content
Get Ranked On TopIllustrative case studyContent Engine plan

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

QUIPU DATA · Retail analytics SaaS, 60 staff · Lima

How a Lima analytics SaaS became the answer AI gives retailers, one topic cluster at a time

+375%

organic trial sign-ups, from 40 to 190 a month

Lima, Peru · 18 months · Technology & SaaS

THE RESULTS, KNOTTED THE INCA WAY
×100×10×11907121264

How to read it: on a quipu, the place on the cord is the decimal place. Single knots high up count hundreds, lower down tens; at the end, a long knot counts units by its turns and a figure-eight knot means one.

THE CHALLENGE

The questions were being answered. Just not by Quipu.

Retailers asking how to measure basket size or which analytics suit a small chain got answers citing competitors. Quipu’s 143 blog posts formed no clusters.

A shop owner in Arequipa asks how to measure basket size; an operations lead in Bogotá asks which analytics suit a chain of twelve stores. Those are the moments just before a trial. Before the work, the answers came from competitors’ blogs, and AI assistants summarised them with competitors’ links.

Quipu’s blog had plenty of posts, but they hung loose: no topic owned, no definition to cite, no page that answered the question first. The job was to tie them into cords that a search engine, and an AI model, could read.

THE TANGLE · MONTH 0

  1. 143 posts, no clusters

    143 posts over seven years, filed by date rather than topic. No pillar page was linked from more than three of them.

  2. 41 posts fighting each other

    41 posts competed for the same nine terms, so Google rotated between them and none held page one.

  3. 9/120 AI prompts cited Quipu

    Of 120 evaluation prompts we tracked in AI assistants, 9 cited Quipu. Competitors were cited in 88.

  4. 0 definition pages

    No page defined basket size, sell-through or stock cover: the very metrics retailers ask about and AI answers quote.

  5. 62% of posts in English only

    Three in four target searches were in Spanish, but most posts existed only in English.

  6. 1.2 internal links per post

    Mostly to the pricing page. Readers and crawlers alike hit dead ends.

THE STRATEGY

Six cords, one main cord, every knot in its place.

Built on our Content Engine plan: content and AI-search work where every piece has a cluster, a format and a question it answers first.

  1. TOPIC CLUSTERS

    Six cords hung from one main cord

    Every post was mapped against 2,300 evaluation queries. 41 were merged into stronger pieces with 301 redirects and 29 rewritten into the clusters; each cluster got a pillar guide that links to every piece on its cord, and back.

    6 clusters · 64 pieces

  2. GLOSSARY PAGES

    One single knot per definition

    30 glossary entries, each opening with a one-sentence definition, then the formula, a worked example in soles and the mistake people usually make. Short, quotable and marked up as DefinedTerm.

    30 entries

  3. COMPARISON PAGES

    Figure-eight knots: two options, one page

    12 comparisons for the moment a retailer is choosing: spreadsheet or dashboard, build or buy, tool categories side by side. Each one says plainly when Quipu is not the right fit.

    12 comparisons

  4. AI-ANSWER CITATIONS

    Answers an assistant can quote

    Answer-first summaries, tables of real figures from Retail Pulse, Quipu’s anonymised benchmark of 400 stores, and named authors from the data team. The same 120 prompts were re-run every month to see which pieces got cited.

    9 → 71 of 120 prompts

TIE THE KNOTS

Each cord a topic, each knot a piece of content.

Scrub through the eighteen months or tie one knot at a time. The knot shows the format; a cord tied off shows the AI citations and sign-ups it now brings in. Modelled data.

1614121199
Pieces tied
64 of 64
AI prompts citing Quipu
71 of 120
Trial sign-ups a month
190

Main cord · all six clusters

Basket metrics

Basket size, average ticket, units per transaction: the numbers every store manager is asked for and few can define.

Tied off in month 12 · Feb

Knots on this cord

  • Guides · long knot5 of 5
  • Glossary entries · single knot7 of 7
  • Comparisons · figure-eight knot1 of 1

Searched as

  • cómo medir el ticket promedio
  • how to measure basket size

Cord tied off: this is what it now brings in.

Queries in Google’s top 10
52
AI prompts citing Quipu
16 of 20
Trial sign-ups a month
38

THE RECORD

Eighteen months, knot by knot.

  1. MONTHS 1–3

    Mar – May

    Unpicking the tangle

    143 posts audited, 2,300 evaluation queries mapped into six clusters and 120 AI prompts set up for monthly tracking. 41 overlapping posts merged, with 301s.

    Sign-ups / month
    44
    AI citations
    10 of 120
  2. MONTHS 4–6

    Jun – Aug

    The first cords hang

    Pillar guides for basket metrics and small chains, the first glossary entries, and Retail Pulse: anonymised, aggregated figures from 400 stores that agreed to share them.

    Sign-ups / month
    61
    AI citations
    17 of 120
  3. MONTHS 7–9

    Sept – Nov

    Definitions people quote

    A steady rhythm of two glossary entries a week, DefinedTerm markup, and the first AI answers citing Quipu on basket questions.

    Sign-ups / month
    88
    AI citations
    28 of 120
  4. MONTHS 10–12

    Dec – Feb

    Honest comparisons

    Spreadsheet-or-dashboard and build-or-buy pages go live. The basket cord is tied off in month 12.

    Sign-ups / month
    137
    AI citations
    48 of 120
  5. MONTHS 13–15

    Mar – May

    Spanish first

    New pieces written in Spanish first and adapted into English, with hreflang between them. The small-chains and inventory cords are tied off.

    Sign-ups / month
    177
    AI citations
    66 of 120
  6. MONTHS 16–18

    Jun – Aug

    Tie off and count

    The last cords are tied. Each trial sign-up is attributed to its landing page in the product: 190 a month from organic search.

    Sign-ups / month
    190
    AI citations
    71 of 120

READ THE CORDS

Eighteen months on, the knots add up.

+375%

Organic trial sign-ups a month
50100150200BeforeMonth 6Month 12Month 18

9 → 71

Tracked AI prompts citing Quipu (of 120)
BeforeMonth 18

14 → 212

Evaluation queries in Google’s top 10
Basket52Chains44Stock36Stores31Tools27Promos22BeforeMonth 18

64

Pieces knotted into six clusters
Guides · 22Glossary entries · 30Comparisons · 12

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

CORD BY CORD · EVALUATION QUERIES

Google positions for evaluation queries before and after the work
Query, as searchedCordMonthly searchesMonth 0Month 18Change
cómo medir el ticket promedioBasket1,300412Up 39 places
how to measure basket sizeBasket590573Up 54 places
retail analytics for small chainsChains260721Up 71 places
analítica para cadenas pequeñasChains320882Up 86 places
cómo calcular la rotación de inventarioStock2,400344Up 30 places
sell-through rate formulaStock1,600616Up 55 places
ventas por metro cuadradoStores1,000293Up 26 places
footfall conversion rateStores480665Up 61 places
retail BI tools comparisonTools390957Up 88 places
hoja de cálculo o software de analítica retailTools210483Up 45 places
cómo medir el impacto de una promociónPromos720525Up 47 places
promo lift calculationPromos210808Up 72 places

TIE YOUR OWN CORDS

Get a plan like this

Topic clusters built around the questions your buyers really ask, written to be found in search and quoted in AI answers.

← Back to all case studies