Skip to content

Omnishown Blackbox · Keyword intelligence

A keyword layer built to explain, not just to list.

Blackbox keeps its own keyword index: phrases and buyer questions with their market, intent, topic and source. It uses that index to match your pages to real demand and rank what to do first.

  • Every record keeps its source and date
  • Grouped into topics, not a flat list
  • Mapped to the page that should answer it
What to do firstExample data
  • same-day custom cake in GachibowliTransactional · maps to /menu86
  • eggless birthday cake near meTransactional · maps to New page78
  • how far ahead should I order a wedding cakeInformational · maps to New FAQ64
  • midnight cake delivery HyderabadTransactional · maps to /delivery59
Score = relevance 30, demand 25, feasibility 20, business impact 15, content gap 10. A guide to priority, not a ranking promise.

Three kinds of record, kept apart.

A phrase, a local version of it and a measurement taken on a given date are different things. Blackbox stores them separately so they are never confused.

The phrase

The query as typed or asked, its normalised form, its language and its search intent.

The market

The country, region or city the phrase applies to, because demand in one city says little about another.

The observation

A metric, such as estimated volume, recorded with where it came from and when. New observations are added, not overwritten.

Inside a record

What Blackbox knows about each phrase.

Each record can hold the phrase, its language and location, estimated monthly demand, estimated difficulty, cost per click when available, search intent (informational, commercial, transactional or navigational), search result features when observed, the topic and parent phrase it belongs to, and its source and last update.

Fields with no reliable source stay empty. A blank is more useful than a number nobody can check.

  • Phrase, language and location
  • Volume, difficulty and CPC (estimated, sourced, dated)
  • Intent and topic cluster
  • Source and last updated
omnishown-research.xlsxExample data
KeywordVolumeIntentRecommended URLScorePriority
same-day custom cakeest. 1,300Transactional/menu86High
eggless birthday cakeest. 880TransactionalNew page78High
wedding cake lead timeest. 320InformationalNew FAQ64Medium
midnight cake deliveryest. 590Transactional/delivery59Medium
Volumes carry their source and date in the real export. Here they are made-up estimates.
How it is used

From your site to a ranked shortlist.

Blackbox extracts the services, products and places on your site, generates candidate questions, retrieves matching records, filters them by market, language, intent and relevance, groups them into topics, maps them to your pages and flags gaps.

Each opportunity then gets an Opportunity Score from 0 to 100, built from relevance, demand, feasibility, business impact and content gap. Omnishown Research shows the full method and the weights. The score is a guide to what to do first, not a ranking forecast.

  • Retrieve, filter and group
  • Map to existing pages
  • Find gaps
  • Score and explain
What to do firstExample data
  • same-day custom cake in GachibowliTransactional · maps to /menu86
  • eggless birthday cake near meTransactional · maps to New page78
  • how far ahead should I order a wedding cakeInformational · maps to New FAQ64
  • midnight cake delivery HyderabadTransactional · maps to /delivery59
Score = relevance 30, demand 25, feasibility 20, business impact 15, content gap 10. A guide to priority, not a ranking promise.
How the AI is used

Retrieval first, then specialised scoring.

Blackbox does not need a new foundation model. It combines fixed audit rules, retrieval from its own keyword index and your audit findings, and a language model that reasons over what was retrieved.

The plan is to start with retrieval, then tune the ranking and classification with properly authorised outcome data, then improve how recommendations are chosen using measured feedback from real use.

  • Phase 1: retrieval over trusted data
  • Phase 2: tuned scoring and classification
  • Phase 3: improvement from measured feedback

application/ld+json

Facts a machine can read

Business
Sweet Crumbs, custom cakes
Area
Gachibowli, Hyderabad
Hours
Daily, 9:00 to 21:00
Ordering
Same-day on request

Example business, not a real customer.

What the keyword layer does not claim.

  • Volumes, difficulty and cost per click are estimates. Each one says where it came from and when.
  • A current ranking page is shown only when that data is actually available.
  • The keyword data sources and their licensing are being finalised for the pilot.

Questions, answered plainly

Still unsure? Join the pilot and ask us directly. We'll answer before you connect anything.

Where does the keyword data come from?

The intended sources are licensed keyword datasets, authorised keyword or advertising APIs, Search Console data you choose to connect, permitted public search observations, and your own site's content. The final sources and licensing are being decided for the pilot. Every figure will show its source and date.

How is this different from Omnishown Research?

Research is the product you use: the ranked list and the Excel export. The keyword layer in Blackbox is the index and the matching behind it.

Does it support my country and language?

The first target market and language are being decided with the pilot. Tell us yours when you join and we will say honestly whether it is covered.

Will it tell me exactly how many people search for a phrase?

No tool can know that exactly. Blackbox shows an estimate with its source and date, and leaves the field blank when it has no reliable source.

Find out which questions matter for your business.

Join the pilot and tell us your market. We will say honestly what the keyword layer covers today.

  • No card, no commitment
  • Every fix arrives as a pull request you approve
  • Planned pricing from ₹2,999 per site per month
I am