FR

Features

Everything Kaiku measures,and how.

The full list of shipped features, laid out the way the app is: five destinations, each with its tabs. For every feature: what it does, where its data comes from, and a glimpse of its screen.

Start for freeWhat each plan includes

The visuals on this page are illustrative examples: their figures show what the screens look like, not customer results.

How it works

Measure, understand, act, prove.

A one-week loop that starts again every Sunday. Each step runs on real data; when a figure is missing, the interface says so.

  1. 1

    Measure

    Every Sunday, Kaiku asks ChatGPT, Gemini, Claude and Perplexity your prompts, records Google's AI Overview, your rankings and your external presence, and reads Search Console, GA4 and Bing. On Sundays, on the Pro and Enterprise plans, ChatGPT, Gemini and Claude search the web before they answer, just like their consumer versions; each engine is asked each prompt once.

    GPTGemClaPerAIO
  2. 2

    Understand

    Two separate scores, SEO and GEO, the SEO×GEO Matrix, your rank inside AI answers and an AI analysis of each period tell you where you stand, and why.

    SEOGEO
  3. 3

    Act

    This week's actions are ranked by estimated impact, with a step-by-step guide and the tools that do the work: briefs, schemas written to WordPress, IndexNow.

    123
  4. 4

    Prove

    When an action is published, the score component it targets is frozen; at the next scan, at least 7 days later, it is read again. Impact proof and reports show the before and after.

01 · Home

Home

Where the project stands this week and where to start — on one page, plus the tools that follow you everywhere in the app.

Period indicators, with their trend

Illustrative example

In the app: Home

Home

What it does
A summary of the selected project: SEO Score and GEO Score, the action of the week, visibility by AI engine, your competitors and your traffic.
How it works
The tiles read the latest Sunday scan plus Search Console, GA4 and Bing over the chosen period. Each one draws its trend once it has two readings, and an engine that failed is flagged rather than counted as zero.
Sunday AI scanSearch ConsoleGoogle Analytics 4Bing WebmasterDataForSEO

Two separate scores, never combined

Illustrative example

In the app: Home

SEO Score and GEO Score

What it does
Two scores from 0 to 100: the SEO Score reflects what Google and Bing show of you, the GEO Score what AI answer engines show of you. They are never added together.
How it works
The GEO Score blends presence in answers, share of voice, rank inside the answer, pages cited as sources and AI Overviews; the SEO Score blends Google rankings, the Search Console trend, coverage, click-through rate and Bing rankings. “Where the score comes from” breaks down each component.
Sunday AI scanSearch ConsoleDataForSEOBing Webmaster

What is left before everything is measured

Illustrative example

In the app: Home

Getting started

What it does
The list of what is left before Kaiku can measure everything — keywords, competitors, connections, first scan — and what each step unlocks.
How it works
It reads the project's setup and folds into a single line once everything is in place. When you create a project, “Run my first check” asks two engines about a few keywords so you can see within minutes whether you are cited, ahead of the full Sunday scan.
Project data

Clicks per week, against the compared period

Illustrative example

In the app: Home

Period bar and history

What it does
Pick 7 days, 28 days, 3 months, 12 months or exact dates, and compare with the previous period or the same period last year.
How it works
Kaiku keeps a dated copy of your data at every scan and deletes nothing: the longer a project is tracked, the richer the comparisons. History depth depends on the plan; Search Console stays capped at 16 months by Google.
Project data

A cross-reading of the page's figures

Illustrative example

In the app: Home

AI analysis of the period

What it does
On statistics pages, an AI analysis tab sits next to the Data tab: it reads the figures of the period on screen and connects them.
How it works
Claude receives the page's figures and returns a short summary, then strengths, points to watch and actions, written in the project's language.
Claude analysis

⌘K palette: find a page, even by its former name

Illustrative example

In the app: Home

All tools, ⌘K palette and Help Center

What it does
The All tools page lists every page in the app, filterable by destination; the ⌘K palette finds a page by its name, even its former name.
How it works
The Help Center opens as a side panel on the article for the tab you are on, in English or French. On a phone, a five-icon bar replaces the menu.
Project data

02 · Get cited by AI

Get cited by AI

ChatGPT, Gemini, Claude, Perplexity and Google's AI: who cites you, who cites the others, which sources they lean on, and what makes a page citable.

Share of answers citing the brand, by engine

Illustrative example

In the app: Get cited by AI › Overview

Overview · GEO Score

What it does
The GEO Score and what moved it, the period's indicators, and one bar per AI engine: the share of answers in the latest scan that cite your brand.
How it works
Every Sunday, Kaiku asks ChatGPT, Gemini, Claude and Perplexity your prompts and records Google's AI Overview on your keywords. A tick marks the previous scan; an engine that did not answer gets no bar.
Sunday AI scan

Google AI Overview: where the figure comes from

Illustrative example

In the app: Get cited by AI › AI Citations

AI Citations

What it does
Who gets cited on your SEO keywords in Google's AI Overview, ChatGPT and Perplexity, and through which sources.
How it works
For Google, the page reads the actual results page and the sources of its AI summary; for ChatGPT and Perplexity, the question is put to the engine. A funnel shows how you get from keywords analyzed to keywords where your brand is cited.
Live queryDataForSEO

This scan, prompt by prompt

Illustrative example

In the app: Get cited by AI › AI Responses

AI Responses

What it does
Read word for word what each AI answered to each of your prompts, with your rank in the answer (“2nd of 3”) and the citation order.
How it works
Full answers from every scan are archived. A prompt-by-engine grid shows where you are cited; a failed request is kept apart from an absence.
Sunday AI scan

Brand position in the answer, scan after scan

Illustrative example

In the app: Get cited by AI › AI Responses

Response history

What it does
Follow how answers to the same prompt change, scan after scan: a competitor moving ahead, a mention appearing.
How it works
An engine-by-scan grid built from the archived answers; the number in a cell is your brand's position in the answer.
Sunday AI scan

Most cited domains in AI answers

Illustrative example

In the app: Get cited by AI › Sources & authority

Sources / Citations

What it does
The domains and URLs that AI engines cite as sources in your market, and where your site sits among them.
How it works
Kaiku collects the links given in your scans' answers and counts them by domain and by engine. The detailed table is one click away.
Sunday AI scan

Your pages against the other cited sources

Illustrative example

In the app: Get cited by AI › Sources & authority

Citation URLs

What it does
A page-level view of citations: your URLs' share of all cited URLs, by named entity and by engine.
How it works
Same material as Sources, grouped by page rather than domain, with its change from one scan to the next.
Sunday AI scan

Where citations come from, by source type

Illustrative example

In the app: Get cited by AI › Sources & authority

Source Authority

What it does
What models lean on when they talk about your sector: Wikipedia, B2B reviews, forums, media, video, documentation.
How it works
Every URL cited in your scans is grouped by root domain, then sorted by authority type. You see which sources are worth pursuing.
Sunday AI scan

You against competitors, scan by scan

Illustrative example

In the app: Get cited by AI › Mentions & sentiment

AI Mentions

What it does
A log of every appearance of your brand or a competitor in AI answers, filterable by engine, tone and period.
How it works
Each scan's answers are parsed to spot named entities and sort them: your brand, a competitor, or other.
Sunday AI scan

Sites that mention you — in violet, those AI already cites

Illustrative example

In the app: Get cited by AI › Mentions & sentiment

Web Mentions

What it does
Who mentions you on the web, and which of those sites AI already cites: your next PR targets, in order.
How it works
Mentions come from DataForSEO; Kaiku cross-checks them with the sources found in your scans. Tone is shown only when the data exists.
DataForSEOSunday AI scan

Tone of mentions over time

Illustrative example

In the app: Get cited by AI › Mentions & sentiment

Sentiment

What it does
The tone of your mentions — positive, neutral or negative — over time and by AI engine.
How it works
Each mention is classified from its sentence and paragraph in the answer. Net sentiment sums up the gap between positive and negative.
Sunday AI scanClaude analysis

Who is named before you: average rank

Illustrative example

In the app: Get cited by AI › Mentions & sentiment

Share of voice

What it does
Your share of mentions against competitors, plus “Who is named before you”: average rank, answers where you come first, change since the previous scan.
How it works
Computed on scan answers. Every brand is clickable and opens its split by engine, its tone, its prompts and its cited URLs.
Sunday AI scan

Share of answers citing you, by engine

Illustrative example

In the app: Get cited by AI › Score by engine

Score by engine

What it does
Your performance engine by engine, because the same site can be strong on Perplexity and missing from ChatGPT.
How it works
One bar per engine and a scan-by-scan curve from 0 to 100%, with pointers specific to each engine. On Sundays, on the Pro and Enterprise plans, ChatGPT, Gemini and Claude search the web before they answer, just like their consumer versions; each engine is asked each prompt once.
Sunday AI scan

Coverage sorted by identity anchor

Illustrative example

In the app: Get cited by AI › External presence

Monitoring

What it does
Save the expressions to watch and Kaiku brings back whatever cites them: articles, posts and public discussions.
How it works
Expressions that can only mean your client become identity anchors: every result is confirmed, to check or dismissed, and only confirmed ones count.
Google NewsDataForSEOHacker News and Bluesky

Your brand as an entity

Illustrative example

In the app: Get cited by AI › External presence

Entity & Brand

What it does
Does your brand exist as an entity in AI's eyes? Kaiku checks Wikidata, Wikipedia, your LinkedIn page and your site's schema.
How it works
Each source found adds points to an entity score, with the steps to take for the missing ones.
Wikidata and WikipediaReads your site

Articles kept per month

Illustrative example

In the app: Get cited by AI › External presence

Press

What it does
Articles that mention your client: headline, outlet, date and excerpt, grouped by type of media.
How it works
Articles come from Google News, in the edition of the project's market, filtered by your monitoring expressions and identity anchors.
Google News

Your best video, keyword by keyword

Illustrative example

In the app: Get cited by AI › External presence

YouTube

What it does
Who owns YouTube on your keywords, and whether it is you. AI engines often cite videos as sources.
How it works
For your first SEO keywords, Kaiku reads the top videos through DataForSEO, in the project's market, and looks for your brand in the title and channel name.
DataForSEO

Where the conversations are

Illustrative example

In the app: Get cited by AI › External presence

Reddit

What it does
Reddit threads that mention your brand, and threads on your topics where you could take part.
How it works
Threads are found in Google search results through DataForSEO, with the most present subreddits and a Refresh button.
DataForSEO

Mentions per month

Illustrative example

In the app: Get cited by AI › External presence

Social buzz

What it does
Your mentions on Hacker News and Bluesky, two open public communities that AI reads.
How it works
Both platforms are queried through their public interfaces, with no account to connect: threads, posts, reactions and dates.
Hacker News and Bluesky

Your visibility against the market, as % of the usual level

Illustrative example

In the app: Get cited by AI › External presence

Sector benchmark

What it does
What normal looks like in your sector: demand on the sector's non-brand queries, and your visibility against that market.
How it works
Demand comes from the sector's Google search volumes, your visibility from Search Console. The sector is inferred from your keywords, or chosen in settings.
DataForSEOSearch Console

One tile per source, with its curve

Illustrative example

In the app: Get cited by AI › External presence

History

What it does
What your external sources said about you months ago: press, entity, Wikipedia, web mentions, Reddit, YouTube and social buzz.
How it works
Every External presence reading is dated and kept. One tile per source shows the latest value, the change and its curve.
Google NewsDataForSEOWikidata and Wikipedia

Your keywords by quadrant

Illustrative example

In the app: Get cited by AI › SEO×GEO Matrix

SEO×GEO Matrix

What it does
The heart of Kaiku: every keyword placed by its Google performance and its AI visibility, in four quadrants — Urgent, Maintain, Watch, Ignore.
How it works
Your rank inside AI answers is added keyword by keyword and engine by engine (“2nd of 4”, “cited alone”, “not cited”); what moved between two readings is highlighted.
Sunday AI scanSearch ConsoleDataForSEO

Weighted GEO score on 5 axes

Illustrative example

In the app: Get cited by AI › Citability & content

GEO Content Audit

What it does
How likely a page is to be cited by AI, with a weighted score on five axes: citability, readability, authority, technical access, multi-modality.
How it works
Kaiku reads the page's HTML and measures concrete signals: self-contained answer blocks, a direct answer at the top of each section, structure, author, freshness, data.
Reads your site

What makes your pages citable (out of 25)

Illustrative example

In the app: Get cited by AI › Citability & content

Citability

What it does
Which pages of your site are the most citable, and which criterion holds you back.
How it works
Four gauges out of 25 combine organic traffic (GA4), Google clicks (Search Console), content type and conversions; the lowest one is your first lever.
Search ConsoleGoogle Analytics 4Reads your site

Completeness of each schema found

Illustrative example

In the app: Get cited by AI › Citability & content

Schema checker

What it does
Finds and validates a page's JSON-LD structured data, then suggests the schemas it is missing.
How it works
Kaiku reads the live page, checks syntax, required fields and schema.org types, and hands you the full code for each suggested schema.
Reads your site

Dated capture of an engine's answer

Illustrative example

In the app: Get cited by AI › Citability & content

Snippet Preview

What it does
A dated capture of what Perplexity, Google's AI Overview and Gemini answer to a prompt: visual proof for a report or a meeting.
How it works
Kaiku opens each engine's answer page itself and takes a screenshot, stored in a library.
Live query

Pages carrying the traffic, and gaps

Illustrative example

In the app: Get cited by AI › Citability & content

Content analysis

What it does
The pages that carry the traffic, the position of their main keyword, and the gaps: topics your competitors cover and you do not.
How it works
Domain and keyword data come from DataForSEO; a gap with no page on your side is flagged.
DataForSEO

03 · Rank on Google and Bing

Rank on Google and Bing

Classic search, on your real data: rankings, traffic, keywords, opportunities, competition, local, app stores and technical health.

Tracked keywords by position bracket

Illustrative example

In the app: Rank on Google and Bing › Overview

Overview · SEO Score

What it does
The SEO Score, your real clicks and your rankings for the period, with tracked keywords by position bracket against your competitors.
How it works
Clicks come from Search Console and Bing, never from estimates; rankings from weekly tracking through DataForSEO.
DataForSEOSearch ConsoleBing Webmaster

Average position, you and a competitor (1 = top)

Illustrative example

In the app: Rank on Google and Bing › Positions

Positions

What it does
Your Google position on every keyword, desktop or mobile, with the ranking URL, moves in and out of the top 10, and your competitors.
How it works
Rankings are collected weekly through DataForSEO, in each tracking list's market; depending on the plan, several lists can follow different countries and languages.
DataForSEO

Clicks day by day, against the compared period

Illustrative example

In the app: Rank on Google and Bing › Analytics

Google Search Console

What it does
Clicks, impressions, click-through rate and position, against the period of your choice, with winning queries, declining ones and opportunities.
How it works
Read-only Google connection. Google keeps 16 months of Search Console data; Kaiku archives it from the moment you connect and deletes nothing.
From StarterSearch Console

From your queries to clicks to defend

Illustrative example

In the app: Rank on Google and Bing › Analytics

Google and AI

What it does
Your clicks against Google's AI answers: protected where the AI Overview cites you, at risk where it cites other sites, with the sites cited in your place.
How it works
Kaiku checks the AI Overview on your Search Console queries that get clicks. Search Console's “Generative AI” report, which Google does not expose in its API, is imported by hand from the export.
Search ConsoleSearch Console exportDataForSEO

Sessions from AI, per week and per assistant

Illustrative example

In the app: Rank on Google and Bing › Analytics

Google Analytics 4

What it does
Organic sessions, conversions and landing pages, plus traffic from AI: sessions per week, per assistant, and what it brings in.
How it works
Read-only, through the same Google connection as Search Console. The Traffic tab isolates organic search; a dedicated tab covers traffic from AI assistants.
From ProGoogle Analytics 4

Bing against Google: queries only Bing brings you

Illustrative example

In the app: Rank on Google and Bing › Analytics

Bing Webmaster

What it does
Queries, clicks, impressions and positions on Bing, which powers Microsoft Copilot, including the queries only Bing brings you.
How it works
Connected with an encrypted Bing Webmaster API key. Bing only dates daily traffic; its queries are a rolling total of about six months, and the page says so.
From StarterBing Webmaster

Estimated organic traffic by domain (DataForSEO estimate)

Illustrative example

In the app: Rank on Google and Bing › Analytics

Traffic

What it does
Your organic traffic rank against the compared domains: share of traffic, gap to the leader, ranking keywords.
How it works
This is a DataForSEO estimate and is labeled as one: your real visits stay in GA4 and Search Console.
DataForSEO

Where the good targets are: volume × difficulty

Illustrative example

In the app: Rank on Google and Bing › Keywords

Keyword Magic

What it does
From one seed term, hundreds of variants and long-tail queries, grouped by intent and added to the project in one click.
How it works
Suggestions come from DataForSEO with volume and difficulty; a volume × difficulty scatter brings out the reachable targets.
DataForSEO

Cumulative monthly volume per group

Illustrative example

In the app: Rank on Google and Bing › Keywords

Clusters

What it does
Your keywords grouped by topic: each group is a candidate pillar page or article.
How it works
Grouping compares what keywords mean; each group shows its total volume, your Google positions and their difficulty.
DataForSEO

Shared URLs between two keywords' top 10

Illustrative example

In the app: Rank on Google and Bing › Keywords

SERP clustering

What it does
Groups keywords by what Google actually ranks: two queries that surface the same pages can be served by one page.
How it works
Kaiku reads each keyword's Google top 10 through DataForSEO and counts shared URLs pair by pair: many means one page, several means one group, few means neighboring topics.
DataForSEO

AI against Google on the largest volumes

Illustrative example

In the app: Rank on Google and Bing › Keywords

AI Volume

What it does
How often your keywords are asked of conversational engines each month, against their Google volume.
How it works
A DataForSEO estimate (AI Keyword Data) on the project's SEO keywords; the source is stated on the page for your reports.
DataForSEO

Authority of the top 10 results — in violet, the winnable spots

Illustrative example

In the app: Rank on Google and Bing › Keywords

Keyword overview

What it does
Everything about one keyword: volume, difficulty, intent, SERP, and the authority of each top-10 result. Domains below the median are winnable spots.
How it works
Pick the country and index language; data comes from DataForSEO, and keywords you already opened reload from the cache.
DataForSEO

Search interest over 12 months

Illustrative example

In the app: Rank on Google and Bing › Keywords

Trends

What it does
Where demand is heading: your keywords sorted by trend (“breakout”, “growing”), AI demand against Google, and the last 12 months of curves.
How it works
Interest curves come from Google Trends, through DataForSEO.
DataForSEO

Who is ahead: your gaps where each competitor ranks

Illustrative example

In the app: Rank on Google and Bing › Opportunities

Gap analysis

What it does
Keywords your competitors rank for and you do not, and who is ahead on each.
How it works
Kaiku pulls your competitors' ranking keywords through DataForSEO and cross-checks them with your positions; sort by volume, or by difficulty for quick wins.
DataForSEO

Impressions split between your pages

Illustrative example

In the app: Rank on Google and Bing › Opportunities

Cannibalization

What it does
Queries where several of your pages compete with each other, ranked by severity.
How it works
Detected in your Search Console data: an even split of impressions between two pages means Google is hesitating.
Search Console

Coverage by intent: ranking page or page to create

Illustrative example

In the app: Rank on Google and Bing › Opportunities

Intents

What it does
Each keyword sorted by intent — informational, comparative, transactional, local, navigational — and the share served by a page on your site.
How it works
Wording and SERP analysis. An intent that is mostly “page to create” is one your site does not serve; send it to the Content queue.
DataForSEOSearch Console

Most frequent results-page features

Illustrative example

In the app: Rank on Google and Bing › Opportunities

SERP

What it does
What a keyword's Google results page is made of — AI Overview, local pack, “People also ask”, featured snippet, videos — and your organic position.
How it works
The real results page is read through DataForSEO; your position counts organic results only, so it matches Search Console.
DataForSEO

SERP consensus: format of the top 10 results

Illustrative example

In the app: Rank on Google and Bing › Opportunities

SXO analysis

What it does
Why a well-optimized page does not rank: its format does not match what Google rewards for that query.
How it works
Each top-10 result and your page are sorted into eight formats (service page, comparison, article, local page…). The SXO Gap Score measures the mismatch.
DataForSEOReads your site

Most relevant competitors (score out of 100)

Illustrative example

In the app: Rank on Google and Bing › Competition

Competitor discovery

What it does
Who your real competitors are, according to AI and Google: each one gets a score, “tracked” or “new, to add”.
How it works
The score keeps the strongest signal among shared keywords (DataForSEO), mentions in your AI scans and rankings.
Sunday AI scanDataForSEO

Against your sector, as % of the average

Illustrative example

In the app: Rank on Google and Bing › Competition

Benchmark

What it does
Your metrics against your sector's average: above on the right, below on the left.
How it works
Your scores, mentions, share of voice and rankings are compared with those of your tracked competitors.
Sunday AI scanDataForSEO

Two of your projects side by side

Illustrative example

In the app: Rank on Google and Bing › Competition

Comparison

What it does
Two of your projects side by side, line by line: handy for agencies comparing clients, or one project at two dates.
How it works
Built from the data Kaiku already holds for each project; mentions gained and lost between two dates are listed.
Sunday AI scanProject data

Domain's estimated traffic, month by month

Illustrative example

In the app: Rank on Google and Bing › Competition

Domain overview

What it does
A portrait of any domain, yours or a competitor's: ranking keywords, estimated traffic, top pages, links and trend.
How it works
DataForSEO data for the market shown; analyses can be saved to follow the trend month by month.
DataForSEO

Links gained and lost per month

Illustrative example

In the app: Rank on Google and Bing › Competition

Backlinks

What it does
Where your domain's authority comes from, what it gained and lost, what its anchors say, and the sources that cite your competitors but not you.
How it works
Computed on the project's main domain with DataForSEO's Backlinks API, authority on a 1,000 scale. A figure that was not measured reads “not available”, never 0.
DataForSEO

Profile views, against the compared period

Illustrative example

In the app: Rank on Google and Bing › Local

Google Business Profile

What it does
How your business profile performs — views, actions, the searches that showed it (brand, competitor, service, place) — compared from one period to another.
How it works
Read through the official Google Business Profile API, with your permission. Google publishes its statistics a few days late, so Kaiku stops at the last published day.
Google Business Profile

Rating distribution

Illustrative example

In the app: Rank on Google and Bing › Local

Reviews

What it does
Understand, prioritize and reply to your Google reviews without leaving Kaiku.
How it works
With the profile connected, every review is read through the official API and archived; without it, recent public reviews come from DataForSEO. An AI analysis draws out the themes.
Google Business ProfileDataForSEOClaude analysis

NAP consistency (name, address, phone)

Illustrative example

In the app: Rank on Google and Bing › Local

Directories

What it does
Whether your name, address and phone (NAP) match from one platform to the next, and your local keywords.
How it works
Kaiku compares your Google profile's details with those on local platforms and flags every variation.
Google Business ProfileDataForSEO

Your local-pack rank, point by point

Illustrative example

In the app: Rank on Google and Bing › Local

Local heatmap

What it does
Your local-pack rank point by point on a grid around your business, with your share of local voice (SoLV).
How it works
Each point is queried through DataForSEO, the grid centered on your Google profile by default. Grids are kept and compare from one date to another.
DataForSEOGoogle Business Profile

Nearby businesses in your category: number of reviews

Illustrative example

In the app: Rank on Google and Bing › Local

Local competitors

What it does
Nearby businesses in your category — number of reviews, rating, your rank — and the local pack on your tracked keywords.
How it works
Only businesses whose primary Google category matches yours are kept; the others stay visible separately so you can check the sorting.
DataForSEOGoogle Business Profile

Saved pages: cities × services covered

Illustrative example

In the app: Rank on Google and Bing › Local

Local pages

What it does
Pages per city and per service, generated with their title, meta description, sections and target keywords.
How it works
A cities × services grid shows which combinations are covered; red text flags a risk of truncation in Google.
Claude analysis

Average rank on your keywords (1 = top)

Illustrative example

In the app: Rank on Google and Bing › App Stores

Ranking

What it does
Your rankings on your keywords in the App Store and Google Play, with your listing, reviews and competitors, across several countries and languages.
How it works
Enter your app's identifier: rankings are collected per keyword and per country, with a curve once there are two readings.
From StarterApp Store and Google Play

Your listing against your competitors' average

Illustrative example

In the app: Rank on Google and Bing › App Stores

Listing & naming

What it does
Score your listing on both stores, measure the gap with your competitors, and pick a title that holds up.
How it works
The audit reads the stores' public listings on the spot; the naming adviser first reads the titles that dominate your keywords.
App Store and Google Play

Store queries: popularity and difficulty

Illustrative example

In the app: Rank on Google and Bing › App Stores

ASO keywords

What it does
The queries people actually type in the App Store, and tools to build and check your keyword field.
How it works
Suggestions come from the App Store's official autocomplete, by popularity; on Android, Google suggestions are a stated fallback.
App Store and Google Play

Several apps compared side by side

Illustrative example

In the app: Rank on Google and Bing › App Stores

App explorer

What it does
Analyze any published app — a competitor's or a prospect's — and compare several side by side.
How it works
Search for an app by name, or paste its App Store or Google Play address: the identifier is extracted automatically.
From ProApp Store and Google Play

Page GEO score, by category

Illustrative example

In the app: Rank on Google and Bing › Technical audit

Page audit

What it does
One URL checked against more than fifteen AI-citability criteria — structure, citability, AI-friendly content, authority — scored out of 100.
How it works
Kaiku fetches the page's full HTML and assesses each criterion. The URL's curve appears from its second audit, and the page suggests the keywords it should target.
Reads your site

Detailed criteria

Illustrative example

In the app: Rank on Google and Bing › Technical audit

SEO audit

What it does
A page's classic SEO checklist: title, meta description, headings, images, canonical, hreflang, HTTPS, robots.txt and sitemap.
How it works
The live page is read, scored by category with detailed criteria. Built to be used alongside the Page audit.
Reads your site

Bot-by-bot verdict (robots.txt)

Illustrative example

In the app: Rank on Google and Bing › Technical audit

AI Crawler Access

What it does
The GEO prerequisite: can AI crawlers (GPTBot, ClaudeBot, PerplexityBot, Google-Extended…) read your site?
How it works
Kaiku reads your robots.txt, checks the content readable without JavaScript, the XML sitemap and the llms.txt file, then gives a bot-by-bot verdict.
Reads your site

Field data (p75, mobile) against Google's thresholds

Illustrative example

In the app: Rank on Google and Bing › Technical audit

Core Web Vitals

What it does
LCP, INP, CLS, FCP and TTFB, on mobile and desktop, against Google's thresholds, with recommendations.
How it works
Field data comes from the Chrome UX Report (CrUX, 28 days, p75) — what your real visitors experience — alongside Google PageSpeed lab measurements.
CrUX and PageSpeed

Crawl health

Illustrative example

In the app: Rank on Google and Bing › Technical audit

Site crawl

What it does
A crawl of your site: pages with errors, non-indexable or orphan pages, click depth, tags and slow pages.
How it works
Kaiku walks your pages and records HTTP codes and load times; the page-by-page detail is one click away.
Reads your site

Changes found at each check

Illustrative example

In the app: Rank on Google and Bing › Technical audit

SEO drift

What it does
“Git for SEO”: spot what a deployment broke — a removed canonical, an added noindex, a vanished schema — before rankings drop.
How it works
Kaiku snapshots the SEO state of your key pages (the baseline), then compares it at every check using content fingerprints.
Reads your site

URLs submitted to Bing, per day

Illustrative example

In the app: Rank on Google and Bing › Technical audit

IndexNow

What it does
Tell Bing within seconds that a page was just published or updated; Copilot relies on Bing's index.
How it works
Generate the project key, place the file at the root of your site and have it verified: every submission is then logged.
IndexNow

04 · Act

Act

Turn what was measured into published pages, live schemas and fixes — then measure what each one changed.

Where your actions stand

Illustrative example

In the app: Act › Overview

Actions overview

What it does
What to do now, and did what was done work? The To do, In progress, Published, Verified cycle, your publishing pace and your priority actions.
How it works
Everything is read from the project's action log. A published action freezes the score component it targets; the following Sunday scan, at least 7 days later, reads it again.
Project data

How to fix: an action's guide

Illustrative example

In the app: Act › Actions

Actions of the week

What it does
The strongest action to do by estimated impact, then all your actions by status. One click opens “How to fix”: the problem, the targeted component, concrete steps and the Kaiku tools that do the work.
How it works
Actions come from your data: AI scans, Search Console, rank against a competitor, recommendations. A detailed AI-written plan is available on request, in the project's language.
Sunday AI scanSearch ConsoleDataForSEOClaude analysis

Remaining action plan: impact × effort

Illustrative example

In the app: Act › Actions

Recommendations

What it does
The project's action plan, ranked by priority, with an impact × effort grid where quick wins are framed.
How it works
Kaiku rereads everything the project has gathered — matrix, mentions, history, Search Console, rankings, GA4 — and recomputes after every scan. Your “done” and “not relevant” statuses are kept.
Sunday AI scanSearch ConsoleGoogle Analytics 4DataForSEO

Before and after each action, on equal windows

Illustrative example

In the app: Act › Impact proof

Impact proof

What it does
What each action really changed: the same measures over two windows of equal length, right before and right after its date.
How it works
A recommendation marked done, a schema confirmed live or an action added by hand becomes a dated marker. Pick 30, 60 or 90 days; a window that has not elapsed stays “pending”, never counted as an effect.
Search ConsoleGoogle Analytics 4Sunday AI scanProject data

Queue progress

Illustrative example

In the app: Act › Content queue

Content queue

What it does
Your editorial pipeline, from detected opportunity to written content: to do, brief created, written.
How it works
Opportunities are added in one click from Search Console (queries with no dedicated page), intent analysis or the page audit; the age of each item is tracked.
Search ConsoleProject data

This brief at a glance: optimization score

Illustrative example

In the app: Act › Briefs

Briefs

What it does
The content workshop: a quick SEO template, a full brief, and a review of the written text, with a score and the target length against the top 10.
How it works
The brief starts from the real Google top 10 (DataForSEO) and your GEO signals, then Claude writes it. The generated article goes to WordPress as a draft, never published without your review.
DataForSEOSunday AI scanClaude analysisWordPress

JSON-LD written into the page, then checked live

Illustrative example

In the app: Act › Schema Injector

Schema Injector

What it does
Build a JSON-LD schema (Organization, Article, FAQPage, HowTo, Product, LocalBusiness…) and place it on your WordPress site.
How it works
Kaiku writes the JSON-LD into the page through the WordPress API, using an encrypted application password, then checks that it is live; a vanished schema turns “to re-inject”.
WordPressReads your site

Suggested Google preview and lengths

Illustrative example

In the app: Act › Generators

Meta Generator

What it does
A title and meta description written to earn the click in Google and the citation in AI answers.
How it works
Claude writes them from the keyword and the page; the tool checks keyword presence and length against the truncation limit, with the suggested Google preview.
Claude analysisReads your site

Structure of the written article

Illustrative example

In the app: Act › Generators

GEO Content

What it does
A full article written to be cited: question headings, short answers, hard data, a closing FAQ and its schema.
How it works
Claude writes from a keyword and, if you like, a brief; the text downloads as Markdown, ready to review.
Claude analysis

Impressions × click-through rate: high-potential pages

Illustrative example

In the app: Act › Generators

Landing pages

What it does
The landing pages to work on first: many impressions but few clicks, or good conversions but little visibility.
How it works
The page combines Search Console (impressions, clicks) and GA4 (sessions, conversions); if either is not connected, it says so.
Search ConsoleGoogle Analytics 4

Who captures your tracked prompts

Illustrative example

In the app: Act › GEO Lab

Prompt → Page

What it does
For each tracked prompt, which page should capture it: captured by your site, by third parties, cited without a URL, or missed.
How it works
Kaiku matches your prompts with the URLs engines actually cite and derives a status, a priority and an action: a brief, or an audit.
Sunday AI scan

Your prompts by your best rank

Illustrative example

In the app: Act › GEO Lab

AI Rank

What it does
Your place in the answer, engine by engine, and who comes first: “on Perplexity, you are recommended third”.
How it works
Rank comes from the order in which brands appear in scan answers; the top 3 is the threshold that matters in a conversational answer.
Sunday AI scan

The words engines use to describe you

Illustrative example

In the app: Act › GEO Lab

AI Perception

What it does
The words engines use to describe you: vocabulary, tone, associations, strengths and reservations, engine by engine.
How it works
On request, Claude rereads the mention excerpts collected by your scans. Every analysis is kept and compares with the previous one.
Sunday AI scanClaude analysis

Accuracy by engine, out of 100

Illustrative example

In the app: Act › GEO Lab

AI Accuracy

What it does
Is what AI says about your brand true? Each answer is checked fact by fact, with a correction for every error.
How it works
Kaiku asks ChatGPT, Perplexity, Gemini and Claude a neutral question about your brand, then compares each answer with your official site, used as the reference.
Live queryReads your siteClaude analysis

Questions suggested for a keyword

Illustrative example

In the app: Act › GEO Lab

Prompt Optimizer

What it does
From one keyword, the questions a customer would ask an AI, sorted by search angle.
How it works
Prompts are suggested from the project's brand and sector; test them in the Simulator or add them to the project.
Claude analysisProject data

Verdict by engine, live

Illustrative example

In the app: Act › GEO Lab

Simulator

What it does
Ask AI engines a question, right now, and see whether they cite you, with the competitors they name.
How it works
The question goes live to the engines, in the language you pick. It is a one-off test: it does not replace the Sunday scan.
Live query

Prompts × engines: brand and competitors

Illustrative example

In the app: Act › GEO Lab

Heatmap

What it does
For each prompt and each engine, is your brand cited? Cells marked “absent, competitors cited” are the most urgent.
How it works
The grid fills in from your scans' answers; until a prompt has run, it says so.
Sunday AI scan

05 · Project

Project

Everything around the data: settings, team, reports, alerts, billing.

The setting that drives almost all your data

Illustrative example

In the app: Project › Settings

Settings

What it does
Keywords, prompts, competitors, project market (country and language), Google and Bing connections, Business Profiles, sends and notifications.
How it works
The market drives rankings, volumes, domain analysis and the press edition searched. Search Console, GA4 and Bing connect read-only.
Project dataSearch ConsoleGoogle Analytics 4Bing WebmasterGoogle Business Profile

Members, roles and projects

Illustrative example

In the app: Project › Team

Team

What it does
Invite people and assign them projects in one go, with a role: Owner, Editor or Viewer.
How it works
A viewer looks without running anything; an editor works without touching connections or sharing. The number of collaborators depends on the owner's plan.
Project data

Build a report, section by section

Illustrative example

In the app: Project › Reports & sends

Reports

What it does
Build a client report section by section — GEO, SEO, Analytics, Local — with AI commentary, as a PDF or a shareable link.
How it works
Saved templates reapply in one click. Automatic reports and sends are switched on project by project, in settings.
From ProProject dataClaude analysis

The Monday email

Illustrative example

In the app: Project › Reports & sends

Weekly digest

What it does
A Monday email that sums up the week: brand mentions, competitors rising or falling, newly cited URLs, recommendations.
How it works
It goes to each project's owner; a preview shows the email exactly as it will arrive, and a button sends it to a test address.
From ProSunday AI scanProject data

Grouped alerts: one line, its occurrences

Illustrative example

In the app: Project › Alerts

Alerts

What it does
Hear about what matters: a score drop, a new competitor, a negative review, a lost ranking.
How it works
Each alert type has its own threshold and frequency; an alert that keeps returning on the same query stays on one line, with its dates.
Sunday AI scanSearch ConsoleGoogle Business ProfileDataForSEO

The Sunday scan, and a frequency of your own

Illustrative example

In the app: Project › Scheduling

Scheduling

What it does
The full scan runs every Sunday; scheduling lets you choose its frequency.
How it works
A custom frequency comes with the Enterprise plan. The rest of the week, pages read the latest scan's data.
Project data

GEO Score and your actions, on one timeline

Illustrative example

In the app: Project › Timeline

Timeline

What it does
One thread that puts your scans, publications and actions at their date, on the curve of your indicators.
How it works
Built from the project log; a partial reading, when an engine was down, stays marked as such.
Project dataSunday AI scan

Report preview in the agency's brand

Illustrative example

In the app: Project › White label

White label

What it does
Deliverables in your brand: name, logo, colors and footer, project by project.
How it works
Settings apply to the white-label report, custom reports and the shareable report link; the preview updates as you type.
EnterpriseProject data

Plan usage gauges

Illustrative example

In the app: Project › Billing

Billing

What it does
Your plan, your usage, and what each plan unlocks in measurements.
How it works
Usage gauges appear where limits apply, with a one-click upgrade. What each plan includes is listed on the pricing page.
Project data

What this project used, day by day (USD)

Illustrative example

In the app: Project › API costs

API costs

What it does
What this project used in data, day by day and API by API, in US and Canadian dollars.
How it works
Only the project owner sees this page: costs are not the collaborators' business.
Project data

Your projects, and those shared with you

Illustrative example

In the app: Project › Projects

Projects

What it does
All your projects — one site, client or brand each — the ones shared with you, and creating a new one.
How it works
Each project has its own keywords, prompts, competitors, connections and history, kept apart from the others.
Project data

Where the data comes from

Real, named sources.

Every figure in Kaiku comes from a source you can name in front of a client.

  • Sunday AI scans

    Your prompts asked of ChatGPT, Gemini, Claude and Perplexity, and Google's AI Overview recorded on your keywords. Full answers archived.

  • Search Console and GA4

    Read-only Google connection, revocable at any time. Nothing Kaiku archives is ever deleted.

  • Bing Webmaster

    Encrypted API key, never shown in clear again. Bing powers Microsoft Copilot.

  • Google Business Profile

    Official API, with your permission: profile statistics, searches and reviews.

  • DataForSEO

    Rankings, volumes, results pages, links, YouTube, Reddit, trends and traffic estimates, labeled as such.

  • Your site

    Read live for audits, robots.txt, schemas and SEO drift; changed only when you place a schema or send a WordPress draft.

  • Public sources

    Google News, Wikidata, Wikipedia, Hacker News, Bluesky, the App Store, Google Play and the Chrome UX Report.

  • Claude

    Period analyses, briefs, writing and fact-checking, in the project's language.

When a figure is missing, the interface says so: never an invented zero, never an estimate passed off as a measurement.

Get started

See who cites you, from the very first check.

One-month free trial, no credit card. What each plan includes is laid out on the pricing page.

Start for freeWhat each plan includes