x-octo home Business judgment on AI products
中文
← Back to funding radar

FUNDING DESK · Sweden · AI search visibility / generative engine optimization (GEO) content automation

Miracly

Miracly is an AI search visibility platform that studies the questions buyers ask AI, generates structured content pages, and continuously tracks citations so brands get named and recommended by ChatGPT, Claude, Gemini and Perplexity.

Company website ↗

First-party material retrieved

Last retrieval:2026-11-15

Evidence gaps:Product and features · Technical documentation

Work it replaces

It replaces the traditional SEO/GEO agency audit-plus-manual-rewrite workflow, and the ad hoc internal work of researching AI search questions, writing landing pages and manually tracking rankings.

Business judgment

It turns generative engine optimization from a one-off audit into an ongoing content-production and citation-maintenance subscription, closer to outsourced content operations than to an analytics tool; whether that holds depends on clients being able to verify a causal link between AI citations and actual leads.

The following is editorial analysis based on public material. Inferences and open questions are labeled in the text. Funding is not evidence of revenue or product-market fit.
Product evidence checked 2026-11-15

01

What the product does

According to the site, Miracly crawls and analyzes a customer's website to check whether agents such as ChatGPT, Claude, Perplexity and Google can reach it and identify the brand, then generates new content pages around high-intent questions people ask AI. Pages are organized around structured facts (service scope, prices, local context, FAQ, reviews) and are added alongside the existing site without changing it. The product also offers AI citation rank tracking, competitor ad and content monitoring, social post generation, and a weekly re-check that rewrites pages whose citations slip. The pricing page shows tiers differentiated by tracked prompts, page caps, daily publishing volume and number of markets, with the full research, writing, scheduling, publishing and rank-tracking engine included from the Starter tier.

02

Users, buyers and demand

The target customer is a small or mid-sized business that depends on local or niche search for leads, such as the Swedish painting company in the case study, plus restaurants, clinics and professional services that must answer questions about location, price and booking; the pricing page also lists an Agency tier and a referral commission program, making agencies an explicit paying segment. Users are typically owners or marketing leads whose pain is that buyers increasingly ask AI directly and act on the answer, while the business neither knows how AI describes it nor has staff to keep producing agent-readable content. Payers are usually single-brand owners billed per brand with unlimited seats, so the decision chain is short but budgets are sensitive.

03

The actual workflow

The flow starts with a URL: the system analyzes the industry and site, lists questions people ask AI, clusters them into subjects and plans pages; it then writes one page per question, constrained by brand voice and customer-supplied rules such as blocked words or compliance requirements, in 40+ languages. Pages publish directly into the existing stack via WordPress, Shopify, Squarespace, Next.js, React plugins or an API, with no re-theming needed. After launch, a dashboard shows position in each question and each agent's answer in real time, and pages are re-checked weekly: cited pages are kept, slipped ones rewritten, never-cited ones retired.

04

Pricing and unit economics

The pricing page publishes monthly subscription tiers from Starter through Pro and Scale to Enterprise, with roughly 5% savings on annual billing, no credit card to start and cancellation anytime; Claude, Perplexity, Grok and other engines are add-ons on lower tiers, while the compliance module, social module, managed onboarding and VIP support appear only at higher tiers. Billing per brand rather than per seat means expansion comes mainly from page volume, markets and add-on modules. Unit economics are not disclosed: inference costs for generation and weekly multi-engine re-checks, hosting and crawling costs, and customer lifetime are unknown, so any margin inference is conditional only.

05

Adoption evidence and gaps

Verifiable adoption evidence is mostly company-reported: the site's case study says Swedish painter Spektrum grew from 5 indexed pages to 1,140 in about 90 days, with Google Search Console showing large impression and click growth, 99% non-branded traffic, and ChatGPT recommending the brand unprompted on relevant questions. Those figures are attributed to Search Console but compiled and disclosed by the company, not independently audited; the "5 to 10 leads a day" claim is a quote from the owner and is company-reported. Beyond that, customer count, retention, churn and revenue are undisclosed, and the case page shows only one named customer, a very small sample.

06

Competition and defensibility

Against general-purpose models, Miracly's difference is not writing quality but a loop built around being cited by agents: question discovery, structured-fact templates, multi-engine citation tracking, automatic rewriting when citations slip, and publishing integrations with major site builders. Against legacy SEO workflows, it collapses audit, writing, publishing and monitoring into one subscription, lowering execution friction. But the barrier is weak: prompt research, content generation and rank scraping are replicable, and a real moat requires accumulating cross-industry, cross-engine citation data to learn what makes pages citable, plus getting clients to attribute results to the product rather than natural variance.

07

How to read this round

This is a Pre-Seed round; amount and stage are rendered by the data layer, the investor list includes institutions and several individual angels but does not indicate a lead, and use of funds is undisclosed. The stage shows the company can articulate a product shape with public pricing and a case study and has raised capital to test acquisition and retention, not that demand is validated at scale. The next milestone should be verifiable paying-customer counts, monthly retention and evidence linking citations to leads; with use of funds unknown, it is impossible to say whether capital goes to product, sales or content capacity.

08

Where it could fail

First, attribution risk: the chain from AI citation position to real leads can be confounded by search rankings, seasonality or organic brand growth, and clients will churn quickly if leads do not move after subscribing. Second, platform dependency: value rests heavily on how ChatGPT, Claude and other engines crawl and cite, so a change in answer generation, crawler access or citation display could invalidate the page strategy wholesale. Third, content-at-scale risk: publishing multiple pages daily could trigger search quality judgments or dilute a client's site, and whether the thousand-page expansion generalizes to other industries is unproven. Fourth, competition and pricing risk: general models and site builders could embed similar features and compress subscription prices.

09

What you can take from it

A transferable practice is decomposing "being recommended by AI" into an executable product loop: a free site-readability check as the acquisition hook, structured-fact templates to make content agent-parseable, and weekly re-checks with automatic rewriting to turn one-off delivery into a recurring subscription. Another is pricing per brand rather than per seat to reduce team-expansion friction while keeping room to grow through add-on modules.

10

What to watch next

Watch whether paying-customer counts and retention become public; whether verifiable Search Console data appears across industries beyond Spektrum; whether a testable link between citation changes and client leads is established; and how quickly the product adapts when AI engines change their rules.

Sources and verification