Skip to main content
Marketing & SEO

ChatSEO Review: What the Free Start Shows (We Tested It)

· 7 min read

This article contains affiliate links: we may earn a commission. Learn more

We typed our own domain into ChatSEO and got a working diagnosis in about fifteen minutes, with no card. It read 57 pages, first described our business wrongly, then correctly spotted a real flaw on our homepage. The detailed plan to fix it sits behind a card-based trial. We ran the test on November 14, 2026, on a free ChatSEO start, with a new account on our agent mailbox.

This is a test of the free part, not a review of the paid product. We did not pay, so everything about the paid plans comes from the pricing page. How we test explains why our agents do the testing.

What does ChatSEO do?

It’s an SEO assistant you talk to, which reads your site and proposes what to fix first. The home page promises “the SEO expert that makes you visible on Google”, free to start with no credit card. You give it a domain, and an agent crawls it, picks competitors, and answers questions in a chat. The paid plans add 14 “SEO agents”, an AI writing agent and publishing to a CMS.

We care about one thing here: what a small site with almost no traffic gets before paying. Ours is the right test case. On November 13, 2026, Search Console showed us 31 to 45 impressions a day and 0 to 1 click.

What happens when you enter your domain?

It starts without an account. We typed hayka-pacha.io into the box on the home page. The onboarding opened at once and crawled the site: “57 pages crawled. Here’s how Google shows you today”, with our title and meta description as Google shows them. No form first.

The account step comes after the analysis: “Continue with Google”, or an email with a sign-in link or a password. We chose the email link, and it arrived within a minute. The onboarding then restarted from the first screen, so we did the steps a second time.

Did it understand our site?

Not at first. The offer it wrote for us was: “An AI-first automation agency that builds and operates self-running digital infrastructure using AI agents instead of human teams”. We are a media that tests AI tools. Its first competitor suggestions followed from that error: feedough.com, amenitytech.ai and ninetwothree.co.

The step asks “Your offer, is that right?” and lets you edit the text, so we rewrote it as a trilingual media that tests AI and automation tools. The suggestions changed to isthereaiforthat.ai, toolify.ai and mrkhan.digital. We kept the first two. Then it measured authority:

DomainAuthority (out of 100)
toolify.ai73
hayka-pacha.io0
isthereaiforthat.ai0

“Your domain has almost no authority: that’s normal, it’s young. Nobody links to you yet.” That’s accurate and blunt. If you accept the wrong offer without reading it, everything after it drifts, so read that screen.

Is the diagnosis of our homepage right?

Yes, on the main point. We asked “Analyze my page /en/ and tell me precisely what to change so it gets more clicks from Google.” In under two minutes it said our homepage competes with itself: the title promises tool testing, the H1 is a brand slogan, and the body is an agency pitch. It added that Google gets no single intent to match.

We checked the live page. The title is “AI and Automation Tools, Tested | Hayka Pacha”. The H1 is “Ancient wisdom. Digital precision.” The three H2s are “Why Hayka Pacha?”, “From the field” and “Build something together”. So the flaw is real, and the wrong offer from the onboarding came straight from our own page. That one is on us, and we haven’t fixed it yet.

Its numbers were close but not exact:

  • Images: it said 38 of 39 had no alt text. We count 41 images, 38 with an empty alt (the decorative icons).
  • External links: it said 38. We count 40 links to other domains, 21 of them distinct.
  • Body length: it said 300 words. We count about 375 words of text on the page, menus included.

An empty alt is normal for a decorative icon, so “no alt” overstates that one. The tool says itself that it can make mistakes. Check its figures before you act on them.

Which keywords did it suggest?

Two lists, and only one is useful. On the competitor step it suggested queries to target “where your site can win”: “slock ai” (210 searches a month), “is spicychat down” (720) and “snapappraise” (320). None has anything to do with testing tools. It had no Search Console data from us, since we clicked Later when it asked to connect.

On the homepage analysis it read keyword data for four candidates, US market:

KeywordVolume/monthCPC
best ai tools6,600$19.48
ai automation agency4,400$25.22
ai automation tools2,400$32.19
ai automation services720$37.56

That list fits our niche. It stopped there: “I need you to pick the target before I write the plan”, and under it a button, “Unlock my full analysis”.

Where does the free part stop?

At the plan itself. The button opens a choice of plans, each with “Try 7 days for free”. The home page says “no credit card required”; the pricing FAQ says the trial gives 7 days and 30 credits “with nothing to pay upfront”, and that the card “lets your plan start without interruption on day 7”, with a reminder 3 days before the first charge. We didn’t start the trial, because our agents never enter a card.

Here’s what we read, on November 14, 2026:

PlanPer monthCreditsSitesNotable
Starter€29 (€23.20 yearly)1001Search Console, AI writing agent
Pro€49 (€39 yearly)2003adds Analytics, API and MCP
Ranker€79 (€63 yearly)40010adds white-label exports, priority support

A credit is one message to the agent, “however complex”. An article costs 2 credits. So Starter pays for roughly 100 questions a month, or fewer if you ask for articles.

Our verdict

Worth fifteen minutes as a free site check. We can’t yet say it’s worth €23 a month. The free start found a real problem on our homepage, which our own checks hadn’t flagged, and it did so from a bare domain. Its weak points were visible: a wrong first reading of our offer, off-topic keywords until we corrected it, and numbers that were close but not exact.

We didn’t test what you pay for: Search Console data, the article agent, the CMS publishing. The site claims “used by 20,000+ website owners” on the home page and “900+ founders” inside the app, and we can’t check either. If your site already has traffic, connect Search Console and judge it on your own queries. If you’re deciding on SEO tools more broadly, our AI visibility tools compared tests the other side of the problem, and GEO vs SEO explains why the two now overlap. For another free-start test, see what an agent gets on systeme.io’s free plan. To check whether AI engines cite your site at all, start with a free GEO audit.

FAQ

Partly. On November 14, 2026, we ran the onboarding and a first page analysis on a new account with only an email address, no card. The full plan of fixes sat behind a 7-day trial that asks for a card, charged on day 7 unless you cancel. Paid plans start at €29 a month, or €23 billed yearly.

Not to start. It offers to connect Search Console, read-only, and we clicked Later. It then worked from our site and public rankings, and said itself that without real data it can only give an estimate. Search Console is included in every paid plan.

Pricing page, November 14, 2026: Starter €29 a month (100 credits, 1 site), Pro €49 (200 credits, 3 sites, Analytics, API and MCP), Ranker €79 (400 credits, 10 sites). Annual billing takes 20% off, so Starter is €23.20 a month. One message to the agent costs 1 credit, an article 2.

On our homepage, yes: it found that our H1 is a brand slogan and that the body reads like an agency pitch, and we checked both. It was off on details (we count 41 images where it said 39) and it first described our offer wrongly. Its keyword suggestions only made sense once we corrected the offer.

Yes. The link on this page is an affiliate link, and we earn a commission if you pay for a plan, at no extra cost to you. We didn't use the link for the test: our agent typed the domain on chatseo.app with a new account.

Related articles