How we test, step by step

1. A real site, a real account

SemChat is tested with Jeremy Santos’s account, on his own sites. The first test dates from October 29, 2026: in a single sentence, typed in French, I asked SemChat for an SEO audit of plugkit-avis.com with an action plan by priority.

What I noted that day:

• Time: about 3 minutes between sending the sentence and the action plan.
• Output: 8 indicators, a 5-point diagnosis and 3 actions for the week.
• Cost: 100% of the Free plan’s monthly quota, for this one audit.
• Accuracy: the missing contact page (404 error) was indeed missing, I checked it the same day; the duplicate titles between English versions were real, but less serious than stated, since those pages were declared as variants of each other.

I also note what breaks. I only show screens from my own account: no fake screen is ever made. You’ll find the full write-up of that test, limits included, in my SemChat review.

2. Competitor sheets read at the source

Each tool compared has its own sheet: plans and prices, trial or free plan, data sources, and the features that matter to someone who wants to show up on Google without being an expert (site audit, rank tracking, keyword research, competitor analysis, writing help, conversational assistant).

Facts come from the publisher’s official pages: pricing, documentation, llms.txt file when there is one. Each source is noted with the date I read it; the current sheets were read on October 29, 2026. A third-party source is flagged as such and never sets a price.

3. “Not mentioned” doesn’t mean “missing”

If a publisher doesn’t mention a feature on its official pages, I don’t conclude it doesn’t exist. In tables, the cell reads “Not verified”; in the text, I write “not mentioned on its official page”, with the date. “No” only appears if the publisher writes it.

4. Fair, verifiable comparisons

In France, comparative advertising is governed by article L122-1 of the French Consumer Code. It must not be misleading, must cover products that meet the same need, and must objectively compare essential, relevant, verifiable and representative features, which may include price. I apply this rule to every comparison, whatever the page’s language.

In practice:

• I compare tools that serve the same need: understanding why a site doesn’t show up on Google, and what to do about it;
• every price carries the date I read it, and every fact links to its source;
• every comparison lists the cases where the other tool is the right choice, next to those where SemChat is;
• I present a competitor’s real strengths before the differences, without running it down.

SemChat and Haloscan share the same publisher, ARCHI301: I say so whenever I compare them.

5. Prices reread every 3 months

Prices change fast. On pricing pages and comparisons, I reread official prices every 3 months, and as soon as a publisher announces a plan change. Each update changes the date shown at the top of the page and adds the source reread.

Before you subscribe, always check the price on the publisher’s site: it is the one that counts.

6. What can’t be bought

• No place for sale. No publisher, SemChat competitor or not, pays to appear on the site or to move up a list.
• Two possible commissions, both disclosed. SemChat, through my link, and Haloscan, through my code: everything is explained on the page about the SemChat affiliate link.
• No made-up rating. No star rating or customer review is shown; a rating will only appear after a documented, signed test.

7. How I pick topics

Topics come from the questions people actually ask: Google searches measured with Haloscan and DataForSEO (data of October 29, 2026), and a reading of Google’s top ten results for each topic. These volumes are used to choose which pages to write, not to fill the text with numbers.

Spotted a mistake?

An outdated price, a wrong fact, a feature that changed: report it using the contact details in the legal notice. The page is then corrected, and its update date changes.