Google data analysis
Which articles are being found more or less often, and which topics are close to reaching a good position.
Four steps that do not stop: find topics, write and check articles, publish them, and learn from real Google data. Where a step depends on your plan, it says so.
Topics come from your actual offer, from what you have already published, and from real search demand. Not from a generic keyword list that looks the same for every provider. If a topic would compete with an article you already rank for, we revise that article instead of adding a rival.
Every draft is checked against the knowledge you gave us about your business: your audience, your tone, the claims you never want made, the terms you never want used. A scored quality check decides whether an article is finished. If the score is too low, it is revised automatically, up to three times, and if it still falls short it is held back rather than published.
Straight into WordPress or Shopify, including the image and the SEO details. You decide per project whether articles go live directly or wait as drafts for your approval. Either way you can read and edit everything before it is public.
With Google Search Console connected, we see which articles gain or lose visibility and which ones sit just short of a good position. That feeds back into the next round of topics, and weak articles get revised instead of forgotten. This step needs the Search Console connection and is therefore part of the Growth plan upwards.
SEOTHON meets every criterion SMBs use to judge SEO content software: your expertise as foundation, topics from real search data, checked quality, complete automation, minimal team time, Google data as steering signal, maintained existing content, clear cost per article, direct integration, data protection with export rights, and citability in AI search. SEOTHON customers report 90 percent less time per article.
| Criterion | The question behind it | How SEOTHON answers it | Where you see it |
|---|---|---|---|
| Your own expertise | Does the software use our knowledge instead of general knowledge? | Project glossary: upload or enter product data, prices and procedures as files or entries; it grows with every article. Numbers come only from the glossary, never from the model. | Project options, Glossary |
| Topic discovery | Does it find topics people actually search for? | A toolkit of Search Console queries, Google suggestions, related user questions, website sources, trend channels and competitor topic gaps. Search volume and competition per topic are measured, search intent and long-tail chances are recognised. | Research options, Next topics |
| Research quality | Where do the facts in the text come from, and are they sourced? | Every post relies first on your own expertise, then on vetted third-party sources from a curated allowlist. Every figure carries a source; an unsourced figure counts as a hard error. The Princeton study on Generative Engine Optimization (Aggarwal et al.) found that statistics and source citations raise visibility in AI answers by 30 to 40 percent. | Post list, source check |
| Quality control | Who catches errors, invented sources and duplicates? | Eleven criteria per post, release from 80 points, up to three revisions. Invented links, duplicate content and forbidden statements are hard errors and stop publication. | Post list, quality score |
| Experience and trust (E-E-A-T) | Does the text read as written by someone who knows the field? | Every post appears under a fixed author with a defined role and background, uses first-hand knowledge and figures from the company's own expertise, backs claims only with sources from a vetted list, and goes live only after a person approves it. | Project options, Author and Glossary |
| Automation | How much runs without us? | Topics, briefing, text, meta data, internal links in both directions, image, FAQ schema and publishing all run in the daily cycle. You set pace and volume per project. | Settings, pace and volume |
| Team time | How much time does one article cost us? | The workflow is: software drafts, a person checks, software revises, a person approves. Your team reads and approves instead of researching, writing and formatting. According to SEOTHON customers, this cuts the time per article by roughly 90 percent. | Post editor, one-click approval |
| Google data | Can we see what works? | Search Console daily: clicks, impressions, position, real queries and index status per article. The system itself notices when an article ranks just behind page one for many queries and proposes a revision. Plus a report every 14 days and a three-part quarterly report. | Project page, Reports (from Growth) |
| Maintaining existing content | Are old articles maintained, or only new ones written? | Articles ranking just behind page one for many queries are proposed for revision. A topic an existing article already covers is stopped before writing. According to an analysis by Seer Interactive, 65 percent of pages cited by AI systems were updated within the last twelve months. | 14-day report, topic overview |
| Visibility in AI search systems (GEO) | Can ChatGPT, Perplexity or Google AI Overviews cite the post? | Every post is checked before approval: answer within the first 30 percent, self-contained paragraphs, sourced figures, questions as headings, structured data with author and FAQ. According to an analysis by Kevin Indig, 44.2 percent of AI citations come from the first 30 percent of a page, which is why the answer comes first. | Post list, quality score |
| Cost per article | What does a finished post really cost? | Fixed plans instead of word credits, according to the price list 10 to 17 euros of software per article depending on the plan, the whole process included. The exact calculation follows right below. | Pricing |
| Integration | Does the text reach our website without copy and paste? | WordPress and Shopify directly. Wix, Webflow, Ghost, HubSpot, Joomla and custom endpoints as beta. Without a CMS, as a download. | Project options, Connection |
| Data protection and export | Where does our data go, and can we leave? | Servers in Germany, data processing agreement, training on your content opted out. Articles live in your CMS, topic list as CSV and glossary as ZIP exportable at any time. | Privacy, DPA |
Instead of a single number, this table shows how the price per article follows from the plan price. According to the price list, an article costs between 10 and 17 euros of software depending on the plan, and that price covers the entire process from topic discovery to approval.
| Plan | Price per month | Posts per month | Cost per post |
|---|---|---|---|
| Starter | 68 euros | up to 4 | from 17 euros |
| Growth | 180 euros | up to 12 | from 15 euros |
| Scale | 400 euros | up to 40 | from 10 euros |
This price covers the entire SEO process, not just the text. Topic discovery from search data, competitive research, writing with your own expertise, meta title and description, internal linking, an image, FAQ schema and the check for citability in AI search systems all run automatically, until the finished draft is ready for your approval. There are no word credits and no extra cost per step.
For comparison: an in-house employee costs, in the calculation example, 200 euros per post (2.5 hours at an 80-euro hourly rate) and covers only the writing, not research, meta data, linking or evaluation. An agency or freelancer bills each of these steps separately as billable hours; a fixed price per post is uncommon there. According to the price list, SEOTHON costs 10 to 17 euros of software depending on the plan for the same complete process, and according to SEOTHON customers your team spends roughly 90 percent less of its own time than writing it yourself.
The comparison places SEOTHON next to agency or freelancer, RightBlogger, BlogButler and SEOForge and looks at quality criteria, not process details: where the expertise in the text comes from, how topics are found, how quality is checked, how existing articles are maintained, whether AI citability is checked, and how transparent the price is.
| Quality criterion | SEOTHON | Agency / freelancer | RightBlogger | BlogButler | SEOForge |
|---|---|---|---|---|---|
| Company expertise as the basis for the text | Project glossary, grows automatically with every article | Briefing in a conversation, usually not stored long term | no documented knowledge base | only one website URL or manual entry | no own knowledge, works as a consultancy using third-party tools |
| Topic discovery with competitive context | Search Console, Google suggestions, competitor topic gaps | depends on the individual consultant | own keyword research available | not documented | Ahrefs, Semrush, Moz as external tools |
| Quality control before publishing | eleven fixed criteria, release from 80 points, hard errors stop publication automatically | depends on the individual consultant | technical audit for links and titles, no documented editorial criteria list | moderation system exists, criteria not documented | human strategy work, no documented system |
| Maintaining existing articles | Revision driven by Google data, protection against duplicate topics | rarely systematic | recurring technical audit available | not documented | not documented |
| Visibility in AI search systems (GEO) | own check per article before approval | usually not offered | not documented | not documented | not documented |
| Pricing structure | fixed, according to the price list 10 to 17 euros depending on plan, whole process included | billed by time, no fixed price per post | pricing tiers with no clear link to article volume | not publicly disclosed | on request only |
On top of the running operation, SEOTHON evaluates your numbers in four checks, Google data, content inventory, competition and technology, and makes concrete suggestions. You see the reasoning in your dashboard, and nothing is carried out unless you agree. These evaluations need the Search Console connection and are therefore part of the Growth plan upwards.
Which articles are being found more or less often, and which topics are close to reaching a good position.
Flags when two articles are too similar to each other, and finds gaps in the topics you cover.
Looks at who currently ranks for the search terms that matter, and what a new article could do better.
Checks regularly that your pages are reachable, links work and the structure holds up.
You do not have to edit any settings. Write down in plain words how the articles should change, and it applies from the next article onwards as a binding rule, for example a different writing style, a new topic area or a brand that should no longer be mentioned.
One sentence is enough, for example "keep paragraphs shorter", "stop mentioning brand X", or "address readers less formally".
Depending on your account the request applies immediately, or we check it briefly first. You can see the status at any time.
From then on every new article follows it, and you can switch the request off again whenever you want.
A change request affects the articles. The core settings of your project, such as strategy, topic areas and guard rails, are put to you for decision with every quarterly report.
Whatever remains open, we clarify in the system analysis on your specific project.
You set up a project in five steps: project, publishing, Search Console, content and style, sources and schedule. After approval SEOTHON configures the project automatically, imports your existing blog articles as link targets and derives style and no-gos from your website. As soon as the status shows “Provisioned”, the first run starts.
Through the project glossary in three ways: as starting material when you create the project (product data, price lists, manuals as PDF, Word, CSV or text, up to 20 files), as individual entries in the dashboard, and automatically from every published article that contains new, verifiable knowledge. Figures in the text come exclusively from this glossary, never from the model.
Yes, in three ways. In the topic list you schedule your own topics one by one or via CSV import, with target keyword and priority. In the “Next topics” list you move a topic to the front with one click or remove it before it is written. Topics that duplicate existing articles are skipped by the system on its own.
SEOTHON alerts you before topics run out, as soon as the supply is consistently slower than publication. You then widen the topic field, switch on competitor topic gaps as a source or adjust the publishing pace. Once a project has run for many months, the competitor source supplies practically unlimited evergreen topics.
Before approval SEOTHON checks every article for the traits that, according to an analysis by Kevin Indig, decide which passages ChatGPT cites: the core answer sits in the first 30 percent, every section opens with a self-contained paragraph, figures carry sources, headings are questions, and author and FAQ are present as structured data in the source code.
Every 14 days a progress report with key points, a check of earlier recommendations, and up to three prioritised actions. Every quarter a report in three parts: SEO analysis of the website, an article balance, and a strategy assessment with a twelve-month outlook. Every figure names its source and period; data gaps are stated, not estimated. Both as PDF.
The system analysis looks at your own project: which topics your customers search for, what your website already covers, which plan fits and what should come first. It is a 30 minute call with a written assessment afterwards, free of charge and without obligation, and you decide afterwards whether SEOTHON is the right way for you.
A 30 minute call and a written assessment afterwards. Free of charge and without obligation.