The stack a freelance consultant usually pays for
Priced separately, the tools a solo SEO needs add up quickly: a keyword and backlink platform, a content optimization tool, an AI detector, a plagiarism checker, and something to answer the AI visibility question clients have started asking. Each is defensible on its own and together they are most of a month's profit on a small retainer.
The cost is not only money. Every tool boundary is an export, a reformat and a re-import, and that work is unbillable. A brief leaves the keyword tool as a spreadsheet, gets retyped into the writing tool, comes out as a document, goes through the detector, then gets pasted into a CMS. Four handoffs, none of which the client sees or pays for.
Collapsing that into one place is worth more to a freelancer than to an agency, because an agency can put a junior on the handoffs and you cannot.
Pricing that follows a variable book
Client counts move for freelancers in a way they rarely do for agencies. Three retainers becomes five after a good quarter and back to three when one is acquired and another takes the work in-house. A tool that assumes a stable book is a tool that overcharges in the quiet months.
Client workspaces are the meter here, so the entry tier is genuinely usable rather than a demo. Seats are included, which for a solo consultant mostly matters because it means you can add a client's developer or a writer you subcontract without a per-user charge appearing.
There is a seven-day trial with no card, covering one workspace, a hundred visibility checks, two AI articles and one site audit. That is deliberately enough to run a real client through rather than enough to look at the interface.
The AI visibility question you are now being asked
Clients read that people are asking assistants which supplier to use, and they want to know where they stand. Answering with an opinion is uncomfortable, and answering by typing the question into ChatGPT yourself is not something you can put in a report.
A visibility run puts your client's prompts to ChatGPT, Gemini, Perplexity, Claude and Google's AI Overview, and records whether the brand was named. Detection is a plain string match rather than another model grading its own work, so two runs a month apart are measured by the same rule and a change in the number means something changed in the world.
It also returns the sources each engine cited. That is the actionable half: when a competitor is named and your client is not, the cited pages usually explain why, and writing the page that closes the gap is billable work that came out of the measurement rather than out of a hunch.
What it will not do for you
There is no white-label mode, no branded PDF export and no client portal. Runs are stored and re-openable, so the underlying data for a report is there, but assembling something with your name on it is still your job in whatever you already use. Some plan tables list a white-label entitlement ahead of the feature shipping.
It is also not a rank tracker with daily position history, and it does not do outreach for you or manage your projects. It covers the analysis, the writing and the measurement. The client relationship, which is the part you are actually paid for, stays yours.
Where the hours actually go, and which ones are recoverable
A solo consultant sells hours whether or not the invoice is worded that way, so the only meaningful question about a tool is which hours it gives back and whether those hours were billable. Most of the answer is unglamorous.
The recoverable ones are the handoffs. Exporting a keyword list, reformatting it into a brief, retyping the brief into a writing tool, running the finished draft through a detector on another site, pasting the result into a CMS and fixing the headings it broke. None of that is work a client would agree to pay for if it appeared on an invoice as its own line, and it is often an hour per article.
The second recoverable block is remembering. When you carry five clients alone, a real part of the week is reconstructing where you left off: which pages you already fixed, which keywords you already covered, what the site looked like before. Runs stored per client with dates on them replace that reconstruction with opening a workspace.
What is not recoverable is judgement. Deciding whether a keyword is worth a page for this particular client, whether an issue the crawler ranked high actually matters on their site, whether the draft says something true. That is the part you are paid for, and no measurement removes it. A tool that promised otherwise would be selling you out of your own job.
Starting without disrupting the work you already have
The lowest-risk way in is a client you already understand. Run the audit on their site and see whether the severity ordering matches what you would have said yourself, since a tool that disagrees with your judgement on a site you know well is one you should not trust on a site you do not.
Then import a published page by URL into the editor. It scores existing text against targets measured from the pages currently ranking, which usually surfaces a page or two sitting just off the first page for a term worth having. That is a quick, real piece of work rather than an evaluation exercise, and it tells you inside an afternoon whether the tool earns its place.
The one thing worth doing before you commit is a visibility run on a client whose market you know well. Answer-engine results are noisy by nature, and the useful check is not whether the score flatters anyone. It is whether the competitors the engines name are the ones you would have named, and whether the pages they cite are the pages you would have pointed at. If both match your read of the market, the measurement is describing something real and you can put it in front of a client. If they do not, you have learned that cheaply, in a week, on a site where you were able to tell.