Why client count is the honest thing to charge for
Pricing tells you what a tool thinks it is worth to you. Per-seat pricing says the value is in access, which is why it grows when your team does and stays flat when your client list doubles. For an agency that is exactly the wrong shape: you added six clients on the same headcount and paid nothing extra, then hired one coordinator and the bill went up.
Client workspaces track the other line. Ten clients cost more than three because ten clients earn more than three, and the tool becomes more expensive at the same moment it becomes more valuable. That is the only version of a bill that survives a hard month without an argument about who still needs a login.
- One login per account, so team size never changes the price
- Workspaces are the meter: 1, 5 or 15, and the plan follows the number
- Setting up a client's site is free, and shows what to fix first before you pay
- Annual billing is available where the retainer is annual too
What a client retainer looks like in here
Month one is usually the audit. Point the crawler at the domain and it reports what is actually on the live pages: redirect chains, canonical and indexability problems, duplicate or missing metadata, duplicated body content, broken internal links. Connect their Search Console and the fix list reorders itself around the pages that hold impressions, which is the difference between a list your client will pay to work through and a list they file.
Month two is usually content. Compare pulls the searches a rival wins and they do not, and the client report names the pages the site has no answer for. Pick one and it opens as a document with targets measured from the pages currently ranking for that search in that market.
From month three the job is proving it moved. Re-run the audit and the old crawl sits beside the new one rather than being replaced. Re-run the visibility check and this week sits beside last week. Both are things you can put in front of a client without asking them to take your word for it.
Reporting: what this does and what it does not
Everything a run produces is stored and re-openable at no cost, per client, with the date it was run. That is the raw material of a client report and it is genuinely useful: you can open March next to June and show what changed.
What is not here is a branded report builder. The client report exports as a PDF and a share link, but it carries our name, not yours: there is no white-label mode and no client-facing portal yet, so treat them as unavailable until they appear in the product.
We would rather say that on the page you are reading than let you find out during a client handover. If a branded deliverable is the thing you are shopping for, this is not that tool yet, and the honest answer is worth more to both of us than a signup.
AI visibility, per client, as a service line
Clients have started asking whether ChatGPT mentions them, and most agencies do not have an answer that survives a follow-up question. This runs their prompts across five answer engines, records whether the brand was named by a plain string match rather than a model grading itself, records where it was named, and lists the URLs each engine actually cited on the way to its answer.
The citation list is the part that turns it into work you can bill. When a competitor is named and your client is not, the cited pages usually explain why, and the page that closes the gap gets written in the same product. Prompts and engines are chosen per workspace, so a local services client and a software client are not measured against the same questions.
Using an audit to win the client in the first place
The hardest part of an agency pitch is being specific before you have access to anything. Prospects have heard the general version of your process from four other agencies that week, and a proposal describing a methodology is indistinguishable from the proposal next to it.
A crawl needs nothing from the prospect. Point it at their public domain and you can walk into the call with the actual redirect chains, the duplicate titles, the canonical tags quietly excluding pages they care about, and how many pages are affected by each. The conversation stops being about whether you are good and becomes about whether the twenty-two broken internal links they did not know existed are worth fixing.
The same applies to the answer-engine question, which is currently the easiest way to say something a prospect has not heard. Run their brand against the questions their buyers ask and you can tell them which of the five engines name them, which name a competitor instead, and which specific competitor pages those engines cited. Almost nobody pitching them has that.
One caution worth building into how you use this: a run costs money and a prospect is not yet paying you. Each button says what it uses before you press it, so treat pitch crawls as a deliberate spend on named opportunities rather than something you fire at every inbound lead.
What this replaces, and what it does not
It covers the work most agency retainers are actually made of: technical audit, content gap analysis, briefing, writing and optimization, AI detection, plagiarism, publishing, and AI visibility tracking. Most teams pay for those separately; here they are on one bill, with the pieces aware of each other.
It is not a rank tracker with daily position history, and it is not a project management tool. If your process depends on a shared task board or on client-facing dashboards with your branding on them, keep those. What you can drop is the pile of single-purpose tools that each know one thing about a site and nothing about each other.