Show any client how AI really sees them

Yes — completely. The WordPress edition runs a full AI Readiness, Agent Readiness, GEO and SEO scan of your own site, includes the AI files manager, a local security-header check and a branded PDF, and needs no account and no API key. Every check runs on your own server and nothing is sent to any external service.

Yes. Daryo89 Multi Scan and AI Assessment is the single official name of the plugin, and it is the name you will find on WordPress.org. Earlier material referred to it as “D89 AI Visibility” or “the D89 AI Visibility Suite”; those describe the same product and the same engine. We have unified the name so that search engines, AI assistants and directories all resolve to one entity rather than two.

Two things. First, evidence for the pitch: a dated, traceable, white-label report showing which questions the AI engines answer without naming the prospect, and which competitors they name instead. That gap is the brief, and it is far harder to argue with than an opinion. Second, proof of delivery: every report carries a delta line against the previous scan, so at renewal you can show what the work moved rather than describing it. The Marketing Agent module (formerly Article Autopilot) then executes the content half of the plan on a schedule, so what you sold actually ships.

All four share one engine. The WordPress edition scores and helps you fix your own site locally, free, forever. Client and Agency install free but need an active licence to run — that licence comes with any audit purchase. Client is activated by a single audit and covers one business; it adds live AI-citation testing across Gemini, Claude and ChatGPT, Lighthouse and security scoring, per-engine brand analysis with a Claude synthesis, and a client-ready report. Agency starts at the five-audit bundle and adds scanning other people’s sites, full white-label branding, a background queue with email delivery, native-language reports, the local and national market analysis, and the Autopilot Add-Ons dashboard for managing client installs. M is the flagship partner tier: a fourth engine (Perplexity), the exclusive report design, your own AI provider keys for unlimited unmetered scanning, Companies House verification of the registered business details, AI crawler analytics, a bounded technical and on-page site audit, and a built-in MCP server. Offered to accredited partners on a monthly subscription.

With any audit bundle of five audits or more. The five-audit bundle is £250 (£50 per audit) and is the Agency entry point — it carries the licence that switches the Agency edition on. Larger bundles reduce the per-audit rate: 10 for £400 (£40 each), 20 for £700 (£35 each) and 50 for £1,250 (£25 each). A single audit does not unlock Agency — it activates the Client edition, which covers one business and scans only its own site.

It depends on whose API keys do the work. On Agency, generation runs on the Daryo89 infrastructure so your client sites need no keys of their own: a £50/month access fee unlocks the Marketing Agent on your account and includes 20 slots, then each client site where it runs is £15/month. Each site carries a monthly generation allowance of roughly four articles and eight social posts, shown in the dashboard as a percentage with its reset date; £5 top-up credits extend it when a site runs hot. On M, access is included with the subscription (also with 20 slots) and each site is £5/month, because generation always runs on your own provider keys and you carry the model cost directly. Need more than 20 sites? Slot blocks add +20 or +50. For a single site without the suite, the standalone licence is £129 a month with generation included, or £79 a month on your own AI keys.

Yes. The standalone licence is in the shop at £129 a month per site with generation included, or £79 a month on your own AI keys, annual at ten months. Agency and M accounts license it per installation from inside the plugin, because a licence binds to the installation it runs on. At activation it locks to that website’s host and to your account licence, and gets a server-issued installation ID — that is what stops a single key being spread across a dozen client sites. A cart on this page could not do that safely, so the licence is issued from inside the plugin on the site where it will run. Install your edition, open Daryo89 AI → Autopilot Add-Ons (Agency) or Daryo89 AI → Autopilot (M), and license the site from there. Moving a licence to a different domain later is handled by support.

A slot is permission to register a Marketing Agent installation, not a free installation. Holding an active Agency access fee — or an active M subscription — grants 20 slots, so you may register the Marketing Agent on up to twenty sites; each site you actually activate is still licensed at its own per-site monthly rate. If you need more than twenty, slot blocks add +20 or +50 to the pool. When every slot is in use, the next registration is refused with a plain explanation rather than a silent failure — buy a block, or deactivate an installation you no longer need.

No. They are separately metered on purpose. Audit credits pay for premium multi-engine scans; Marketing Agent generation is billed against its own monthly allowance on its own licence. Writing content can never quietly consume the audits you bought for client reporting, and a paused or capped Marketing Agent never touches your audit pool.

It pauses rather than overspending. New generation stops, the dashboard shows the site as capped with its reset date, queued work finishes and already-scheduled posts still publish. You can either wait for the monthly reset or apply a £5 top-up credit to extend that site’s allowance. The allowance is always presented as a percentage with a reset date, so you know where a site stands without doing token arithmetic.

Separate providers, not a single model. Every paid edition queries Gemini (Google), Claude (Anthropic) and ChatGPT (OpenAI) live and independently. The M edition adds a fourth — Perplexity — which runs on your own API key and is therefore exclusive to M. Each engine is asked the same questions because a business can be cited by one and ignored by another, and the results are reconciled into a single cross-engine view — so you can see true AI visibility, and exactly which engine is leaving you out.

A single grid: every question down the side, every engine across the top. A green Yes means that engine named the business in its answer. A red No means it didn’t. A grey dash means that engine returned no answer for that question, so nothing is counted for it — we’d rather show an honest gap than quietly average it away. The branded question is marked separately as brand recall. It’s the fastest way to see not just whether a business is cited, but on which questions and by whom.

Because they measure completely different things, and conflating them flatters everybody. A branded question names the business — “what does Acme Ltd do?” — and almost every established company scores well, because the engine simply reads their website. That’s recall, not demand. A non-branded question is what a real buyer types — “who’s a good X in Y?” — and the name only appears if the engines have a reason to name it, usually from sources other than the company’s own site. That’s the number that wins work, and it’s the one most tools never separate out. We score both, per engine, and report them apart. The gap between them is usually the whole story.

When you start a premium scan, the plugin gathers the site’s signals, then securely contacts the daryo89.com infrastructure to run the live multi-engine analysis and build the report. On the Agency edition the scan can run in the background and email you when it’s ready; on Client it runs on-screen while you watch the progress. Either way, the audit balance on your licence is checked and counted centrally at daryo89.com, so it’s accurate across every site you run.

No. If you don’t want or need a plugin, you can request an audit directly on our website — the same live multi-engine analysis, delivered as a branded report, with nothing to install. The Marketing Agent is the exception: because it publishes to a WordPress site, it does require the plugin.

A premium audit is a deeper, live analysis: it queries the major AI engines to measure whether and how they cite a business, adds Google Lighthouse performance and security scoring, per-engine brand analysis and a fully branded PDF report. An audit credit is simply one premium audit — buy a single one to activate Client, or a bundle from five upward that activates Agency and tops up the balance on your licence.

Not for the free, Client or Agency editions. The free WordPress edition needs no keys at all, and on Client and Agency the live AI-citation testing and Lighthouse scoring run through the daryo89.com infrastructure, so keys are managed centrally for you — there’s nothing to configure. That applies to the Marketing Agent on Agency too: your client sites publish without ever holding an API key. Only the M partner edition runs on your own API keys, which is exactly what makes it unmetered — it’s why Perplexity is M-exclusive, and why the Marketing Agent is cheaper per site on M.

It’s recorded honestly and it costs you nothing. In the citation matrix that engine’s cell shows a grey dash for that question, and its answer is excluded from the rates rather than counted as a miss. If you switch an engine off in Settings, the Scan Log says so in plain words — that it was a configuration choice, not a system error, and that no audit credit was used for it. A metered client should never read a deactivated engine as an undelivered token.

Just as robots.txt guides traditional search crawlers, llms.txt is a plain-text file that tells AI assistants where key content lives, so they can find and cite it accurately. All editions include a manager that helps you create and publish it, along with ai.txt and your robots.txt AI-crawler rules.

Yes — that’s exactly what the Agency and M editions are for. They can scan any site (not only the one they’re installed on) and are fully white-label, so every report carries your own branding. Start with the five-audit bundle at £250 to unlock Agency, and you can run AI visibility as a productised service.

On the M edition, yes. It includes a built-in MCP server (Model Context Protocol) that you switch on from Settings. Once enabled, an AI assistant such as Claude, Cursor or any MCP-compatible client can query the site’s AI-visibility data in plain language — crawler activity, AI referrals, discovery-file status and the last readiness scan. Every tool is read-only and free: nothing runs a new scan or spends an audit credit, so an assistant can explore the data safely. Access is protected by a private token you control and can rotate at any time.

Yes, on the M edition. The AI Crawler Analytics view shows which AI crawlers — GPTBot, ClaudeBot, PerplexityBot, Google-Extended and more — are actually reaching the site, how often, and their share of all bot traffic, so you can tell whether AI engines are even seeing the content. It’s measured server-side (not estimated), and an optional lightweight Cloudflare Worker captures crawlers even on fully cached sites that never reach WordPress. Alongside it, the M edition’s technical & on-page site audit runs a bounded crawl of your own sitemap and flags redirects, broken links, duplicate titles and meta, indexability and structured-data gaps — each with a fix and an overall health score.

On the M partner tier, the report doesn’t take the website’s word for it: the registered business name, company number, status, incorporation date, nature of business and registered office are checked live against the Companies House public register, and the report states whether the details published on the site match the official record. When the location verifies, the market comparison is built on that confirmed address — so the competitive analysis reflects where the business is really registered, not just where it claims to be.

Every scan — free or premium — is stamped with a unique 10-character Scan ID. It appears in the report header, in your Scan Log, and on our audit service, so any figure in any PDF can be traced end to end to the exact run that produced it. For agencies it means every client deliverable is reproducible and defensible; for everyone it means the numbers are accountable.

Partly — and the distinction matters. Every technical score (AI Readiness, GEO, SEO, Agent Readiness, security headers, performance) is fetched live at scan time, so a fix shows on the very next scan. Purge the CDN cache first, or the scanner may be served the pre-fix files. Citation results are different: they’re live web searches, but the web itself takes roughly one to four weeks to index a change. So a scan the day after a fix will honestly show “the site is fixed, the market hasn’t noticed yet”. That gap is real and we won’t paper over it — it’s why the report carries a delta line against the previous scan, and why a re-test after about thirty days is the number that actually tells you whether the work landed.