FAQs

Getting started

Do I need to know how to code? No. Productised is built for coaches, consultants and agencies, not developers. You never write code. The only technical steps are optional — pasting an embed snippet into your website, or adding a DNS record for a custom domain — and both come with copy-ready instructions.

How long does it take to build a product? Build with AI makes a working first version in about a minute. Describe your idea, choose what people get at the end, choose how the product opens, and pick a theme. Refining the questions, scoring and result page usually takes another 10–30 minutes.

What is the difference between an AI product and a chatbot? A chatbot is an open-ended conversation with no defined goal. An AI product has a clear goal: it asks the right questions, works through the answers using your method, and ends in a defined outcome — a personalised result, a designed result page, or an ongoing Advisor session.

Can I build products from Claude? Yes. The Productised connector is verified in Claude's connector directory. Connect it once, then build, edit and publish products by asking Claude. See Claude connector.

Building products

What is an Outcome? An Outcome is what happens at the end of the conversation — a Standard written result, a designed result page, a Document, a live Advisor session, or nothing at all if you only want the lead (Collect only). On the canvas, choose it in the AI Product node's Outcome tab. See What are Outcomes?.

What result page designs are there? There are 24, each built to fill itself from the conversation — including Growth Diagnostic, Readiness Scorecard, Maturity Ladder, Audit Checklist, Benchmark, Coaching Fit Check, SWOT Analysis, Priority Matrix, Cost of Inaction, Best-fit Recommendation, Action Plan, Timeline, Scope & Price Estimate, Budget Allocation, Strategy Report, Proposal, Future-state Story and Content Kit. Build with AI shows 19 of them as preview cards. All 24 — including Risk Register, Radar Profile, Persona, Deep Insights and Grid Scorecard — are available through the Claude connector.

What is the MCP node? The MCP node lets your product use other tools while someone is answering. Add it to the canvas and enter the web address of any hosted MCP server — your CRM, your calendar, a search provider, your own API — plus an access token if it needs one. Claude can then use that server's tools during the conversation and in an Advisor session. Servers that only run on your own computer, or that need you to sign in through a browser, can't be connected. See MCP connections.

Can I customise how the product looks? Yes. Choose from 20 themes, or adjust colours, fonts, text size, animations and the background (an image or an animated effect) in the AI Product node's Design tab. You can also choose how the product opens — a landing page, a title page, a chat widget or straight into the chat. For fully custom pages, design them in Claude with the connector.

Can my product run in another language? Yes. Each product speaks one language: English, German, French, Spanish, Italian, Dutch or Portuguese. Set it on the AI Product node under Conversation → Collection Instructions → Language, or ask Claude: "Set the language of [product name] to French." Auto (the default) follows the language your instructions are written in. The questions, the analysis and the result page all use it. See Understanding your AI Product.

Can I embed the product on my website? Yes. Product Settings → Deployment shows your public link and embed code at any time. They work for visitors once Publish Product is switched on in the same tab. To embed the product on your own site, add your site's domain to Allowed Domains for Embedding — if that field is empty, other websites can't embed it.

Can I use my Lovable, WordPress, Webflow or other website? Yes — no embed code required. Copy your product's public link from Product Settings → Deployment and add a button or link on your site that points to it. If you want the chat to start straight away with no welcome screen, open the AI Product node's Design tab and choose Direct Start under Welcome Screen.

Can I use my own domain? Yes. Connect a domain to a workspace in Settings → Client settings → Custom Domain, and every product in it lives at yourdomain.com/product-name. New products get their web address automatically, and you can change any product's address on the same page, or ask Claude. See Custom domain.

Will visitors see Productised? On a custom domain, the page title and the AI disclosure carry only your name. A few things show Productised until you replace them, product by product:

• the "made by" badge in the bottom-left corner — upload your own logo and link in Product Settings → Branding (paid plans), or ask Claude to set custom branding;
• the browser tab icon — set a favicon in Product Settings → Page SEO, or a default for the whole domain in Custom Domain settings;
• the link preview image when someone shares the link — set your own share image in Page SEO.

Workspaces and teams

What is a workspace? A workspace keeps its own products, leads, Brand Voice, variables, branding, Brain, custom domain, data retention and AI disclosure settings, so different brands — or different clients — stay separate. Your plan, credits, team, prompts and account Claude key are shared across the account, though any workspace can have its own Claude key. Each settings page says whether it applies to This client only or Every client on the account.

Can I invite my team? Team seats come with the Agency plan: 5 seats including you, so you can invite 4 people from Settings → Account & billing → Team. See Agency.

What can Members and Admins do?

• Members work on client workspaces: products, pages, leads (including deleting a lead), Brain, Brand Voice, variables, prompts and custom domains. They can also connect Claude.
• Admins can do everything a Member can, plus manage billing and Usage & cost, the team, Claude and API keys, data retention, AI disclosure and logs.
• Only the owner can delete the account. Owners and Admins can change other people's roles on the Team page.

Data, privacy and AI

Where is response data stored? In the United States: the app runs on Fly.io (Ashburn, Virginia), the database on Supabase (AWS us-east-1), and files on Cloudflare R2. See Legal.

Who can see responses? In the app, only people on your account. Each visitor's result page also has its own private, unguessable link, and anyone holding that link can open it — for example, when a visitor forwards their result.

Which AI services see my data? Products run on Anthropic's Claude. A few features use other services: AI Insights, Talking Points and the Growth Playbook (OpenAI), voice input in the chat (OpenAI), parts of Brain (OpenAI), files visitors upload in a chat (Supermemory) and web search (Parallel). Productised doesn't train AI models on your data, and Anthropic and OpenAI don't use API data to train their models.

Can I delete a lead? Yes. Open the product's Analytics → All Results table and delete the lead — it's a permanent delete, not hidden. Use this when someone asks to be erased. Files they uploaded, and anything already sent on by webhook or email, aren't removed by this.

Can leads be deleted automatically? Yes. In Settings → Client settings → Data retention (Admins only), choose how long a workspace keeps its leads, for example 365 days. Leads older than that are deleted in a daily clean-up. Leave it empty to keep leads indefinitely (the default).

How do I tell visitors they're talking to AI? In Settings → AI & connections → AI Disclosure (Admins only). Your landing, chat and result pages then show "You're engaging with AI. By continuing, you acknowledge our Privacy Policy and AI Notice." Link your privacy policy or write it in the app, and edit the AI Notice we draft for you. It's on for new products by default, and you tick which existing products show it. This helps you meet the EU AI Act's transparency rules.

Is AI-generated content marked? Anthropic is adding invisible, machine-readable marks to text its Claude models generate, so it can be identified as AI-generated. This doesn't change what your visitors see. See our AI Content & Disclosure Policy.

Do test runs count as leads? Runs from your product's demo link (ending in ?demo=true) are test runs. They show in your analytics by default, marked Test, so a new product looks populated when you demo it — switch on Hide test runs in a product's Analytics to leave them out everywhere in the workspace. Test runs never send lead emails, notifications, webhooks or Zapier data, and never use up a submission cap.

Can I export my data? Yes. From any product's Analytics, export all responses, form data, leads, or the full JSON. See Exporting data.

Billing and plans

What are credits? When Productised's own Claude key runs your product, each visitor conversation uses 1 credit, whatever it generates. Building products, Build with AI and Brain don't use credits. Your plan's allowance resets each billing period, and your balance is on Settings → Account & billing → Subscription, under the Usage tab. If you have your own Claude key, credits stop being spent altogether — see Credits.

Can I use my own Claude API key? Yes, on any plan. An Admin adds it in Settings → AI & connections → Claude API Key. To cover one workspace, switch to it and use the This workspace card on the same page. Your products' conversations and results then run on your Anthropic account and use no credits, as do products you build through the Claude connector. Brain, in-app Build with AI, voice input, AI Insights and Talking Points still run on Productised's services, at no credit cost.

The key isn't a fallback for when credits run out — it takes over from the moment it's saved, and Anthropic bills you for that usage directly. Your included credits then sit unused, as a safety net if you ever remove or rotate the key. To see what the key has actually cost, use Settings → Account & billing → Usage & cost, or ask the Claude connector about your usage.

How do I see what each workspace costs? Settings → Account & billing → Usage & cost shows the AI cost of each workspace by month, with a breakdown by product for the workspace you're in. Download the monthly figures as a CSV for client billing.

AI reliability and failures

What happens if the AI fails while someone is using my product? Their answers are saved. A lead is recorded as soon as the visitor sends their first message, and if the AI can't respond, the conversation so far is kept in your Analytics. The visitor sees a short, friendly message — never a technical error — and if they try again, it carries on the same lead rather than starting a new one.

Will I know if my product stops working? Yes. If your own Claude key is rejected, its Anthropic account runs out of credit, or your account runs out of Productised credits, Owners and Admins get an in-app notification naming the product and the reason in plain terms. You'll get at most one alert per product per hour. Visitors' answers are saved either way.

I use my own API key. What happens if it runs out of credit or is revoked? Visitors' answers are still captured, you're alerted, and visitors see a short "please try again" message. Top up or replace your key in Settings → AI & connections → Claude API Key — new messages use it straight away, with no need to republish. Productised never quietly switches to its own key when yours fails: your key means your billing.

Does a brief provider outage break my product? Usually not. In the conversation, rate limits and provider hiccups are retried automatically, and most resolve without the visitor noticing. Problems a retry can't fix, like an invalid key, fail straight away and alert you.

Can I choose which Claude model runs? On AI Agent nodes, yes — the model you pick there is used whichever key runs it. The AI Product conversation itself uses a Claude model chosen by Productised.