Frequently asked questions

No. There are no automation builders, no flowcharts, and nothing to configure in a scripting language. You pick an assistant role from the library, answer a short set of plain-English questions about your business and how you want it to sound, connect the apps you already use, and it starts working. Everything after that happens in normal conversation and in a review queue.

Minutes, not weeks. Setup is self-serve: you sign up, choose a plan, hire an assistant from the library, and walk through its setup wizard — typically 10–15 minutes of questions. Your assistant is provisioned into your workspace as soon as you finish. There's no waiting list, no scoping call, and no engineer building anything behind the scenes before you can use it.

Not currently — we'd rather be straight with you than run a trial that quietly auto-charges you later. Every plan is a rolling monthly subscription you can cancel at any time, so the entry tier is the low-commitment way in: take a month, hand over your most draining task, and see what comes back. If it isn't for you, cancel before your next renewal date and that's the end of it.

The roles with a full self-serve setup today are the Social Media Assistant, Blog Writer, Lead Generation Assistant, CRM Data Assistant, Accounts Receivable Clerk and Minute Taker. Other roles in the Assistant Library are at different stages — anything still in private beta is clearly marked Coming Soon and offers a waitlist rather than letting you hire it. We'd rather show you the roadmap than sell you a role that isn't finished.

A human VA works set hours, needs training, and costs many times more. Your digital assistants work to a schedule you set, come pre-trained for their role, track their own goals, and never go on holiday — while you keep final approval on everything that goes out the door. The honest trade-off: a human VA can use judgement in situations nobody briefed them on, and your assistant can't. It's very good at the repeatable work that eats your week, and it always shows you its output before it commits to anything.

When you first set up your assistant, you'll go through a short onboarding process where you share a few things: your brand name, the tone you want (friendly, professional, bold, etc.), examples of content you love, and who your customers are. Think of it like briefing a new team member on their first day. From there, your assistant picks up patterns from everything it produces for you. The more you use it — and the more you approve or tweak its output — the sharper it gets at sounding like you, not a generic robot.

Not at all. The initial setup takes around 10–15 minutes, and from there your assistant works from the preferences and examples you gave it. You won't be sitting at a dashboard feeding it instructions every morning. That said, the more specific feedback you give it over the first couple of weeks, the faster it dials in. After that, most customers find they barely need to touch the settings.

This will happen — especially early on, and that's completely normal. Every piece of content your assistant produces lands in a review queue for you to approve before anything is published or sent. Nothing goes live without your approval unless you've specifically turned on automatic publishing for a particular workflow. If something's off, you reject it with a note explaining why, and the assistant regenerates the piece using that feedback. Think of it as a first draft from a very fast copywriter, not a final decision-maker — and check anything factual, because AI output is confident even when it's wrong.

Your assistant can do both — it's up to you. For recurring work (say three social posts a week, or a blog post every fortnight), you set the cadence once during setup and your assistant drafts to that schedule on its own. You don't need to be logged in. For one-off jobs or anything that needs your specific input, you can open a chat and brief it whenever you like. Most customers set up their routines once and then check the review queue a couple of times a week.

Think of it as a second pair of eyes built into the process. Before your assistant's content reaches you for final approval, the Quality Reviewer checks it against your brand guidelines — flagging anything that sounds off-brand, factually shaky, or inconsistent in tone — and can suggest a rewrite you apply with one click. It's the difference between handing you a rough draft versus a polished near-final draft, and it saves a lot of back-and-forth on customer-facing content.

On Tier 1 and Tier 2, your assistants are designed for individual or small-team use within a single workspace. Tier 3 adds workspace collaboration, so your whole team — account managers, content leads, ops staff — can work through the same assistants with shared context. If you're running an agency or a team of more than a handful of people, Tier 3 or our Custom Enterprise option is the right fit.

You do. As between you and us, you own all rights in the outputs your assistants generate from your inputs, and we assign to you any rights we might otherwise have in them. Two honest caveats, both set out in our Terms of Service: AI outputs aren't guaranteed to be unique, and how far copyright protects AI-generated work is still unsettled law in the UK and US. Before you publish anything commercially important, it's on you to check it doesn't tread on someone else's intellectual property — the same check you'd do on a freelancer's work.

Emails your assistant drafts and sends on your behalf carry a short AI disclosure footer — this is a transparency requirement under the EU AI Act, and it's applied by the platform rather than left to the assistant to remember. You can customise the wording (as long as it still makes the AI involvement clear), and disclosure settings for other content types live in your workspace under each assistant's guardrails. If disclosure matters to your industry, ask us before you sign up and we'll walk you through exactly what gets labelled where.

We support a curated list of integrations rather than a "thousands of apps" directory — each one is built and maintained properly instead of half-working. Today that covers social and publishing (Facebook, Instagram, LinkedIn, X, Threads, TikTok, YouTube, WordPress.com and developer blogs), email and messaging (Gmail, Outlook, Slack, Intercom, Zendesk), business systems (HubSpot, Salesforce, Notion, Xero, QuickBooks) and creative tools (Canva). The live list — and which of them your plan can use — is on the Integrations page inside your workspace. If a specific tool is make-or-break for you, ask us before you sign up and we'll tell you straight.

Each account you link to Be More Swan counts as one connection. Connect Gmail, that's one. Add Instagram, that's two. Every individual account is its own connection, so linking both a Gmail inbox and an Outlook inbox uses two slots rather than one. Your plan's connection allowance is shown on the pricing page and in your workspace.

Standard Apps are the everyday tools most small businesses already use — email, social platforms, messaging, and scheduling. Premium Apps are the specialist, higher-complexity integrations: CRM platforms like HubSpot and Salesforce, finance systems, and connections into your own internal software. Premium Apps need more setup and ongoing maintenance on our side, which is why they unlock from Tier 3 onwards. If you're unsure which category a tool you rely on falls into, ask us before you sign up.

Nothing breaks and nothing gets disconnected without your knowledge. You'll simply receive a notification letting you know you've reached your plan's limit. Your existing connected apps continue working normally. To add another connection, you'd either need to disconnect one of your current apps or upgrade to the next tier. We'll never silently disconnect something — you're always in control.

You can swap them out at any time from your workspace. If you connected Instagram during your first week but want to replace it with LinkedIn next month, just disconnect Instagram and connect LinkedIn — it counts as one change, not an additional slot. Your total number of active connections just can't exceed your plan's limit at any given time.

Connections do occasionally break — you change your password, a platform expires an access token, or you revoke permissions from the platform's own settings. When that happens the connection is badged in your workspace as needing attention and we notify you; anything queued for that account waits rather than failing silently. Reconnecting is a two-click job from the Connections area. It's worth glancing at your connection statuses if a scheduled post doesn't appear when you expected it.

Our integration library covers the most widely-used business tools. If your tool isn't on the list, it doesn't automatically mean we can't support it — it may just require a custom integration, which falls under our Custom Enterprise professional services. Get in touch and we'll be straight with you about what's possible and what it would involve.

One task is one piece of work your assistant does for you: drafting a post, writing a blog article, reviewing a draft for quality, rewriting something you rejected, or researching a lead. Messages you send your assistant in chat also count — each one puts it to work, so each one draws from your monthly allowance. What doesn't count: browsing your dashboard, approving or scheduling something that's already drafted, changing settings, or connecting an app. There's no daily cap and no per-message limit; the only ceiling is your plan's monthly allowance, and your live usage is always visible in your workspace.

You'll receive a notification at 80% of your monthly allowance, telling you how many tasks you have left. At 100% your assistants pause until the allowance resets — you'll be prompted to upgrade or wait for the reset. The limit is a hard ceiling, not a meter: we never charge you for going over it, so the bill you agreed to is the bill you get. Work already drafted and waiting in your review queue stays there, and upgrading lifts the pause immediately.

No — task allowances do not roll over. Your allowance resets on the 1st of each calendar month (this is separate from your billing date, so a mid-month subscriber gets a full allowance straight away). AI media credits are different: those do roll over.

Tasks cover the thinking and writing work. Media credits cover generated imagery and video, which cost considerably more to produce — one credit per generated image. Every plan includes a monthly allocation of credits (the amount is on the pricing page), AI video generation unlocks from Tier 2, and unlike tasks, unused credits roll over rather than expiring at month end. Not every image needs a credit, either — your assistant can also pull from stock libraries or from photos you've uploaded yourself.

No. Every plan is a fixed flat fee — there are no per-task or per-compute charges, and going over your monthly allowance pauses your assistants rather than billing you more. Your bill is exactly the same amount every month unless you choose to change plan.

No. All plans are rolling monthly subscriptions. Cancel or upgrade at any time — no notice period, no penalties, and no phone call to sit through. Cancelling is not refunded for time already paid: you keep access until the end of your current billing period, and you simply won't be charged again after that.

It means there's no fixed end date and no long-term commitment. You pay month to month, and your plan renews automatically each month on the same date you originally signed up. If you cancel, your access continues until the end of that billing period — you won't be charged again after that, but the payment you've already made for the current period is not refunded. There are no cancellation fees and no exit penalties.

Yes, from the Billing page, and the two directions work differently on purpose. Upgrades take effect immediately — your new assistant slots, connections and allowances unlock the moment you confirm, and you receive a prorated credit for the unused portion of your current period against your next invoice. You'll see a preview of exactly what you'll be charged before you commit. Downgrades take effect at the start of your next billing period, so you keep everything you've paid for until then, and you can cancel a scheduled downgrade any time before it lands.

Yes — the annual plan is billed as a single payment at the start of your year, and in return you pay 20% less than the monthly rate (roughly two and a half months free). If you'd prefer not to pay a lump sum, the monthly rolling contract is always there. If you cancel partway through the year, your access continues to the end of the annual period you already paid for, but the remaining months are not refunded — so pick annual only if you're confident, and start monthly if you're not.

Yes. Email billing@bemoreswan.com and we'll move you across and account for the part of the month you've already paid for. We'll confirm the exact figures with you in writing before anything is charged — you won't lose time you've already paid for.

Our plans are priced in pounds sterling. The pricing page can display approximate equivalents in other currencies to help you compare, but those are indicative — the exact amount and the exact currency you're charged are shown on the checkout page before you confirm payment, and that's the figure that matters. If your card is issued outside the UK, your bank may apply its own conversion rate.

Prices on our pricing page are shown exclusive of VAT. The full amount you'll be charged is always displayed on the checkout page before you confirm payment. You can store your legal company name, address and VAT/Tax ID in Billing so they appear on your invoices. If you need a specific VAT treatment for your jurisdiction — a reverse charge, for example — email billing@bemoreswan.com before you subscribe and we'll sort it out with you directly.

Every payment generates an invoice automatically, and your full invoice history is available to download from the Billing section of your workspace at any time. Add your legal company name, address and VAT/Tax ID under Business Information first and they'll appear on every invoice we issue from then on.

We'll email you immediately if a payment doesn't go through, and your account enters a 7-day grace period during which everything keeps running normally while you update your card. If it still hasn't resolved after that, your assistants stop working — but your data, settings and drafts are fully preserved. As soon as a payment succeeds, everything picks back up exactly where it left off.

Yes — if you're going quiet for a season, going on leave, or just don't need the output this month, you can pause your account from the Billing page instead of cancelling outright. Your workspace, brand settings and content stay exactly as they are, ready to pick up when you come back. You'll be offered the pause option as part of the cancellation flow, so you don't have to decide before you start.

From the Billing page in your workspace, in a few clicks — no notice period, no cancellation fees, and nobody will try to keep you on a call. Your assistants keep working until the end of the period you've already paid for. Cancelling stops future billing; it doesn't delete your data, so if you also want your workspace erased see What happens to my data if I cancel? below.

We do. You get a personal referral link from your workspace, and every person who signs up through it and starts a paid plan earns you a token. One token is worth £10 of credit; bank five and you unlock a free assistant for a month. There's no cap, no sales call, and you can track conversions in your reward vault. Full details on the Referral Program page.

Your workspace lives in its own secure sandbox with tenant isolation enforced at the database layer — your data is never pooled with, or visible to, any other customer. Everything is encrypted in transit (TLS 1.3) and at rest (AES-256), the OAuth credentials connecting your business tools get an additional layer of AES-256-GCM encryption and are never logged or returned to a browser, and payments are handled entirely by Stripe (we never see your card details). The full detail, including our sub-processor list, is in the Trust & Security Center.

Every assistant is governed by the Be More Swan Safe Content Benchmark — a built-in safety layer applied to everything your assistants create. It runs at the highest priority in the system and cannot be disabled, overridden or bypassed by any other setting, including your own instructions. It's included on every plan, and combined with the review queue it means you stay in control of what carries your name.

Only if you explicitly configure it that way. By default every output waits in your review queue until you approve it. Once you're confident in the output quality you can switch specific workflows to publish automatically — and even then there are conditions that always force a human review, such as anything using an AI-generated image. You can turn automatic publishing off again at any time.

No. Your data — including any emails, documents, messages, or content your assistant handles — is never used to train public AI models or shared with third parties for model development. That's not just our policy: our contracts with the AI providers we use carry an explicit no-training-on-API-data clause. What happens in your workspace stays in your workspace.

We build on leading commercial foundation models rather than training our own from scratch, accessed through their business APIs under data processing agreements. Anthropic's models do the main work — drafting, chat, and quality review. We also use OpenAI for content moderation screening and tour narration, and a specialist provider for AI image and video generation. That means you get frontier-quality output, and we can move to a better model as they improve without you having to do anything. Each provider, what it processes and the legal basis for the transfer is listed in our Privacy Policy, and a full sub-processor register is available on request.

In normal operations, no human at Be More Swan reads your connected app data. Access is handled programmatically by the assistant — software, not staff, is doing the reading and writing. The only exception is if you raise a support ticket and grant us temporary, scoped permission to investigate a specific issue. Even then, access is logged in an immutable audit trail, time-limited, and confined to the minimum data needed to resolve your problem.

Your workspace database is hosted in the UK/EEA. Being straight with you: some processing does happen outside the UK, because the AI model providers, our payment processor and our email provider are US-based. Those transfers run under Standard Contractual Clauses (or the UK–US Data Privacy Framework in Stripe's case), backed by a documented Transfer Impact Assessment and data minimisation on what's sent to model APIs. Every sub-processor, its location and its transfer basis is listed in our Privacy Policy. If you're in a regulated industry with strict residency requirements, talk to us first.

Yes — it's your data and your right under GDPR. Request a personal data export and we'll prepare a machine-readable file covering your account, assistants, content, connections and billing records, with a secure download link valid for 24 hours. Exports are limited to one per day to keep them secure. Email privacy@bemoreswan.com from your account address if you can't find the option in your workspace.

Cancelling your subscription and deleting your data are two separate things, deliberately. Cancelling stops the billing and stops your assistants; your workspace and content stay put in case you come back or need to export something. When you ask us to delete your account, there's a 24-hour cooling-off window with a cancel link in case it was a mistake, after which your assistants, connections, drafts and published-post records are erased — permanently completed within 30 days, as UK GDPR requires. We keep only an anonymised record of the erasure itself and anything we're legally required to retain, such as billing history. Posts already published to your own social accounts stay live on those platforms; only you can remove those. Full detail on the Data Deletion page.

Yes. If you're processing your own customers' personal data through Be More Swan, you may need a DPA in place for GDPR compliance. Every agreement you've accepted with us — terms, privacy, DPA and the responsible AI use agreement — is listed with its version and date under Settings → My Agreements in your workspace. To request our standard DPA, or to negotiate a custom one as an Enterprise customer, email privacy@bemoreswan.com.