Help Centre
Every question, answered in full.
The full list, rather than the flattering half of it. If yours is not here, write to [email protected] and it will be added.
Is this a CRM?
No, and it reads from yours. A CRM is the authoritative record of who your customers are; this is the layer that decides what should happen next on each opportunity and makes sure it happens. Most companies need both, and the two should not compete to be the master — two customer records diverge within a month and people trust whichever they opened last. If your problem is that nobody knows who a customer is, fix that first.
How is this different from a sales engagement tool?
The unit of work. A sequence is complete when it has sent; a revenue motion is complete when a meeting happened or money arrived. A motion carries a commercial goal, a qualification bar, an agent with bounded authority and a named human owner — and it changes channel when the customer's behaviour changes rather than sending step three on day seven regardless. If your outbound is genuinely high volume and low complexity, a sequencer is faster at that specific job and we would rather say so.
Will it send cold email to a list I upload?
It can run outbound, and that motion carries the most limits in the product: an audience you approve explicitly, a maximum of three attempts, hard stops at anything resembling a commitment, and consent checked per contact and per channel before anything sends. It will not contact somebody whose consent state does not permit it, and there is no setting that changes that. If what you want is volume cold email with minimal constraint, this is the wrong product and it will frustrate you.
Does the AI say it is an AI on calls?
Always, at the start of every call, and it is not configurable off. Calling hours are enforced in the contact's own timezone, attempts are capped, do-not-call lists are honoured and recording notices follow what you configure per country. All of that is checked by the platform before a call is placed rather than described to a model and hoped for.
What stops an agent offering a discount it should not?
A commercial authority ladder, enforced outside the model. Standard pricing an agent may send; up to 5% it may offer; between 5 and 10% a manager approves; above that the Commercial Director does. Contractual terms and service levels are a person's decision only, on every plan and in every configuration — if that could be unlocked under commercial pressure it was never a limit. Exceeding authority produces an approval request with the reasoning attached, not an improvised number.
Will it try to sell to a customer who has a complaint open?
No — that is one of the features we would point at first. With your service records connected, expansion and renewal outreach is held automatically while a critical case is unresolved, and the hold releases itself when the case closes. Their need for a second service will still be there next week; the relationship will not survive being sold to while something is broken. A sales tool that cannot see the complaint will recommend the outreach confidently, which is the problem.
Does it replace salespeople?
It is not designed to, and the product would be worse if it tried. What agents take is research, admin, remembering, routine follow-up and the low-value repetition — the work that gets dropped when somebody is busy. What stays with people is judgement, negotiation and the relationship. Nothing in the product measures a person by activity volume either: no view ranks anybody by emails sent, because that measures the wrong thing and teaches a team to hide things.
How do you know which campaign actually works?
By keeping the chain intact from the advert through the website visit, the agent's email, the human call and the proposal to the invoice and the payment. Leads is the wrong measure and this demonstrates why: one campaign produces sixty leads, another produces twenty and more collected revenue. Some revenue genuinely cannot be traced, and that share is shown as untraceable rather than distributed across channels to make the totals balance.
What does it cost to run a calling motion?
A plan price plus metered AI and telephony usage. Before any calling motion launches you see an estimate — runs, model usage, voice minutes and a total — and afterwards the actual against that estimate. A motion placing around 500 calls a month typically runs €40–€120 in usage. You should not have to model voice seconds to predict a bill, so the platform does that arithmetic and takes the risk of having done it wrong.
Can I see it before connecting my customer data?
Yes, and we would suggest it. There is a demo workspace with three fictional hotel groups, real-looking intent, an approval request waiting and an account deliberately on hold because of an open complaint. You can send an email, simulate a call, approve a discount and book a meeting against data that belongs to nobody. It includes the awkward cases on purpose — a demo where everything is going well shows the least interesting version of the product.
Will our data be used to train models?
No. Your conversations, customer records, call transcripts and commercial data are not used to train models made available outside your account. That is a contractual commitment rather than a setting, and the same prohibition binds the model providers we use.
What happens when the AI does not know something?
It says so and hands over. Agents answer from knowledge your company has approved and cite it; asked something outside that, they escalate with what they have already established. Where two approved sources disagree, an agent stops and asks rather than choosing — choosing would mean inventing a company position, in writing, to a customer.
How long before it does something useful?
An afternoon to a first motion, usually inbound lead follow-up, and often a result within the first week — an agent answering an enquiry and booking a meeting while you were doing something else. What an afternoon does not get you is a configured authority ladder and a full playbook across five motions; those need decisions from whoever owns pricing, which is a conversation rather than a setup step.
Can we get our data out?
Yes, on every plan, throughout the term and during any wind-down: opportunities, motions, conversations, call transcripts, engagement history and attribution records. Your customer records were never ours in the first place — they stay in your customer system and this platform references them.
Does this make us compliant with marketing and calling rules?
No. It enforces the rules you configure — calling hours, disclosure, attempt caps, consent, opt-outs — deterministically and before anything is sent. Configuring them correctly for your market is your decision and needs your own advice. No configuration of any software makes an organisation compliant with anything, and a product claiming otherwise would be selling you a false sense of security in an area where the penalties are real.
Make sure every opportunity gets a next action.
Let AI handle the repetitive engagement, keep follow-up moving, and bring your team in when a customer needs a real conversation.
Agents identify themselves on every call · Nothing commits without you · [email protected]