Discovery Call

What a Discovery call is

A discovery call is the one way to start an AI audit or a bespoke training programme. It is a short talk, not a proposal. Nothing is quoted or scheduled before it. What it does is establish which of your processes are in scope and who needs to be in the room.

It is the first step for both. An audit begins with it, and so does a bespoke programme. The set courses have published outlines and can be requested directly, which is why the conversation is the front door for the two things that have no standard shape.

Why it comes first

The audit page puts it in one line. The rate is published and the scope is not. What an audit is charged at is stated on the pricing page. What the conversation settles is what that rate applies to: which processes are worth putting inside the work, and how many days that comes to.

The reason the scope cannot be published in advance is given honestly on the audit page. It depends on what the first two steps find, and neither side knows that yet. So the conversation exists to agree a sensible starting boundary before anything is fixed.

What it establishes

The audit page lists three things. First, which processes are worth examining, and which are working as they are. Second, who needs to be in the room when they are examined. Third, whether an audit is the right instrument at all.

The third is the one buyers do not expect. If the answer is a training programme, or a decision the organisation can take without the business, the site says you are told so in the conversation and not after a proposal. The booking page repeats it: if the answer is none of the services yet, that is the answer you get.

How the thirty minutes run

The booking page sets out five stages. It opens on your side: what the organisation does and what you want AI to change about it. It moves to current processes, meaning which pieces of work repeat, which queue and which only run because one person knows something they have never written down. Then it names the processes worth putting inside a piece of work, and the ones worth leaving alone.

The fourth stage discusses the recommended route: AI training, an AI audit, AI automation, or none of them yet, and who needs to be in the room for whichever it is. The last stage is a recommendation and a scope, said out loud, with nothing committed on either side. There is nothing to prepare. The site’s advice is to say what your teams currently spend their time on, and that is enough to start.

What you leave with

The booking page lists what the call ends with, and is careful about what that means. It is said out loud. Nothing is quoted or scheduled during the thirty minutes, and no document arrives afterwards. The six outcomes are clear recommendations, priority opportunities, a view of where to start and what follows, suggested training options, suggested automation opportunities and practical next steps.

The last of those has an unusual line attached. The page says the next steps are what to do in the week after the call, whether or not any of it involves this business. That is a useful test of the conversation. It is meant to leave you with something usable, and not only with a reason to buy.

Who needs to be in the room

The second thing the conversation settles is who needs to be in the room. The reason appears in the first audit step. Mapping walks the operating processes end to end with the people who run them, in support, operations, administration and sales. The people who run a process know how work actually moves, which is rarely how the organisation believes it moves.

The audit page adds that nothing needs writing up in advance. A process description prepared for an audit describes the process as it is supposed to work, and that is not the process being audited. So the conversation is where a leader names the processes and commits the people who actually run them to being available.

What a Good scope looks like

A scope is a boundary, and a good one is narrow. The site’s article on what a well-scoped workflow delivers names four principles: a clear objective, a defined scope, built-in guardrails and a business impact that can be shown. It argues that a single, well-scoped workflow gives a result that can be measured, while a broad rollout usually does not.

The audit article makes the matching point about honesty. A genuine audit is willing to name the process that is not worth touching alongside the ones that are, and it starts with a conversation and not a quote. It advises asking to see the ranked list before committing to a tool, and treating an audit that will not say no to anything as a warning sign. The conversation is where that honesty starts.

Where it fits in an Audit

The audit runs five steps in order: map your processes, document how they run, identify the opportunities, test the first one and build and scale. Each produces what the next needs. The conversation comes before all of them and sets which processes the first step will map.

The audit produces process documentation, a register of the knowledge that sits with individual people, a prioritised roadmap, a compliance checklist for the EU AI Act and GDPR, and a tested prototype of the first opportunity. The AI audit page and the audit methodology cover the steps. The audit is a diagnostic and not a legal opinion, and processes outside the agreed scope are not examined.

Where it fits in a Bespoke Programme

A bespoke programme has no standard shape to publish, because the shape follows what the discovery turns up. The bespoke page says it is scoped through direct conversation only, and that duration, session count and content are all set there, so none of them is published in advance.

The published pattern has five steps: a discovery call, an AI audit where one is warranted, automation and agent implementation, training and ongoing support, and white-label delivery where it is asked for. The recommended entry sequence for training is a leadership briefing first and then all-staff foundations, so top-down commitment and bottom-up capability arrive together. The bespoke training page sets it out.

Not needed for a set Course

A core AI course is different. It has a published outline and a published shape, and the review of set courses against bespoke training describes it as the entry point for a team that already knows roughly what subject it needs covered and wants a proven session and not a conversation first. The core AI courses can be requested directly.

The conversation is still open to anyone unsure which route fits. If a team cannot say what it needs, the conversation is where the answer to that comes from. It is one way in and nothing needs choosing beforehand.

A worked example

This is an illustration, not a real client. A 40-person services firm asks for an audit because leadership feels behind on AI. In the conversation, the firm says its people spend a lot of time on client onboarding paperwork, monthly reporting and a shared inbox nobody owns. Six processes come up.

The conversation sorts them. Two repeat often and are slow, so they are worth examining. Three run well and are left alone. The sixth turns out to be a question about staff knowing what they may put into a tool, which a training programme answers better than an audit does, and the firm is told so on the call. The audit scope is the two processes, and the people who run them are named as the ones to be in the room. Nothing was quoted. The firm left with a recommendation and a boundary.

Written or booked

There are two routes in. The booking page is the thirty minute call. The talk to us page is for the other case, a question you would rather put in writing first. It asks you to say which of the three things the business does your question is about, and what your organisation does, and it reaches the same inbox.

Nothing is quoted or scheduled from the form either. The point of both routes is the same: to agree what is in scope before anyone commits.

Mix-ups worth avoiding

Four confusions come up often enough to name. The first is treating the conversation as a sales call with a quote at the end, when nothing is quoted or scheduled in it. The second is treating it as the start of a fixed package, when the scope is what the conversation exists to decide.

The third is treating it as required for everything. A set course can be requested directly. The fourth is treating it as a step that only ever leads to an audit. The site says the answer can be a training programme, a decision you can take yourself, or none of the services yet.

Where it goes wrong

The main failure is arriving with a tool in mind and no process. The site’s advice runs the other way: name the work your teams currently spend their time on. A second is arriving without the people who run the processes, because the map is built with them and cannot be built from a description.

A third is asking for a deployment decision before the process is known. The review of cloud against on-premise AI says a workload’s data and residency needs are exactly what a discovery call has to settle before anything regulated or commercially sensitive goes near a cloud tool. A fourth is treating the scope as fixed once it is agreed. It is a starting boundary, and the first two steps of the audit are what confirm it.

Where it appears on the site

The discovery call is the closing step of most of the site’s audit content, and the get an audit page is built around it. It is also where the ladder of Assist, Recommend, Execute and Never delegate gets applied to a business’s own processes, in the article on automating a task against automating a decision.

The related terms are opportunity identification, the step that ranks what the conversation puts in scope, and workflow design, what one scoped process becomes once it is redrawn end to end. The wider training catalogue is where the set courses sit.