AI strategy

AI audit for your business: what it is, what's included and what it costs

An AI audit looks at how your company works today and shows where AI would save your people real time, and where it would not yet. Here is how Obinova runs one, what you get at the end and how to tell whether it is what you need.

What an AI audit is and what it is for

Many people hear "AI audit" and picture a technical review of their IT systems, or a thick report nobody opens again. The audit Obinova runs starts from your team's daily work and ends with a ranked list of use cases and a roadmap with owners and indicators, which we revisit several times a year. At an estate agency in Bristol, for instance, it would start with three things: two negotiators writing listings in ChatGPT with their own prompts, an administrator copying portal enquiries into the CRM by hand, and a director with two quotes for "AI-powered" software he cannot compare.

An AI audit is a structured review of your business processes with a practical aim. It identifies the tasks where AI could help your team, and the ones better left alone for now.

The output is a set of ranked use cases. A use case is one specific task where AI is applied, such as sorting the documents clients email to an accounting firm or drafting replies to an online shop's customer questions. Each comes with its expected impact, what it needs and how you will measure whether it works.

Choosing well matters. A Harvard Business School and BCG study of 758 consultants found that those using AI produced work of over 40% higher quality and finished 25% faster. The same study found that when AI was used outside its strengths, results got worse. An audit tells those cases apart before you spend money.

Do you need an AI audit, training or automation?

Not every company should start with an audit. McKinsey's The State of AI 2025 reports that 62% of organisations are already experimenting with AI agents. Knowing what to do next is harder, and the right first step depends on which of these four situations you are in.

  • You can't name three tasks with figures: audit

    If the answer to "where would AI help us?" is "lots of places", with no specific task, volume or time attached, the diagnosis is missing. The audit builds it with your team.

  • People already use AI, each their own way: training first

    A few colleagues use ChatGPT or Copilot on their own and results vary widely. Training built on their own tasks usually pays off sooner. An audit can follow for the processes that cross departments.

  • The task is already clear: automation

    If you know which process you want to change, such as logging portal enquiries in the CRM, and how to measure it, you can move straight to building the workflow in n8n.

  • Leadership needs a plan it can stand behind: audit

    When an investment has to be justified to partners or department heads, a document with use cases, estimated effort and a clear order of work makes the decision easier to take and to explain.

How an AI audit works, step by step

Every mission follows the same sequence. What changes from one company to the next is how many processes and teams are in scope.

  1. 01

    Kick-off meeting

    We agree the objectives with leadership, choose the departments to review and name a contact in each team. We also set the working calendar.

  2. 02

    End-to-end process mapping

    A process map describes, step by step, how a request comes in, whose hands it passes through and how it gets closed. We build it by talking to the people who do the work, noting the tools, the time spent, the repetition and where information gets lost.

  3. 03

    Finding and ranking use cases

    Each opportunity is scored on impact, such as time saved, errors avoided or better customer service, and on feasibility, based on the data available, your current tools and the effort needed to set it up.

  4. 04

    Roadmap

    The roadmap is the phased plan that comes out of all this. It shows what comes first, which tool it uses, who owns it and how you will measure whether it works.

  5. 05

    Follow-ups through the year

    We come back several times a year to review what has been put in place, adjust priorities and add new use cases. AI tools change quickly and the roadmap has to keep pace.

What you receive at the end of the audit

  • A map of your processes

    A clear description of how the reviewed processes run today. Teams often use it for decisions that have nothing to do with AI, such as reorganising tasks or reviewing a piece of software.

  • Ranked use cases

    Each one with its expected impact, its feasibility and what is needed to get it running.

  • A phased roadmap

    A realistic order of work, with named owners and one indicator to track the progress of each case.

  • Tool and training recommendations

    Which tasks suit assistants such as ChatGPT, Claude, Gemini or Microsoft Copilot, which call for an n8n automation, and what training each team needs.

  • A presentation session

    A meeting with leadership and team leads to go through the findings and agree the first steps.

Example

What an audit might uncover at the Bristol agency

Back to the Bristol agency, with eight people across sales and admin. The case is hypothetical. At the kick-off, the scope is narrowed to two processes: handling enquiries from the property portals and publishing new listings.

Mapping the enquiries reveals something nobody had measured. Portal alerts arrive by email, the administrator moves them into the CRM when she has a moment, and the negotiator for that area hears about the lead hours later. On the listings side, each negotiator writes in their own style and the vendor is asked for the same property details several times.

The roadmap proposes three steps. First, an n8n automation that logs every enquiry in the CRM and alerts the right negotiator. Next, a short training session with a shared template for writing listings in ChatGPT or Claude. Later on, a single form so vendors provide the property details once.

Each step has its indicator. For enquiries, the time between the alert arriving and the first call. For listings, the time spent writing and the number of corrections before publishing. At the first follow-up, both are compared with the figures measured during mapping.

How long an AI audit takes, team involvement and price

The length depends on scope. Reviewing one department of an estate agency takes less time than covering sales, finance and customer service across a company with several sites. The calendar is set out in the quote and confirmed at the kick-off.

Your team mainly takes part in interviews and mapping workshops. We ask each person for a limited amount of time and schedule it around their day-to-day work. The people who do the job every day know the exceptions, the workarounds and the tasks nobody sees from outside, and the process map is built from what they tell us.

We don't publish fixed rates because every mission is different. The price depends on scope (one process, one department or your whole operation), the number of areas reviewed and how many follow-ups you want through the year. If you would then like us to build automations or train your teams, we set that out separately so you can decide on each part. The quote follows the first meeting, once we understand your situation.

How to prepare before the kick-off meeting

A few hours of preparation make the audit faster and more useful. This is what to have ready.

  1. 01

    Choose a lead contact

    Someone in leadership who can sign off priorities, arrange access to tools and has time set aside for the mission.

  2. 02

    Keep a task log for one week

    Ask each team lead to note their team's repetitive tasks, how often they happen and roughly how long they take. A shared spreadsheet is enough, and it is the best starting point for the process map.

  3. 03

    Gather current AI use

    Which tools people already use, on which accounts and for what. Include informal use, because it usually shows where the demand is.

  4. 04

    List your tools and sensitive data

    Email, CRM, accounting software, spreadsheets. Note as well which information must not leave the company.

  5. 05

    Write down what you hope to achieve

    One or two sentences per area, such as replying to customers sooner or closing the month with less overtime. They help set the objectives on day one.

Frequently asked questions

What is the difference between an AI audit and AI consulting?

The audit is the starting point. It reviews your processes, finds use cases and ranks them. Consulting then supports you while you put them in place. At Obinova both are part of the same mission, with a kick-off meeting and follow-ups several times a year, so the findings actually get put to use.

How much does an AI audit cost?

It depends on the scope, the number of processes and departments reviewed, and how often you want follow-ups. That is why we have no list price. After the free 30-minute first meeting, we send you a quote that sets out the scope, the calendar and everything you will receive at the end of the mission.

How long does an AI audit take?

Scope decides it. An audit focused on one small team is shorter than one that covers several departments. The calendar is agreed at the kick-off meeting and written into the quote. Follow-ups then take place at intervals through the year to review progress and adjust the roadmap.

Is an AI audit worth it for a small business?

It usually is if you have several ideas and no clear priority. In a small company the scope is narrower, because there are fewer processes and fewer people to coordinate. If you already know exactly which task you want to automate, you may not need one and can go straight to building the workflow or training the team.

Do we need technical skills in-house for an AI assessment?

No. We need access to the people who do the work and one contact in leadership. We turn what we observe into use cases and technical requirements. If your team wants to become more independent, the roadmap can include training on ChatGPT, Claude, Gemini or Microsoft Copilot, delivered by video or in person.

What happens after the audit?

You decide how to carry out the roadmap, with your own team, other providers or us. Obinova also builds automations and AI agents in n8n and runs training for your staff. During the follow-ups we review what has been put in place, check the agreed indicators and update the priorities.

Let's talk about your AI audit

Book a free 30-minute video call. Tell us how your team works and we will tell you whether an audit makes sense for you, and with what scope.

Free 30-minute diagnosis →
Free diagnosisWhatsApp

WhatsApp · Reply within 2 hours during business hours