Yes, It Is Time To Adopt AI. But First, Slow Down and Think It Through

Since February 2023, I have been running AI readiness sessions, executive briefings, and practical masterclasses with businesses across the UK, covering everything from solo consultancies to organisations with hundreds of employees.

And over time, the question has changed from “what can AI do” to “where do we actually start, without getting it wrong?”.

A short while ago I sat down, virtually, with a senior team leader from the professional services sector. She was thoughtful, grounded, and by her own admission not particularly technical. She was looking for a sensible way forward.

We had such a brilliant conversation, one of this discussions that reminded me why I do this work. Her questions were exactly the right ones. Her caution was entirely appropriate. And her plan, to try things out herself first and then bring her team along in small chunks, was very sensible indeed. Let’s take a closer look together.

Five practical points came out of our discussion. Each one applies just as much to a 46-person law firm as to any small or medium-sized business.

• Start where you are. Before reaching for new tools, get more from what you already have. Most businesses are sitting on underused AI features inside software they already pay for.
• Security is a design principle, not a barrier. AI adoption in regulated or client-facing work needs data governance from the start, not bolted on later.
• The whiteboard moment matters. Mapping your workflows and wishes before speaking to a tech vendor changes the conversation entirely, from passive recipient to informed buyer.
• People come first. The human side of AI adoption, fear of job loss, resistance to change, the need for reassurance, is not a soft issue. It is the central one.
• Patience is a strategy. AI gets better over time. Going slowly is not a weakness.

These principles are what I see play out in session after session when I help organisations like yours work through this.

1. How AI became everybody’s business

When ChatGPT arrived in November 2022, I remember the jolt of it. I also remember the reactions from the media, it was quite something to hear all the negatives and so little about the positives.

I’ve been working in digital since the mid-1990s, my first website predates Google, and I’ve always known AI existed, but I had assumed it was the preserve of universities and large corporates. Something way outside of most budgets and complex, something simply not for the rest of us.

ChatGPT was like a rock thrown into a pond. Suddenly the ripples reached everyone. Then came Gemini, Copilot, Claude, Perplexity, a race between platforms, each generating more noise. My response was to make it my mission to bring some clarity to all of it, using the podcast series I c0-host as a vehicle to capture my findings and reflections.

This AI solutions have complemented very naturally the 400+ apps and solutions I’ve reviewed to date, always looking for ways to make this world clearer and more accessible.

“I made it my job to bring clarity and sense to all this noise. That’s still what I do.” Pascal

The professional I spoke with had a similar moment of discovery, not with ChatGPT but with a Teams transcription tool called Otter that appeared unannounced after an internal meeting. Her reaction was immediate and instinctively right: she paused, asked her IT provider where the data was going, and switched it off. That single moment of caution is worth more than any number of AI webinars.

2. The trap of chasing productivity

Here is one of the great ironies of the current AI moment: the businesses most eager to adopt AI in the name of productivity are often the ones creating the most risk for themselves. If you rush to automate before you understand your workflows, your data obligations, and your team’s readiness, you may end up with a worse client experience, or worse.

“Caution and prudence are what you should use first, not the chase for productivity at all costs.”

This is particularly true in regulated sectors like law, finance, social care, and healthcare, where client confidentiality is a legal and ethical obligation, not just good practice. The question is not “can AI do this?” It is “should we use AI for this, and under what conditions?”

The professional I spoke with put it plainly: client data confidentiality is sacrosanct. That framing is exactly right. AI adoption in professional services should always start with data governance, not as a box-ticking exercise but as a genuine expression of how you run your business.

3. The whiteboard moment: mapping your wishes before you speak to anyone

One of my strongest recommendations, and something I take every client through at the start of an engagement, is what I call the whiteboard moment. Before you pick up the phone to your case management provider, your IT support, or any AI vendor, gather your team and draw it out. Map the journey. Describe the friction. Articulate, using plain language rather than technical knowledge, what you wish could happen differently.

“Using logic and common sense, not technical knowledge, describe what would be helpful. What would let us reclaim time, and what would we do with that time once we had it?”

This does two things. It gives you a clear description of your needs that vendors can respond to. And it shifts the conversation from “what can your technology do?” to “here is what we need, can you help?” That is a different and more useful position to be in.

In our conversation, the wish list that came out was concrete and modest: transcribe client meetings securely, turn recordings into attendance notes and letters, automate update calls, streamline reception. None of this is complicated. Most of it, it turns out, is already possible within platforms these businesses already use.

4. Squeeze the lemon: get more from what you already have

Before you pay for anything new, get more from what you already have. Most organisations are using perhaps a third of the capability of their existing software. Microsoft 365, Google Workspace, your case management system, your CRM, all of these have been quietly adding AI features, often without making much noise about it.

Why the silence? Partly because tech companies worry that if everyone piles onto AI at once, systems buckle. Partly because they are still testing. But the features are there, waiting to be found. I see this every week in my training sessions: people are genuinely surprised by what their existing tools can already do.

“My way of thinking is: exhaust and squeeze value from what we have currently. Once we reach a limitation, then we look at other options.”

In practice this means contacting your software provider’s customer success team. Ask for a review of what AI is already available within your current licence. Be prepared to be surprised, and perhaps a little annoyed that nobody mentioned it sooner.

For the legal professional I spoke with, the first place to look was clear: her case management system, LEAP, and her Microsoft 365 environment. Both have AI worth exploring before any third-party tool is considered. Similarly, the voice recordings her team already makes on their phones come with native transcription on most modern devices. The technology is already in their pockets.

5. The data question: know your borders

For businesses handling personal or confidential client data, the question of where that data goes is not optional. This came up repeatedly in our conversation, and rightly so.

The safe ground for most UK businesses is the established, well-governed platforms: Microsoft 365 and Google Workspace. These have clear commitments to data residency and GDPR compliance. They are the ones you can be certain will not create problems.

Third-party AI tools, even good ones like Whisperflow or NotebookLM, work differently. Many distribute data across servers in multiple regions, including outside the UK and EU. For some businesses this is manageable, particularly if data is not kept after the session. For others, especially in legal, financial, or healthcare contexts, it may be a hard line.

A practical workaround I have used with clients in regulated sectors is this: use NotebookLM, Google’s free platform originally built to support neurodivergent students, to upload an audio file, generate a transcript, turn it into a meeting report, and then delete everything before closing the session. If nothing is kept, the risk drops sharply.

6. The people dimension: the second half of the story

This is where I always come back, because it is where AI adoption most often fails. The technology is, in many ways, the straightforward part. The people, their fears, their habits, their need to feel valued and secure, that is the real challenge.

When automation frees up time in your organisation, you need to have an answer for what comes next. If you announce that AI will handle certain tasks without explaining what that means for your people, the first thought in many minds will be: am I going to be replaced? If you cannot answer that question clearly and honestly, you will get anxiety, resistance, and eventually a failed adoption.

“You have to be ready with the second half of the narrative: here is what you will do with the time we reclaim together. Not a threat. An invitation.”

The question I encourage leaders to ask their teams is this: if you could get back one or two hours a day from the repetitive work, how would you want to use them? Invite people to come with their own answers. Create the space for an open conversation. This reframes AI not as something being done to people but as something that frees them to do the work that actually needs human judgement, empathy, and expertise.

The professional I spoke with had this instinct naturally. She does not impose things on her team. She takes them on journeys. She plans to try things herself first, then expand to her team in manageable steps. That is the right approach.

7. The 90-day mindset: why patience pays off

I give every organisation I work with a 90-day roadmap where the first phase involves no implementation at all, just investigation, observation, and mapping. This is not time wasted. It is time well spent.

AI gets better over time. The platforms improve, the integrations mature, and your own understanding deepens. The people side, which takes the longest to get right, has time to develop. There is no advantage in going fast.

“Going slowly is the best approach. AI improves over time, and so does your team’s readiness. Patience is a strategy, not a weakness.”

This runs against much of the noise in the AI space, where urgency and fear of missing out are used to push hasty technology purchases. My advice is straightforward: ignore the competition between AI platforms. That is their problem, not yours. Focus on your workflows, your people, and your clients. Everything else follows from that.

In collaboration with my digital assistant, AI-LEEN, I’ve reviewed the conversation and explored the questions, the actions and the pitfalls, you, as a business leader and senior marketer, might need to consider and act on.

These are the questions I would invite every small business owner to consider before taking another step on their AI journey.

“Am I moving at the speed of my tools, or the speed of my team?”

Technology can move faster than people. The risk is not that AI will overtake your business. It is that your adoption strategy will outpace your team’s readiness. Have you had honest conversations about what AI means for each person’s role? Have you listened to their concerns and hopes?

“Do I know what I already have?”

Before spending anything on new AI tools, can you honestly say you have looked at the AI features already inside your current software? When did you last have a proper conversation with your software providers about what is possible now and what is coming next?

“What is the second half of my story?”

When automation saves time in your organisation, what will people do with it? This is not a trivial question. It is the difference between AI adoption that lifts your whole team and adoption that quietly breeds resentment. Have you worked out your answer?

Hold a whiteboard session

Set aside two hours with your key colleagues. Map out your most time-consuming and friction-heavy workflows. For each one, ask: what do we wish could happen differently here? Write the wishes down without worrying about how they would work. The output is your specification, the foundation of every conversation you have with a tech provider from this point.

A tool like Miro or FigJam works well, as does a physical whiteboard. Keep the session to two or three workflows. Record everything, including the rough ideas.

Book a “what’s possible” call with your software providers

Contact the customer success or account management team at every significant software platform you currently use. Ask for a conversation on two things: what AI features are already available within your current licence, and what is on the roadmap in the next six to twelve months. You may be surprised, and you should be prepared to be a little annoyed that nobody flagged this earlier.

When you email them, use this framing: “We are developing our AI readiness plan and want to understand what is already available to us before we look elsewhere.” That positions you as an informed buyer.

Introduce a regular team learning habit

One of the most underrated ways to build AI confidence in a team is shared, informal experimentation. Rather than asking people to learn AI on their own, which rarely sticks, create a regular session: a Friday lunch where colleagues try out a specific feature together, share what they find, and build confidence as a group. Start with something immediately practical, like the native transcription feature on your team’s phones.

Keep sessions to 30 to 45 minutes. Pick one specific task each time. Make attendance welcoming rather than mandatory. Keep a shared notes document so discoveries accumulate.

Buying before you have mapped

The temptation, often pushed by enthusiastic sales calls and convincing demos, is to buy an AI tool before you have clearly defined the problem you are trying to solve. Without a clear workflow map and a specific wish list, you will either buy the wrong tool, pay for features you never use, or run into adoption problems you were not prepared for.

Complete your whiteboard session and your existing-software review before any vendor conversation. If a salesperson wants to show you their product before you are ready, “we are still in our discovery phase” is a perfectly reasonable response.

Ignoring the data governance question

Using a consumer AI tool for business purposes, particularly where client or sensitive data is involved, without knowing where that data goes and how it is used is a serious risk. It is not just a legal risk. It is a reputational one. The instinct to pause, check, and verify, as my legal sector contact did when Otter appeared unexpectedly in her Teams meeting, is exactly right. Never assume that a free or low-cost AI tool has the same data standards as your enterprise software.

Before using any AI tool with client or confidential data, check: where is it held? Is it within UK or EU borders? Does the platform use it to train its model? Does it persist after the session? If you cannot answer those questions, do not proceed.

Underestimating the human cost of change

The biggest risk in any AI adoption programme is not a technical failure. It is human disengagement. If your team does not understand why AI is being introduced, does not feel involved in the process, and cannot see a clear answer to “what does this mean for my job?”, adoption will stall. Burnout and quiet resentment follow when change is announced from above without genuine communication, support, or a people-first approach.

Never announce AI adoption as a done deal. Involve your team from the mapping stage. Communicate openly. Answer the job security question directly and honestly. Pace the rollout in manageable steps, not a big bang.

These are a starting point for UK and US business owners who want to go deeper on AI readiness, digital skills, and responsible adoption.

1. Microsoft Copilot for Microsoft 365, official documentation. A clear overview of what Copilot can do within your existing Microsoft 365 environment. Start here before buying anything else. microsoft.com/copilot
2. Google NotebookLM. A free AI research and note-taking tool useful for processing audio files and generating summaries in a contained environment. google.com
3. NCSC (National Cyber Security Centre), AI guidance for organisations. The UK government’s official guidance on using AI securely, including data governance and risk management. gov.uk
4. MIT Sloan Management Review, AI and the future of work. Research-backed articles on the human side of AI adoption in organisations. mit.edu
5. Human-First Responsible AI Pledge, a values-grounded framework for leaders who want adopt AI in a way that is ethical, human-centred, and commercially viable humanfirstresponsibleaipledge.org

I have been helping businesses find their way in the digital world since the mid-1990s. I have seen the commercial internet arrive, social media take hold, the mobile shift, and now the thing that changes everything more than any of those: AI becoming genuinely accessible to anyone with a laptop and a browser.

Each of these waves brought the same mixture of real opportunity and manufactured urgency. Each time, the businesses that fared best were not the ones that moved fastest. They were the ones that moved with the most thought.

The legal professional I spoke with was very clear about where she stood. Her first responsibility was to her clients, her team, and the standards of her profession. She wasn’t prepared to set those aside because AI is the latest trend, it was a thoughtful assessment of what mattered most, and I respected that.

“This is the biggest change project that small business owners will have to work through since computers first arrived in the office. It deserves that level of respect, and that level of care.”

My role as a content marketing skills and AI strategy consultant is to help businesses find their own voice in the digital world. AI does not change that. If anything it sharpens it. As AI makes it cheaper and faster to produce content, the thing that will matter most is authenticity, expertise, and human judgement. The businesses that do well will be the ones that use AI in service of those things, not as a replacement for them.

So: allons-y. Let us go. But let us go together, carefully, with our eyes open and our teams beside us.