For many Gloucestershire businesses, the hardest part of using AI is not the technology. It is choosing where to begin.
There are plenty of impressive demos online. AI can write, summarise, analyse, search, plan and automate. But if you run a professional practice, trade business, agency, consultancy, clinic, manufacturer, shop or local service business, the real question is simpler:
Where could AI save time or improve quality without creating unnecessary risk?
The easiest answer is usually this: start with one repeatable process.
Not your whole business. Not a grand “AI transformation”. Not a system that tries to replace three people by next month. Just one process that happens often, takes time, and already has a fairly predictable shape.
For many small and medium-sized businesses, the best place to start is enquiries.
Why enquiries are a sensible first AI project
New enquiries are valuable, but they are also easy to mishandle.
They arrive through website forms, emails, phone notes, social messages, referrals and sometimes half-complete conversations. Some are urgent. Some are not a fit. Some need a quick answer. Some need a quote. Some need a proper discovery call. Some simply need to be logged so they do not disappear into somebody’s inbox.
Most businesses already have a rough process for this. It might not be written down, but people know what tends to happen:
- Read the enquiry
- Work out what the person is asking for
- Decide whether it is a good fit
- Ask for missing information
- Draft a reply
- Add it to a CRM or spreadsheet
- Set a reminder to follow up
That makes enquiries a good starting point for AI. The work involves reading, sorting, summarising and drafting, which are all things AI can be useful for. It also has a clear commercial link: better handling of enquiries can mean faster responses, fewer missed opportunities and a more consistent first impression.
Just as importantly, it does not have to be fully automated on day one.
What an AI-assisted enquiry process could look like
A sensible first version might be very simple.
When a new enquiry arrives, AI could help by producing a short internal summary:
- Who the person or business is
- What they appear to need
- Any deadlines or urgency
- What information is missing
- Suggested next action
- A draft reply for a human to review
For example, a Cheltenham accountancy firm might receive an enquiry from a growing local business asking about payroll, bookkeeping and Making Tax Digital. Instead of somebody reading the email from scratch, deciding what matters, and typing the first response manually, AI could prepare a structured note:
Potential new client. Limited company based in Gloucestershire. Needs bookkeeping and payroll support from next quarter. Mentions current accountant is slow to respond. Missing: number of employees, turnover band, software currently used. Suggested next step: offer a 20-minute call and ask three qualifying questions.
It could then draft a polite reply in the firm’s tone, ready for a member of staff to check and send.
That is not glamorous. It will not make headlines. But it is useful.
And useful is where most businesses should start.
Keep the human in charge
The aim is not to let AI decide which customers matter or send unchecked messages on behalf of your business.
A better first step is human-in-the-loop. AI prepares the summary and draft. A person reviews it, edits it if needed, and sends it. Over time, you learn where the AI is reliable, where it needs clearer instructions, and where automation should stop.
This matters because enquiries often contain sensitive or commercially important information. A professional services firm, for example, may receive details about finances, staff issues, contracts or personal circumstances. A local manufacturer might receive drawings, supplier details or pricing information. A healthcare or wellbeing business may receive personal information.
So the question is not just “can AI do this?”
The better question is:
Can we design a process where AI helps, while the business keeps control of data, tone, decisions and risk?
That is where implementation matters.
What to decide before you build anything
Before connecting AI to enquiries, a business should answer a few practical questions.
First, what types of enquiries do you receive? A general “contact us” message is different from a quote request, a support issue or a sales lead.
Second, what should AI be allowed to do? Summarising and drafting are lower-risk than sending replies automatically or updating live customer records.
Third, what information should never be sent to an external AI tool? This is especially important for regulated, professional or data-sensitive businesses.
Fourth, what does a good output look like? If you want a summary, define the fields. If you want a draft reply, define the tone. If you want qualification questions, decide what counts as useful.
Fifth, how will you measure whether it is working? You do not need a complicated dashboard. You could start with simple measures such as response time, number of enquiries followed up, staff time saved, or the percentage of AI drafts that are usable after review.
A lesson from practical AI experiments
In my own AI agent experiments, including the Cheltenham Times local news site project, the useful lessons did not come from asking AI to do impressive party tricks. They came from giving AI a real job, setting boundaries, and watching where it helped and where it needed steering.
The same pattern showed up in my AI-first CRM experiment: the interesting opportunity was not “AI everywhere”. It was using AI to help with real work such as email triage, task handling and deciding what needs attention.
That is the mindset I would recommend for local businesses. Do not start by asking how to use every AI tool available. Start by choosing one business process and making it slightly faster, clearer or more consistent.
Enquiries are often a good candidate because the work is frequent, valuable and easy to review. If AI gets the summary wrong, a human can spot it. If the draft reply is not quite right, it can be edited. If the process works, you can improve it gradually.
Start small, then build confidence
The first AI project in a business should prove something.
It should prove that AI can help with a real workflow. It should prove that staff can use it without feeling threatened or confused. It should prove that the business can manage privacy, cost and oversight sensibly. And it should prove that the output is good enough to be worth continuing.
That is why one enquiry workflow is a better starting point than a vague ambition to “use AI more”.
You can map it, test it, measure it and improve it. You can start manually with prompts and templates, then move towards a more joined-up workflow if it earns its place.
AI does not have to be a big leap. For many Gloucestershire SMEs, the best first step is simply this:
Pick one process. Make it clearer. Let AI assist. Keep a human in control. Measure whether it helped.
If you would like to explore where AI could genuinely help in your business, I can help you identify one practical starting point, design a sensible first workflow and decide what should — and should not — be automated.
Book a short AI consultancy call, or get in touch and ask: “What could AI do in our business without creating unnecessary risk?”
