Is your SME really ready for AI? Five signals to check

Artificial intelligence is everywhere right now.

In one place it is used to write emails, in another to create advertising, prepare quotes, answer customer enquiries, check documents or analyse reports.

So a lot of companies feel the pressure: competitors are doing something with AI, staff are using ChatGPT, everyone is talking about automation, agents and productivity. It looks as though you need to start as quickly as possible too.

But this is where one mistake tends to happen.

A company starts from the tool, when it should start from its own operations.

Simply buying an AI tool doesn’t mean a company has begun applying AI effectively. Someone may write an email or draft a quote faster, but if everything else in the company stays just as scattered, the business may feel no real benefit at all.

For example, if customer enquiries arrive by email, Facebook, Messenger, phone and directly to the manager, AI won’t help simply because a chatbot now exists.

First you have to answer some simple questions:

• Where should each enquiry end up?

• Who is responsible for it?

• What information is needed to prepare a quote?

• Who makes the decision?

• How do we measure whether we answered the customer faster and better?

Only then can AI become a real part of the process rather than another tool everyone uses differently.

Company maturity isn’t about age or turnover

When we talk about company maturity, we don’t mean how many years the company has existed, how many people it employs or what turnover it generates.

Maturity shows how well a company can operate in an orderly, stable way even when the owner isn’t involved in every decision.

In a less mature company, a great deal rests on particular individuals.

One person knows where to find the information. Another knows how to put a quote together properly. The manager remembers what was agreed with the client. A new hire needs a long explanation of how everything works, because the process isn’t written down anywhere.

In a more mature company the situation is different.

It is clearer how the work runs, who is responsible for what, where information is kept and on what basis decisions are made. Such a company can grow faster, depends less on individuals and finds automation easier to adopt.

AI is no miracle here. It simply helps to do faster the work a company already knows, at least partly, how to do in an orderly way.

1. You clearly understand where time disappears in the company

The first signal that a company is ready to start thinking about applying AI is that it understands where its biggest problems are.

Not “we want to use AI”. Specifically: “we want to prepare quotes faster”, “we want to reduce manual order entry” or “we want customers to wait less for an answer”.

• commercial quotes take too long to prepare;

• account managers spend too much time answering similar enquiries;

• staff move data by hand from one system to another;

• customer information is scattered across several places;

• reports are prepared manually;

• a lot of time goes on hunting for documents, prices or previous agreements.

If a company can name one or two processes that slow the work down most, it can already start looking for a solution.

In a services business, for instance, it may turn out that preparing a quote takes a long time not because the person is slow. The problem may be that they have to gather information from earlier emails, price lists, Excel files and colleagues.

AI can help in a place like that. But only once we know where the information should come from and what the final result should look like.

2. The process doesn’t live only in people’s heads

If you asked three employees how the customer enquiry process works, would you get three identical answers?

If not, the process probably isn’t clear enough yet.

In many small and medium companies the work runs like this: whoever sees the enquiry first answers it. Whoever has more experience knows how to do it better. When a problem comes up, everyone asks the manager.

In that situation AI can help individuals, but it is very hard to embed properly into the whole process.

First you need agreement on the basics:

• what the stages of the process are;

• what information is essential at each stage;

• who is responsible for the decision;

• where the result gets recorded;

• when human approval is required.

You don’t need elaborate procedures or dozens of documents.

Sometimes one clear description of the process is enough: from receiving the enquiry through to the final quote or fulfilled order.

3. The information you need isn’t scattered across the whole company

AI can quickly find, summarise, compare or prepare information. But it can’t work properly if the company itself doesn’t know where that information is.

If customer history sits in one person’s inbox, the price list in another Excel file, contracts on a shared drive, and the most important agreements in Messenger conversations, the first step isn’t an AI solution.

The first step is sorting out the information.

That doesn’t mean immediately buying an expensive ERP or CRM. Sometimes it is enough to start with one shared place where enquiries, customers, quotes or documents are recorded.

What matters is that the information is:

• findable;

• current;

• understandable;

• used the same way by everyone.

Only then is it worth thinking about an AI assistant that helps prepare quotes, answers questions from company documents or classifies enquiries automatically.

4. There is someone who can make decisions

AI implementations often fail for reasons that have nothing to do with technology.

They fail because nobody owns them.

One person tries ChatGPT. Another builds an automation. A third buys a different tool. Everyone is trying something, but nobody is accountable for the overall result.

So a company needs someone who can say:

• which process we tackle first;

• what we are aiming for;

• who will be involved;

• what data we are allowed to use;

• how we will verify the result;

• whether the solution is worth extending.

It doesn’t have to be an IT manager or a developer.

Often the right person is the process owner: the sales manager, the project manager, the production manager, the office manager or the owner themselves.

What matters most is that they understand the real work and can make decisions.

5. The company is ready to change the way it works, not just the tool

This is the most important point.

AI almost always changes not just the tool but the workflow itself.

Where someone used to prepare a quote from scratch, after AI they may first receive a draft, then check it, add to it and approve it.

Where customer enquiries used to be sorted by hand, after automation the system can group them by topic, priority or customer, and the person only reviews the harder cases.

Where a report used to take several hours, AI may assemble a first version in minutes — but a person still has to check the numbers and make the decision.

So before implementing anything, it is important to agree what the new process will look like.

Who does what? Where does a person check the result? What happens if the AI returns something wrong? How will we know whether the solution genuinely saved time?

If a company is ready to answer those questions, it is already much closer to real benefit from AI.

What if you aren’t ready yet?

That isn’t bad news.

Plenty of companies aren’t ready to implement complex AI solutions straight away. And right now they don’t need to be.

Often the biggest benefit comes not from an expensive solution but from simple things:

• sorting out the path a customer enquiry takes;

• standardising how quotes are prepared;

• centralising documents;

• writing down the most important process;

• understanding where staff lose the most time;

• picking one process for a first pilot.

A small company usually doesn’t need to start with a complex AI project.

It needs to start with one process that currently involves a lot of manual work, disorder or repeated questions.

Where to start in practice

Don’t start with the question:

“Which AI tool should we buy?”

Start with the question:

“Which process is losing us the most time, money or customers today?”

Then analyse that process.

Look at where information gets stuck, where people repeat the same thing by hand, where the manager becomes the bottleneck, and where errors most often appear.

Only then can you decide whether what is needed there is artificial intelligence, simple automation, a new system, clearer procedure — or all of those together.

AI can help a company grow faster, serve customers better and reduce manual work.

But it works best when the company understands its own processes first.

The first step isn’t an AI tool. The first step is clarity about how your company works today.

If your team is just getting started, a good first step is AI training for companies: shared fundamentals for everyone plus hands-on workshops by function.