Most Irish SMEs don't need an AI consultant yet. Here's how to tell when they do — and what to look for when they get there.
Most Irish SMEs don't need an AI consultant right now. They need to use the tools they already have more effectively, map their workflows, and identify one or two high-value problems to address. That work can often be done internally, with discipline and a clear process.
But there are signals that indicate an SME has moved beyond what internal effort can resolve — and at that point, external advisory makes economic sense. Knowing the difference saves money and time.
The signals that indicate readiness
The clearest signal is that tool adoption has happened but results haven't. The business has been using AI tools for six months or more, some staff are using them regularly, but nobody can quantify the value and the initial enthusiasm is fading. This is the point where a diagnostic — conducted by someone without a stake in the tools already purchased — tends to be most valuable.
A second signal is that the AI opportunity is large enough to justify getting it right. A business turning over €5 million with a clear workflow problem that's costing €300,000 per year in inefficiency has a compelling case for external advisory. The cost of the engagement is a fraction of the recoverable value. A business where the best available opportunity is worth €20,000 probably doesn't.
The third signal is that internal capacity genuinely isn't there. If the person most qualified to lead AI implementation is already at 100% capacity running the business, the implementation won't happen without external resource. The choice is between hiring a consultant to drive it or accepting that it won't get done.
A fourth, less obvious signal: the business is about to make a significant AI investment. If you're considering a multi-year software contract, a custom development project, or any AI spend above €30,000, the cost of a vendor-neutral diagnostic before you commit is almost always justified. The risk of getting the decision wrong is too high to rely solely on the vendor's assessment.
What to look for in an AI consultant
Vendor neutrality is the most important criterion. An AI consultant who derives income from recommending specific tools has a structural conflict of interest. The recommendation will reflect the consultant's vendor relationships, not your business needs. Ask directly: do you receive fees, commissions, or referral income from any AI vendors? The answer should be no.
Diagnostic-first methodology. A good AI consultant should want to understand your business before recommending anything. If the opening conversation is primarily about what AI can do rather than what your specific business needs, that's a poor sign. The diagnostic should precede the prescription.
Relevant sector experience helps but isn't essential. A consultant who has worked with businesses at a similar scale, in your sector, or with analogous workflow problems will move faster and spot patterns more readily. But a rigorous generalist is more valuable than a sector specialist who skips the diagnostic.
Look for an engagement that produces a specific output: a prioritised recommendation with a clear ROI case, actionable without requiring the consultant to remain involved indefinitely. Consultants who are hard to disengage from are not structuring engagements in your interest.
What to avoid
Be wary of consultants who lead with their technology stack. If the first conversation is about a proprietary platform or a "solution" the consultant has already built, you are being sold to, not advised.
Be wary of engagements that start large. A credible AI advisory engagement for an Irish SME typically begins with a diagnostic — a time-bounded piece of work that produces a clear picture and a prioritised recommendation. If the opening proposal is a six-month retainer or a €100,000 implementation project, the consultant has not earned the right to ask for that yet.
And be wary of urgency as a sales tool. The AI landscape is changing fast, but that doesn't mean every decision needs to be made this quarter. A consultant who creates pressure around speed of adoption is working with the vendor's interests in mind, not yours.
What the engagement should produce
An AI advisory engagement for an SME should produce three things: a clear picture of where value is being lost in the current operation; a prioritised set of AI (and non-AI) recommendations with ROI estimates; and a practical implementation plan that the business can execute, with or without the consultant's continued involvement.
If the output is a strategy document with no implementation path, you've paid for a report. If it's an implementation without a strategy, you've paid for tool deployment. Both fall short. The combination of clear diagnosis, prioritised recommendation, and executable plan is what justifies the cost.