How Do I Know If My AI Consultant Is Actually Competent?

1. A competent consultant asks better questions than an average salesperson

The first signal of competence is not the answer they give. It is the questions they ask. A serious AI consultant should want to understand your workflows, systems, constraints, data quality, team capacity, risk profile, and how success will be measured. If the conversation jumps straight to model names, automation magic, or dramatic ROI claims, that is a warning sign.

Good consultants do not just ask what you want to build. They ask what goes wrong today, where the bottlenecks are, who owns the process, how outputs are checked, and what would make the project commercially worthwhile. That is the language of delivery, not theatre.

If someone cannot turn your vague idea into a sharper business problem, they are unlikely to solve it well.

2. They can show evidence, not just enthusiasm

Case studies matter, but quality matters more than quantity. Ask what was built, how success was measured, what changed in the client workflow, and what the limits were. A competent consultant should be able to explain outcomes in plain English, not hide behind confidentiality whenever specifics get uncomfortable.

You should also ask about failure. What type of AI projects have not worked well for them? Where have they had to scale back scope or recommend a simpler solution? Honest practitioners have stories like that because real delivery includes constraints, edge cases, and trade-offs.

In due diligence language, you are looking for proof that they understand dependencies, integration blockers, security implications, and operational reality. If every past project sounds frictionless, you are probably hearing a sanitised sales narrative.

3. They speak clearly about data, governance, and security

Any consultant working with business AI should be comfortable discussing data handling, access controls, supplier risk, and review processes. They do not need to be your lawyer, but they do need to understand how AI affects confidentiality, governance, and compliance.

In the UK, that means they should be able to speak sensibly about UK GDPR exposure, supplier retention questions, logging, permission boundaries, and where human review is still required. If they treat those topics as paperwork for later, that is a serious weakness.

Competence here is often visible in the boring details. Do they talk about rollout guardrails, access policies, prompt handling standards, documentation, and support after launch? Or do they act as though the important part is merely connecting a model and shipping a demo?

If data risk matters in your organisation, you should also read our article on AI security and privacy risks.

4. They are willing to recommend a smaller project or no project at all

This is one of the strongest trust signals. Competent consultants do not need every conversation to become a large engagement. Sometimes the right answer is a workshop, an audit, a low-risk pilot, or even a recommendation to improve the process before adding AI.

If someone always steers you towards the biggest package, the most complex architecture, or a broad transformation programme before they have really understood the basics, be careful. Over-scoping is often a sign that they sell confidence better than judgement.

By contrast, an experienced consultant knows that saying no can build more trust than forcing a sale. Marcus Sheridan makes this point well in TAYA terms: the hard truths are often what win business later.

5. Use a simple scorecard before you sign anything

If you want a practical filter, score each consultant out of five on these areas: clarity of problem understanding, evidence of delivery, honesty about limitations, data and governance maturity, and fit for your business size. That simple framework is often more useful than a slick deck.

You should also ask who will actually do the work. In many agencies the senior person sells and a junior person delivers. That is not automatically bad, but it should be transparent. Ask about handover, documentation, ownership of assets, and what support looks like after launch.

If a consultant scores poorly on transparency, delivery evidence, or governance maturity, walk away. AI projects are hard enough without paying someone to learn on your time.

Before signing, it is also worth reading the red flags to look for in an AI agency contract.

Is This Right For You?

This article is right for you if you are speaking to AI consultancies, freelancers, or agencies and want a better filter than confidence and buzzwords. It is especially useful if this is your first AI project and you need help separating genuine delivery capability from polished salesmanship.

It is less useful if you are already deep into procurement with a formal technical due diligence team. In that case, you may need a detailed vendor scorecard rather than a plain-English buyer guide.

Frequently Asked Questions

What is the biggest red flag in an AI consultant?

Overconfidence without specifics. If they cannot explain the workflow, risks, and delivery steps in plain English, they may not have real implementation depth.

Should an AI consultant show examples of past work?

Yes. They should be able to describe real use cases, outcomes, and lessons learned even when client details are confidential.

Do certifications prove AI competence?

Not on their own. They can help, but practical delivery experience, judgement, and transparent communication matter far more.

Is it a bad sign if a consultant says AI is not right for every problem?

No. That is usually a good sign. Mature consultants know when process fixes, standard automation, or no change at all are better options.