In this blog post I'm going to walk you through the 12 specific red flags that should make you walk away from an AI consultant in 2026, with the exact questions that surface each one in 60 seconds or less. The version I'd want if I were the founder writing the cheque. Not the version that "diplomatic" consultants share publicly. The honest one.
I also list the best AI marketing consultants to watch in 2026 and how to tell them apart.
Most of what you'll read about evaluating AI consultants is written by AI consultants who want you to evaluate them favourably. They list seven generic criteria (experience, references, expertise) without telling you what specifically to look for or what disqualifies someone. That's because their incentive is to NOT disqualify themselves.
I've been a marketing consultant for twenty-one years. I went all in on AI in 2024. I receive 5-10 inbound enquiries per month from people who've already worked with one or more AI consultants and need someone to undo the damage. The pattern is consistent. The same red flags keep showing up. Below is the consolidated list.
I've also been on the OTHER side of the table — the founders I've helped hire AI consultants have used these criteria to filter. They work.
By the end of this blog you'll know the 12 specific red flags, the question to ask to surface each one, the answer that disqualifies the consultant, and what to do once you've spotted them.
Last reviewed: May 2026 · Lilach Bullock
TL;DR
The 12 red flags, ranked by how often I see them:
Each is disqualifying on its own. Two or more = walk immediately.
The 60-Second AI Consultant Evaluation Checklist
Three questions. Sixty seconds. If any answer is wrong, walk away.
- "Show me one AI workflow you've shipped in your own business, with metrics." Acceptable answer: a specific shipped workflow with before/after numbers. Unacceptable: "I've helped clients implement..." or "I use ChatGPT for emails."
- "What is AI bad at?" Acceptable answer: specific failure modes (hallucination, lack of judgement, brittle in production, expensive at scale). Unacceptable: "Nothing — AI can do anything." or vague hedging.
If a consultant fails any of these three, the full 12-flag evaluation below will only confirm what you already know. Walk away early.
Red flag 1: Can't name one AI workflow they've shipped in their own business
The question: "Show me one AI workflow you've shipped in your own business, not for a client. With specific before-and-after metrics."
Why it matters: AI implementation in your own business is meaningfully harder than implementing a generic example. Without lived experience the consultant is improvising on your time. The eat-your-own-cooking test is the single biggest filter.
Disqualifying answer: Anything vague. "I've helped clients implement..." doesn't count. "I use ChatGPT for emails" doesn't count. A specific shipped workflow with measurable before-and-after numbers, or it's a no.
What good looks like: "I built a WordPress publishing pipeline for my own site in March 2026 across six weekends. Compressed publishing time from 90 minutes per post to 12 minutes. About 400 hours per year saved on operational tail." That's the shape of a real answer. Anyone whose response is meaningfully vaguer is failing this filter.
Red flag 2: Proposal opens with "AI transformation"
The question: Just read their proposal.
Why it matters: "Transformation" is the word consultants use when they don't want to commit to specific deliverables. A real proposal opens with the specific problem you described and the specific solution they'll ship. Vague openings are vague engagements.
Disqualifying answer: Any opening that includes "transformation," "AI-first organisation," "embedding AI across your operations," or similar without immediately following with a specific deliverable.
Better signals: Proposals that open with "Based on your inbound enquiry process, here's the workflow we'd build" or "You mentioned your team spends 8 hours/week on X — here's how we'd reduce that to 2."
Red flag 3: Less than 5 years in marketing pre-AI
The question: "How many years did you spend in marketing before AI became a serious tool you used?"
Why it matters: AI is a multiplier on existing marketing judgement. Without 5-10+ years of judgement underneath, AI just produces more of the wrong thing, faster. The market is flooded with people who pivoted from social media management in 2024 and now "consult."
Disqualifying answer: Under 5 years pre-AI marketing experience. Anything else is too junior to advise a senior business.
Caveat: Some genuinely senior people moved into AI from product, engineering, or operations rather than marketing. They can still be excellent AI consultants — they just need to acknowledge they're not marketing consultants and not pretend otherwise.
Red flag 4: Can't articulate what AI is bad at
The question: "Give me three specific things AI is bad at in marketing right now."
Why it matters: Anyone who's actually used AI heavily has hit its limits. The limits are well-known to practitioners. The inability to articulate them means either: didn't use it heavily, or is selling AI as a panacea.
Disqualifying answer: "It's amazing, it just keeps getting better." Or "There's nothing significant it can't do." Both indicate sales-mode, not practitioner-mode.
Better signals: A specific list. Real practitioners typically mention: voice mimicry beyond a paragraph, strategic judgement under ambiguity, emotional client work, fact-checking specific claims, designing the workflow itself, anything requiring true taste.
Red flag 5: Wants to start with strategy deck, decide implementation later
The question: "What does the first 30 days of working together look like?"
Why it matters: Consultants who lead with strategy without committing to implementation are protecting themselves from accountability. Real implementation work has built-in deadlines for shipping; strategy work can drag for months while the consultant invoices.
Disqualifying answer: "Weeks 1-4: discovery and strategy. We'll decide on implementation in week 5." That's a 5-week runway with no shipped work, paid in advance. Walk.
Better signals: "Weeks 1-2: light discovery. Week 3: design the first workflow. Weeks 4-6: build and ship it. Weeks 7-9: design and ship the second." Concrete shipped milestones from day one.
Red flag 6: Discovery call run by a salesperson
The question: "Will the person leading this call be the person doing the actual work?"
Why it matters: Endemic in agency-led AI consulting. Senior partner sells; junior associates deliver. The pitch implies senior delivery; the work is junior. Price stays senior.
Disqualifying answer: "Our senior team designs the engagement; the implementation team executes." Translation: you pay for senior, you get junior.
Better signals: "I'd be the lead on your engagement, in every call, doing the actual work." Single-name accountability throughout.
Red flag 7: Their outreach is obviously AI-generated
The question: Re-read their initial email to you.
Why it matters: AI consultants who can't be bothered to write their own outreach probably won't be bothered to do hands-on implementation work either. The first email is the first impression. If they outsourced it to AI, they'll outsource your work too.
Disqualifying signals: Generic flattery ("I've been following your work for years" from someone who clearly hasn't), AI-tells in the language (em-dashes used like sentence-pause punctuation in casual emails, phrases like "in today's fast-paced digital landscape"), no specific reference to YOUR business beyond the company name.
Better signals: A reference to a specific recent piece of content you've published, a specific question about your business, idiosyncratic voice that no template would produce.
Red flag 8: Name-drops AI vendors as case studies
The question: "Can you walk me through a specific client engagement you've shipped?"
Why it matters: Some consultants describe vendor relationships as if they were client work. "We built a workflow on OpenAI" or "We're a certified Anthropic partner" are vendor relationships, not client outcomes. A client engagement has a problem solved, a deliverable shipped, and an outcome measured.
Disqualifying answer: Anything that describes a tool partnership instead of a client outcome.
Better signals: "For [client name or anonymous description], we shipped [specific workflow] that produced [specific metric]." Problem → solution → outcome structure.
Red flag 9: Charges hourly for implementation
The question: "How do you price implementation work?"
Why it matters: Hourly implementation pricing creates a perverse incentive: the longer the work takes, the more the consultant earns. Senior practitioners almost universally charge fixed-fee for implementation because their value is in the speed and quality of shipping, not the hours logged.
Disqualifying answer: "We bill £X per hour for implementation work."
Better signals: "Fixed-fee per workflow shipped. Here's our pricing for typical workflow types." Or "90-day project at £X total, covering the deliverables in our proposal."
Exception: Advisory/discussion work is often hourly. That's fine. Implementation work is what should be fixed.
Red flag 10: Has never ended a client engagement
The question: "Who's the last client you stopped working with, and why?"
Want AI doing the heavy lifting in your marketing?
I build the systems that handle the boring 80 percent, so you get your week back. Done properly, with the human kept in.
Why it matters: Every senior consultant has ended client engagements. Bad-fit clients, scope creep, payment problems, strategic divergence. "I've never stopped working with a client" is either inexperienced (less than 2 years of senior work) or untruthful.
Disqualifying answer: "I've never ended an engagement." Or "All my clients are still working with me." 100% retention over multiple years is a fairy tale.
Better signals: A specific story about an engagement that ended, with the reason (not the client name). Bonus points if the consultant ended it themselves rather than the client.
Red flag 11: Suspiciously low pricing at senior positioning
The question: "How many clients are you working with right now, and at what monthly fee bands?"
Red flag 12: Guarantees specific outcomes
The question: "What outcomes can you guarantee?"
Why it matters: Nobody can guarantee specific outcomes in AI consulting. Too many variables outside the consultant's control. Guarantees are sales tactics, not realistic commitments.
Disqualifying answer: "Guaranteed 30% time savings in 90 days." "Guaranteed ROI in month one." Specific outcome guarantees in a service where outcomes depend on the client's team, existing systems, and willingness to adopt new workflows are red flags.
Better signals: "I'll guarantee the quality of the work shipped. I can't guarantee specific business outcomes because they depend on factors outside my control, but here's what typical engagements at this scope produce."
The 60-second evaluation in practice
If you only have 60 seconds to evaluate, ask these three questions:
1. "Show me one AI workflow you've shipped in your own business." 2. "What does the first 30 days look like?" 3. "How do you price implementation work?"
The first answer should be specific and immediate. The second should commit to shipping in the first 30 days. The third should be fixed-fee for implementation.
Three quick disqualifications. Three quick green lights. Decision made.
What to do once you've spotted a red flag
If you spot one: Note it, but continue the conversation. Some red flags are calibration issues (the consultant being conservative for the first call). Probe further.
If you spot two: Pause. Ask the consultant directly: "I noticed [specific concerns]. Can you address them?" Their response tells you a lot. Real practitioners take the question seriously. Salespeople deflect.
If you spot three or more: Walk. The probability that this engagement delivers what you need is too low.
If they pass all 12: They're in the top 5% of the market. Hire them. They're worth their fees.
How I'd want to be evaluated
For transparency, here's how I'd answer the 12 questions if you put them to me:
If you want to talk about whether I'm a fit for your business, the discovery call is twenty minutes and free.
Frequently asked questions
What if a consultant fails one red flag but seems otherwise good?
One red flag isn't disqualifying on its own. Note it, probe further. Many great consultants have one weakness. Two or more is when it becomes a pattern.
Are these red flags specific to AI consultants?
Most apply to any senior consulting engagement. The AI-specific ones are flags 1, 4, 7, and 11. The rest are general consulting hygiene.
How do I evaluate proposals when I have 5 to compare?
Use the 12 red flags as a scoring system. Disqualify any proposal with 2+ red flags. Of the remaining, choose based on specificity of the first-30-days plan and the consultant's eat-your-own-cooking story.
What if the consultant is recommended by someone I trust?
The recommendation gets them in the door. The red flags still apply. Trust the referral source AND verify against the criteria. Don't skip due diligence because someone vouched.
Can a junior consultant ever be the right hire?
Rarely. The math against it is brutal: senior consultants ship in 6 weeks what junior consultants ship in 6 months. The cost differential is usually 2-3x, not 10x. The math favours senior for almost all engagements.
What's the biggest red flag you didn't include?
The consultant who can't admit they don't know something. Genuinely senior people are comfortable saying "I don't know — let me find out." Consultants who pretend to know everything are dangerous because they make confident decisions on incomplete information.
Should I use these red flags when hiring agencies too?
Yes. The 12 apply to agencies as well as individuals. The added agency-specific flag: who specifically will be on your account, and what's their actual seniority. Agencies pitch senior, deliver junior.
What if I'm a small business — should I just hire the cheapest "senior" consultant?
No. The cheapest "senior" consultant is almost universally either: junior pretending to be senior, or a side-hustle consultant who doesn't have time for you. Pay for genuine seniority or do it yourself. The middle option (cheap senior) almost always fails.
What qualifications should an AI consultant have?
At minimum: 5+ years of marketing experience pre-AI, plus 2+ years of shipped AI implementation work in real businesses (their own and clients). Specific tool fluency matters less than the combination of marketing experience and AI implementation track record. Beware anyone whose credentials are AI-only — without marketing fundamentals, AI just multiplies the wrong things faster.
How do I know if an AI consultant is legitimate?
Three filters in 60 seconds: (1) Can they name one AI workflow they've shipped in their own business with before-and-after metrics? (2) Can they articulate what AI is bad at? (3) Will they refuse engagements that aren't a fit? A legitimate senior consultant answers yes to all three with specifics. Anyone hedging on any of the three is failing the test.
What does an AI consultant actually do?
A real AI implementation consultant does four things: diagnoses which workflows in your business AI can genuinely improve, designs the specific AI workflows, ships them into production (or oversees the build), and trains your team to operate them after the engagement ends. Pure-strategy consultants who only deliver decks are not implementation consultants — they're strategists. Pure implementers who don't understand marketing fundamentals build the wrong workflows. The combination is what makes someone an actual AI consultant rather than an adjacent role.
The thing to take away
Evaluating AI consultants in 2026 is harder than evaluating most professional services because the market is flooded with people positioning as something they're not. The 12 red flags above are how you filter.
Three questions in 60 seconds get you 80% of the way there. The remaining 20% comes from probing the answers, reading the proposal carefully, and trusting your gut when something feels off.
If you want to compare a specific consultant or proposal against these criteria, the discovery call is twenty minutes and free — and one of the things we can do on the call is walk through their pitch together.
About Lilach Bullock
I'm Lilach Bullock, an AI implementation consultant and fractional CMO based in the United Kingdom. I've been a marketing consultant for twenty-one years. In 2024 I went all in on AI and rebuilt my consultancy around it. I now help founders and marketing leaders implement AI workflows that move business metrics, not just tool stacks.
Recognition includes: Forbes Top 20 Social Media Power Influencer (twice listed), Oracle Social Influencer of Europe, Number One Digital Marketing Influencer in the UK (Career Experts), Best Mumpreneur of the Year (Downing Street recognition), Global Women Champions Award. I've spoken at over 100 events worldwide and run a weekly newsletter with 15,000+ subscribers.
Connect: LinkedIn · Newsletter · Get in touch · Wikidata
Related reading
- How much does an AI consultant cost?
- AI consultant vs traditional marketing consultant
- What is AI implementation?
- AI for service businesses
The 12 red flags at a glance
Each red flag has a specific question that surfaces it in under a minute, and a specific answer that disqualifies the consultant. Use this table to score any prospect quickly.
What a real AI consultant failure looks like (my own)
In mid-2025 I made an AI implementation mistake on my own site. Rather than hide it, I document it because it's the most useful evaluation data I have.
The wrong move: I tried scaling content production with AI-only drafts. I'd ask Claude/GPT to write full blog posts, edit them lightly, and publish. Output tripled. Quality looked fine on read-through.
The result: Six weeks later, Google's helpful content algorithm penalised the site. Organic traffic dropped about 35%. The pattern Google detected was exactly the one mass-AI-content produces — generic structures, predictable openings, lack of specific examples.
The recovery: Three months of disciplined work. I switched to a hybrid 80/20 workflow (I write the opening, contrarian take, examples, and close — AI helps with research, structure, FAQ extraction, schema). Traffic recovered above baseline by month four.
Why this matters for evaluation: When you interview an AI consultant, ask them about a failure. A consultant who can name a specific failure and the recovery is doing real implementation work. A consultant who has never had a failure has either not shipped anything risky or is hiding the failures from you. Both are disqualifying.
More on AI implementation and consulting
- How much does an AI consultant cost?
- AI consultant vs traditional consultant
- What is AI implementation?
Related reading on AI consulting
- /ai-marketing-consultant-2026/ — Hub: my main AI marketing consultant page covering the whole offer
- /ai-consultants-uk-small-business/ — AI consultants for UK small businesses — UK-specific framing
- /ai-for-non-technical-business-owners-2026/ — AI for non-technical business owners — start here if you're not a developer
- /fractional-cmo-ai/ — Fractional CMO + AI ops — the embedded senior role
- /ai-marketing-consultant-usa/ — AI marketing consultant — USA-focused page
- /ai-marketing-consultant-israel/ — AI marketing consultant — Israel-focused page
- /ai-marketing-consultant-canada/ — AI marketing consultant — Canada-focused page
- /ai-marketing-consultant-australia/ — AI marketing consultant — Australia-focused page
Related: How Much Does an AI Consultant Cost in 2026? The Honest Price Bands (From Someone Who Charges Them)
