AI Consulting: What to Expect and How to Choose

What AI Consulting Actually Covers

AI consulting spans a wide range of engagement types, and understanding which type you need is the first step to choosing the right partner. The four primary engagement categories are: opportunity assessment (identifying where AI can create value in your specific business and what that value is worth), AI strategy (building a roadmap for AI adoption across functions, prioritising use cases, and defining governance), implementation (building or deploying specific AI systems, automations, or models), and change management (helping organisations adopt AI tools effectively and overcome resistance from teams whose workflows are changing).

Most businesses entering AI consulting for the first time need either an opportunity assessment or a strategy engagement — before you invest in implementation, you need clarity on where AI creates value versus where it creates noise. The opportunity assessment is the foundation. What are the repetitive, high-volume processes in your business where AI can reduce cost or increase throughput? Where are you currently making decisions with incomplete or delayed data that AI could address? What customer-facing interactions could be enhanced or automated without reducing quality?

The answers to those questions determine your AI roadmap — which use cases to pursue first, in what sequence, and with what success criteria. Skipping the strategy phase and going directly to implementation is the most common and costly mistake businesses make in AI adoption.

Types of AI Consultants

Big Four and Management Consulting Firms (Deloitte, McKinsey, Accenture, BCG) offer AI consulting at enterprise scale. They have deep cross-industry experience, large bench strength, and credibility for board-level conversations. The trade-offs: high cost (typically $250–$500k+ for meaningful engagements), slower delivery, and a tendency toward strategic frameworks over practical implementation. Best for large enterprises with complex organisational challenges and budget to match.

Boutique AI Specialists are smaller firms with deep expertise in specific AI domains — computer vision, NLP, marketing AI, process automation, or industry-specific AI applications. They typically move faster, cost less, and have more hands-on implementation experience than large consultancies. The trade-offs: limited bandwidth, narrower cross-functional perspective, and variable quality (the boutique space includes both exceptional specialists and under-qualified generalists with an "AI" rebrand).

AI-Native Agencies like Lumo AI blend marketing execution with AI implementation — building the automations, workflows, and AI-assisted systems that make marketing operations faster and more effective. This is distinct from pure strategy consulting; the deliverable is running systems, not strategy documents. For businesses whose AI needs are concentrated in marketing, sales, and customer engagement, AI-native agencies often deliver more practical value than traditional consultancies.

Typical Engagement Costs and What Drives Scope

Opportunity assessment / AI audit: $5,000–$25,000 depending on business complexity and depth. This should produce a prioritised list of AI use cases with estimated ROI, implementation complexity, and recommended sequencing — a concrete roadmap, not a generic AI overview.

AI strategy engagement: $25,000–$100,000 for mid-market businesses. Includes use case prioritisation, technology selection, governance framework, team capability assessment, and a detailed implementation roadmap. Big Four strategy engagements for enterprise clients can exceed $500k.

AI implementation project: highly variable. A marketing automation build using existing AI tools might be $15,000–$50,000. A custom machine learning model or enterprise AI platform deployment can run $200,000–$1M+. The primary cost drivers are: custom vs. off-the-shelf technology, data complexity (clean, structured data is dramatically cheaper to work with than messy, unstructured data), integration requirements, and how much change management is needed to get teams to adopt the new system.

What to Ask Before Signing

The questions that separate good AI consulting engagements from expensive disappointments: What will you deliver, in what format, by what date? AI consulting deliverables should be concrete — a working prototype, a prioritised roadmap document with ROI estimates, a configured automation system, or a training program for your team. "We'll provide AI guidance and support" is not a deliverable.

How will success be measured? Every engagement should define 2–3 measurable outcomes at the outset. For strategy engagements: clarity of roadmap, stakeholder alignment, quality of use case documentation. For implementation: the system works as specified, adoption rate, and measurable impact on the targeted process. If the consultant can't define how they'll demonstrate value, they probably can't deliver it.

Who on the consulting team will do the actual work? Senior AI consultants close deals; junior analysts often do delivery. Ask to meet the people who will be hands-on in your engagement before signing. In boutique firms especially, the quality difference between senior and junior practitioners is enormous, and the person who pitched you may not be the person who builds your system.

Red Flags in AI Consulting Proposals

Vague deliverables described in AI buzzwords. "We'll leverage generative AI and LLM capabilities to transform your customer experience" is meaningless without specifics about what system will be built, what data it will use, and what the user interaction will look like. Specificity is the mark of a consultant who knows what they're building; vagueness protects a consultant who doesn't.

Overconfidence about ROI. AI implementations regularly take longer and cost more than initial estimates, and the business value is often more modest in the first 12 months than the pitch suggested. A consultant who promises aggressive ROI timelines without detailed assumptions is selling, not advising. Good consultants present optimistic, base, and conservative scenarios with clear assumptions for each.

No mention of change management. The most technically excellent AI implementation fails if the people using it don't trust it, don't understand it, or don't change their workflows to take advantage of it. AI consulting proposals that say nothing about adoption, training, and organisational change are missing the hardest part of the problem. That's not a gap you want to discover after the contract is signed.