Everyone Selling AI Consulting Says You Need AI Consulting
Most AI consulting content exists to convince you to hire an AI consultant. This post does the opposite. We run an AI consultancy, and a meaningful share of our discovery calls end with us saying some version of "you don't need us yet."
That is not charity. Engagements with the wrong client fail, and failed engagements are expensive for everyone - in money, in time, and in a team that walks away believing AI doesn't work. So here is the honest filter we apply to ourselves, written down. If you land on the wrong side of it, we will point you to the free path instead.
If a consultant never tells you "you don't need us for this," you're not talking to a consultant. You're talking to a salesperson.
You Should NOT Hire Us (or Anyone) If...
1. You haven't tried the tools yourself yet
A Claude Cowork or ChatGPT Business seat costs less than a single hour of anyone's consulting time. If you have not spent a few weeks actually working with these tools on your own tasks, that is where your money goes first - not to us.
Hands-on time changes the questions you ask. Founders who show up having used Claude for a month scope better projects, spot better use cases, and get dramatically more out of an engagement. Start with our free resources, and if you want a running start, our open-source Claude skills library - 172 MIT-licensed skills you can install today - costs exactly nothing.
2. You're under about five people and budget-tight
At that size, you do not have an AI adoption problem. You have a "one curious person spends a weekend on this" opportunity. Consulting overhead - discovery, training sessions, documentation - is built for teams where knowledge has to travel between people. When the whole company fits at one table, it doesn't.
Self-serve instead. Anthropic Academy is free and covers everything from prompting to agents. Combine that with a paid Claude seat and you can get remarkably far. Come back when the team - and the coordination problem - has grown.
3. You want someone to "do AI" without touching your workflows
This one kills more engagements than any budget issue. AI creates value by changing how specific work gets done - how proposals get drafted, how tickets get triaged, how data moves between systems. If those workflows are off-limits, there is nothing left to improve.
No consultant can fix a process you won't let them see. If the honest internal answer is "we want the AI checkbox ticked without anyone changing how they work," save your money. The engagement will produce a deck and nothing else.
4. You expect AI to replace a team you haven't hired yet
We hear this pitch regularly: "We don't want to hire a sales team / support team / ops person - can AI just do it?" Usually, no. AI multiplies existing capability. It makes a good salesperson faster, a good support team more consistent, a good analyst more thorough.
It does not found your company for you. If nobody on the team knows what good looks like in a function, an agent cannot be trained to produce it, and nobody can tell when it is quietly failing. Hire the first human, then multiply them.
5. You just want a chatbot on your website
Buy one. Seriously - website chat widgets are a solved, productized category with good off-the-shelf options at software prices. Paying consulting rates for a commodity is bad for you and boring for us.
Custom consulting makes sense when the work is genuinely custom: your data, your systems, your workflows. A chat bubble in the corner of your homepage is none of those things.
6. Nobody internally owns the outcome
Every successful deployment we have done had an internal champion - someone with the authority and the motivation to make the new way of working stick after we leave. Not necessarily technical. Just accountable.
When a project is "everyone's priority," it is nobody's, and it dies the week the consultants roll off. If you cannot name the person who owns this internally, that is the thing to fix before you sign anything - with us or anyone else.
When a Consultancy IS the Right Call
Flip every filter above and you have the profile of a client who gets outsized value from an engagement:
- You've used the tools and hit the ceiling of what off-the-shelf gets you.
- You're a 10 to 200+ person team where knowledge has to move between people, and self-serve stops scaling.
- You have real workflows with repetitive knowledge work - proposals, reporting, intake, research, support - that eat skilled hours every week.
- Leadership is bought in and willing to let someone examine and change how the work actually gets done.
- You need custom work - MCP integrations with your systems, agents built around your data, training tailored to your roles.
- Someone owns the outcome and will still own it six months after the engagement ends.
We will be direct about the sales part: this profile is exactly the work we do at OneWave. Across 50+ deployments, from SMB to enterprise, the pattern holds - the clients who match this list get results, and the ones who don't, we turn away or redirect to free resources. Every member of our team is Claude-certified, and our core engagements are Claude consulting and team AI training built around your actual workflows.
The Honest Close
If you read the first half of this post and saw yourself, you now have a free roadmap: get a seat, use it for a month, work through Anthropic Academy, grab our open-source skills. That path is real and it works. No hard feelings - come back when the team grows.
If you read the second half and saw yourself, book a free discovery call. And if it turns out you're not a fit, we'll tell you on the call - not after the invoice.