AI Business Consulting: When It's Worth It, When It's Not
AI enters every corner of business. But where does it actually help in consulting and where is it just noise? A guide based on daily use at Plan B AI Lab.
By Eitan Eshtemaker
Every other consultant today brags about 'AI business consulting.' Most of the time it means: I have ChatGPT and also a client. At Plan B we use AI every day - but we have clear lines about when it enters and when it doesn't.
AI is useful in business consulting in 3 main places: fast processing of large data (cash flow, customer messages, customer history), building custom bots for the business (initial response, lead screening), and automatic meeting summaries. AI does not replace judgment in strategic decisions or team conversations.
What AI can do in business consulting
When I sit with an owner and need to see if their numbers work - I can manually process 12 months of cash flow over hours, or upload it to AI and get an analysis of spending patterns within 5 minutes. That's not magic - it's manual computation that used to be expensive.
Same with customer inquiries. A business with 500 monthly inquiries can't personally classify them. AI can - at negligible cost. The result: we discover customers ask the same question 60% of the time and build a dedicated bot to handle it automatically.
What AI cannot do
Strategic decisions. Where the business goes in 3 years, which customers to refuse, when to fire an employee - these have psychological, visual, and value components that can't be reduced to an algorithm. AI presents data. The decision - a human makes.
Human conversations. When an employee needs hard feedback, AI can't do it. When a customer is unhappy - a human has to call. AI can prep the structure. Execution is human.
Where AI saves the most time in a small business
Classic case: an owner who manually responds to every SMS inquiry. A business with 50 inquiries a day burns 2-3 hours daily. Most inquiries repeat: hours, prices, availability. A custom bot can handle 60-70% automatically and route complexity to the owner.
Second case: lead screening. AI can question a lead automatically, remove leads not in category, and bring the owner only leads that passed the screen.
Is ChatGPT enough for business consulting?
Depends. If your question is 'what are the 5 steps to build a marketing plan' - ChatGPT gives an excellent answer. If the question is 'what to do with this specific customer, in this business' - ChatGPT doesn't know your context.
At Plan B we use ChatGPT, Claude, and local models. But we built a process that loads business context before every question. Without context, AI is generic. With context - specific.
How to identify a consultant using AI correctly
First sign: they can show AI artifacts specific to your business. Not 'I looked at ChatGPT' - but 'here's the analysis of your 24 months, here are the patterns we identified.'
Second sign: they know where AI doesn't fit. If a consultant says 'AI can do everything' - run. Third sign: real integration with CRM, with messaging, with business tools.
How much does it cost and what's the ROI?
Building a dedicated bot runs $1,500 to $4,500. If it saves the owner 10 hours/week, and the owner values their hour at $60 - savings are $2,400/month. ROI in 1-2 months.
Lead screening ROI: a business with 200 monthly leads dropping to 30 quality leads gets 3-5x output from the same sales team.
How to start with AI in your business
Step one is not 'which AI.' Step one is 'what's the most repetitive process I hate doing.' Write down 5 things. Track over a week: how long each takes.
Step two: take the first on the list and think about how AI fits. Step three: implement small (one process, 2-4 weeks), measure, then expand.