A 1:1 AI consultation is worth it when you bring one specific decision or process and the cost of getting it wrong yourself (wasted subscriptions, staff hours, a build that stalls) is bigger than the fee. It's rarely worth it before you've tried the basic tools yourself, or if you arrive without a question you want answered.
Treat the question as a sum: what a call really costs you, what doing it alone costs, and where the two cross. The fees used below are illustrative, not anyone's actual prices. Put in the fee you've been quoted and your own numbers, and you'll have an answer for your business rather than a general one.
What a consultation really costs you
The fee is the obvious line. It isn't the only one, and it's often not the biggest.
| Cost | One-off or ongoing | Typical size |
|---|---|---|
| The consultant's fee | One-off | Whatever you've been quoted; free introductory calls exist but cover less |
| Your preparation | One-off | One to two hours gathering examples, volumes and your tool list |
| The call itself | One-off | Usually around an hour of your time |
| Acting on the advice | One-off, then light upkeep | From an afternoon of setup to a funded project, depending on what's decided |
| Any tools it leads you to buy | Ongoing | From nothing (using tools and seats you have) to a few hundred dollars a month |
The fourth line decides whether the others were worth paying. A call that ends in a clear plan nobody carries out costs the fee plus your time and returns nothing.
What working it out alone costs
Doing it yourself is free only if your time is. Count these honestly:
- Research hours. Reading comparisons, watching demos and trialling tools. A single decision, such as which tool should handle customer enquiries, can easily swallow a working day or two.
- Trials that turn into subscriptions. Free trials that roll into monthly charges nobody cancels.
- The wrong tool on annual billing. Three standard team seats at $20 a month on an annual plan is $720 committed before you know whether it fits.
- A build that stalls. A freelancer's automation that half works, has no owner and gets switched off in month two.
- Delay. The problem keeps costing you every week you don't fix it. A task that eats five hours a week costs over 40 hours across two months of indecision.
Added up, the DIY route can cost more than people notice. Take an illustrative domestic cleaning company whose owner wants AI to help with rota changes. Over three weeks she spends about nine evening hours on comparisons and demos, trials two scheduling tools, and forgets to cancel one, which charges $49 a month for four months before she spots it on a card statement. She ends up back on her original spreadsheet. At $40 an hour, that's $360 of her time plus $196 of charges: $556 spent without a decision. None of it was foolish; it's simply what a decision with no one to test it against tends to cost.
Break-even: how many hours a call has to save
The table below shows how many hours of your time a consultation must save (or equivalent spend it must avoid) to pay for itself, including two hours for preparation and the call. The formula is simple: fee divided by your hourly value, plus two. Your hourly value is what an hour of your time is worth to the business, not your salary.
| Illustrative fee | Your time worth $30 an hour | $50 an hour | $80 an hour |
|---|---|---|---|
| $150 | 7 hours | 5 hours | 4 hours |
| $300 | 12 hours | 8 hours | 6 hours |
| $600 | 22 hours | 14 hours | 10 hours |
Two things follow. First, a process that costs you three or more hours a week produces over twelve hours a month, so even a modest improvement clears most of this table within one to two months. Second, avoided spend counts dollar for dollar: cancelling one $50-a-month tool you didn't need recovers $600 a year. If you'd rather measure the task before you decide, setting a baseline takes a week of simple tallies.
Run one row through a real week. An illustrative mobile hairdresser values her time at $30 an hour and has been quoted $150, so the call must save about 7 hours. Booking and rebooking by text takes her roughly two hours a week, eight a month. If the advice cuts that by half (saved message templates, a booking link in every reply, a weekly reminder batch), she saves four hours a month and breaks even in the second month. If the advice cuts it by a quarter, it takes nearly four months. Both are worth knowing before booking: the first is a good bet, the second is one she might try alone first with free tools.
Three cafés, three different answers
These are illustrations, chosen to show how the same question lands differently depending on what the business brings.
A single-site café whose owner hasn't used AI yet
The owner wants to "get into AI" but hasn't used a chat assistant for anything beyond a test question. There's no specific process in mind and no tool bill to review.
Verdict: not yet. Spend two weeks using a free assistant on your own admin first: supplier emails, menu descriptions, the staff rota, replies to reviews. You'll find out what it's good and bad at for nothing, and you'll come back with real questions. Using AI on your own admin first gives you a route through those two weeks.
A café with a catering side
About 25 catering enquiries a week arrive by email. Each needs a reply with menu options, dietary questions, delivery details and a price. The owner spends roughly 15 minutes on each: about six hours a week. She's been sent a demo of catering software that charges monthly and isn't sure whether she needs it or whether her existing email and spreadsheet could do the job with AI added.
Verdict: likely worth it. There's a clear process, real volume, known hours and a purchase decision waiting. If the outcome cuts her time to about 2.5 hours a week, that's around 15 hours saved a month: roughly $750 of her time at $50 an hour, which clears even the $600 row in the first month. That depends on the change actually being made. If the advice is "buy the software", she's still spent the fee to avoid choosing blind.
A two-site café paying for overlapping tools
Two managers pay for personal ChatGPT Plus accounts at $20 a month each, the social scheduler has an AI add-on, and there's a separate review-reply tool. Nobody is sure which of them earns its keep.
Verdict: possibly, but try the free route first. Much of this is an audit you can run yourself in an afternoon, and auditing your AI subscriptions shows how. A call earns its fee here if the question goes beyond "which to cancel" to "what should replace them", such as moving both managers onto a business plan so client details stop sitting in personal accounts.
Six signs you're ready to get value from a call
Tick the ones that apply. Four or more and a consultation is likely to pay; fewer than three and you'll get more from a few weeks of your own experiments first.
- You can name the one process or decision you want help with.
- You know roughly how often it happens and how long it takes.
- You've tried at least one AI tool yourself, even briefly.
- You can share a handful of real examples, with personal details removed.
- You know what software you already pay for and roughly what it costs.
- Someone (you or a named person) has time in the next two weeks to act on the outcome.
The sixth matters most. If nobody has time to act in the next fortnight, book the call for when they will.
Getting the most from the hour
Before. Send a short description of the process, a few examples, your tool list and a budget range a couple of days ahead, so the time isn't spent on background. Preparing for an AI consultation has a checklist. Here is roughly what that note looks like from an illustrative two-person bike repair shop, including how the question sharpened once the owner wrote it down:
WHAT I WANT DECIDED
First version: "How can AI help a bike shop?"
Sharpened: "Should repair customers get AI-drafted status
texts from our booking tool, or is a fixed
template enough?"
THE PROCESS
About 35 repairs a week. Each customer gets 2-3 updates:
booked in, extra work found (needs a price approval),
ready to collect. The mechanic texts from his own phone,
about 4 hours a week, and customers still ring to ask.
EXAMPLES (attached, names and numbers removed)
6 real update texts, 2 of them extra-work approvals.
WHAT WE PAY FOR NOW
Booking and till software, $69 a month (has an unused
SMS add-on). Google Workspace Business Starter, 2 users.
BUDGET
Up to $50 a month to run; up to $500 one-off.
WHO ACTS ON IT
Workshop manager, Tuesday mornings for three weeks.
A consultant who reads that before the call can spend the hour on the actual choice, and may well tell the shop that a fixed template in its unused SMS add-on does most of the job.
During. Ask for decisions, not overviews. "Should we use our existing inbox or buy the catering software?" gets you further than "What can AI do for cafés?". Keep to one or two questions. Write down each decision and the reason as it's made. For the café with a catering side, the notes at the end of the hour might read:
DECISION 1 Don't buy the catering software for now.
Reason Enquiries already arrive by email; the software's main
feature is a form she doesn't need yet.
DECISION 2 Draft replies with the AI in the existing email account,
from a saved menu and price sheet.
Reason Business plan already paid for; no new tool to learn.
FIRST STEP Write the menu-and-prices sheet (owner, Thursday).
CHECK Any reply mentioning allergens is read twice before
sending. Revisit the software if enquiries pass 40 a week.
Five lines of decisions and reasons are worth more than three pages of general notes, because in a month's time the reason is what tells you whether the decision still holds. The "revisit if" line matters too: it records the point at which the answer changes, so you don't have to pay for the same advice twice.
After. Take one action within 48 hours, even a small one, while the reasoning is fresh. Then turn the rest into a short plan; turning the advice into a 30-day plan shows a format that fits on a page.
That's the shape my AI implementation consultation takes: finding the job where AI pays off, working out how to automate it or add AI to the tools you already use, and leaving you with something your team can run. If you have a process in mind, you can book a call here.
Free call, paid call or a full project?
| Type of help | Good for | Not good for |
|---|---|---|
| Free introductory call | Checking fit and whether the problem suits the person | Detailed advice on your process |
| Paid 1:1 consultation | A specific decision, a process review, a plan you can act on | Vague curiosity; a whole-business overhaul in one sitting |
| A project or engagement | Building and testing several automations, with handover | Deciding whether you need AI at all |
The free call is for deciding whether to talk further; don't expect it to answer the question you'd pay to have answered. For an illustrative independent optician asking about AI for recall letters and frame enquiries, a free call can reasonably settle "Is this the kind of thing you work on, and does it sound worth a proper look?" It won't settle "Which of these two tools should we use, and what should the recall letter say?", because answering that means reading the letters, the booking system's settings and a month of enquiries. A project makes sense only once you know what you want built, which is often what the paid call settles.
Why some calls don't pay for themselves
When a consultation disappoints, the cause is usually one of these, and most are avoidable before you book.
- Too many questions. Ten topics in an hour means six minutes each and no decisions. Pick one or two.
- The wrong person on the call. If the manager who actually handles the catering enquiries isn't there, the advice is built on the owner's summary of her job, which is rarely the whole picture.
- No examples. Advice on "our enquiries" without seeing any is generic by necessity.
- Buying on the same day. A recommendation to try a tool is a recommendation to trial it, not to sign an annual plan that afternoon. Picture an illustrative picture-framing shop told a meeting-notes tool might help with customer consultations. The owner signs up for a year that evening at $15 a month, $180 in all, then finds in week two that most consultations happen standing at the counter, where nobody is going to open a recording app. A two-week monthly trial would have cost $15 and given the same answer.
- Expecting a guarantee. Nobody honest can promise a specific number of hours saved before your process has been tested. Treat any such promise as a warning sign.
How to judge, 30 days later, whether it paid
Put a reminder in your diary for a month after the call and answer four questions:
- Did you make the decision you went in with? A clear no counts.
- Did you buy, cancel or avoid anything because of it? Add up the dollars.
- Has anything changed in the process? Compare the hours now with before.
- If nothing happened, why not? Usually the answer is time, not the advice.
Add the hours saved (at your hourly value) to the spend avoided, subtract the fee and your time, and you have your answer. Filled in for the café with a catering side, using the illustrative $300 fee, a first month might read:
| Line | Month one |
|---|---|
| Decision made? | Yes: kept email and spreadsheet, added AI-drafted replies; didn't buy the catering software |
| Hours saved | From 6 to 3 a week (not yet the 2.5 hoped for): about 13 hours, worth $650 at $50 an hour |
| Spend avoided | The catering software's quoted $89 a month (illustrative): $89 |
| Fee | −$300 |
| Her time: preparation, call and four hours of setup | −$300 |
| Net | +$139, then about +$740 a month while the saving holds, less a few dollars of AI running costs |
If the same café had filed the notes and changed nothing, the line would read −$400: the fee plus the two hours of preparation and call. If the result is negative because nothing was done, that's worth knowing before you pay for the next one. For a longer engagement, judging whether your consultant delivered value goes further.
Other questions about paying for a 1:1 call
Should I expect a written summary after the call?
It's reasonable to ask for one, and worth confirming before you book. Even if the consultant doesn't provide it, write your own within an hour of the call: each decision made, the first action and who does it, and anything you need to check. Advice you can't find a week later has mostly been wasted.
Can one call cover my whole business?
Not usefully. An hour spread across sales, admin, finance and marketing produces general observations you could have read anywhere. One call works best on one process or one decision, with a short list of where to look next. If you have several areas, rank them by hours lost each week and bring the top one.
What if the advice is not to use AI for that task?
Then the call may have saved you the most money of all. Plenty of tasks are better fixed with a simpler rule, a template, a change to the process or no change. A consultant who tells you that, and why, is doing the job properly. Count the tool or build you didn't buy in your sum.
Further reads
- What Happens in a 1:1 AI Implementation Consultation? — What actually happens during a 1:1 session.
- Is Paying for AI Help Worth It? Your Time vs an Expert's Fee — The wider sum: your time against an expert's fee.
- Course, 1:1 Call, Audit, or Project: Which AI Help Do You Need? — Course, call, audit or project: matching help to the problem.
- What Is the Cheapest Way to Get Expert AI Advice? — Lower-cost routes to expert input if budget is tight.
- Free AI Consultations: What You Get and How to Use Them Well — What a free introductory call will and won't give you.
- How Much Does an AI Consultant Cost for a Small Business? — What consultants typically charge and why fees vary.
- AI Tools and AI Development: The Complete 2026 Guide — the AI hub, including every tutorial in the AI-for-business series.