Maybe it’s that February is always slow and you never feel prepared for it.
Maybe it’s that one service you keep pushing that never seems to convert as well as the one you do without thinking.
Maybe it’s that your best customers seem to come from one particular neighborhood or job type and you’ve never quite followed that thread.
You have a sense of these patterns. You’ve been running this business long enough to feel them. What you might not have is a clear picture…the kind that lets you say with confidence: “here’s where my money actually comes from, here’s what’s growing, here’s what I should be doing more of next quarter.”
This example captured it perfectly: I worked with a client that was spending 80% of their time and resources prospecting for brand new clients. It “felt” productive and familiar. Yet, when we actually analyzed their sales, 70% of their business came from happy, repeat customers. That insight should really affect where they should spend their time and resources, right?
That clarity used to require a bookkeeper, a spreadsheet habit you never had, or a consultant you couldn’t justify. Now, the first step can be an honest conversation with an AI tool and whatever data you can scrape together in the next ten minutes.
The data doesn’t need to be clean, and it doesn’t need to be complete. It just needs to be honest.
Many small business owners don’t do sales trend analysis: 40% aren’t using data and analytics to enhance outputs. Not because it’s hard, but because it feels like something that requires a dashboard, a data analyst, or a tidy spreadsheet they’ve never gotten around to building.
None of that is true. Here’s what AI actually needs from you to do useful sales analysis.
YOU HAVE THIS | ALSO WORKS | WHAT CLAUDE DOES WITH IT |
• A list of jobs from the last 3–6 months • Invoices in QuickBooks, Excel, or even paper • A rough sense of your busiest and slowest months • Notes on which customers came back and which didn’t | • A mental summary typed into Claude as a paragraph • Screenshots of your invoicing app totals • A list of your top 10 jobs by value this year • Revenue by month, even if it’s approximate | • Identifies which services or job types drive the most revenue • Spots seasonal patterns you might have stopped noticing • Flags which customer types repeat and which don’t • Asks the follow-up questions that sharpen the picture |
Any of these is enough to start. | Rough numbers beat no numbers. | Analysis you’d pay a consultant for. |
The insight that surprises people most consistently: revenue and margin are not the same thing. Your biggest jobs by dollar value are often not your most profitable jobs per hour of actual work. AI can surface this from a basic job list in about 30 seconds. Most owners who see this output have an immediate, specific reaction about one service or customer type they’ve been undervaluing or over-serving.
The second insight is about patterns you’ve stopped seeing. When you’re inside a business every day, seasonal rhythms become background noise. You know February is slow, but you’ve stopped thinking about what that means for October, when you could be building the pipeline that gets you through it. An outside eye on your revenue by month, even rough numbers, spots these patterns in seconds.
The third is the follow-up question. The most useful thing about using AI for this isn’t the initial analysis. It’s that it will ask you something back. The fourth question in the prompt below specifically invites this — Claude tells you what it would need to sharpen the picture. That question almost always points at the number you’ve been avoiding.
A gut feeling about your business and six months of rough data is enough to start. The analysis gets sharper every time you add one more honest answer. |
Below are the four questions worth asking. These aren’t generic, they’re the ones that consistently produce a specific action for small service businesses, regardless of whether the data behind them is a clean export or a list typed from memory.
Which service makes me the most money per hour of work? Not per job — per hour. The answer is almost never what owners expect. | Which months are slow, and is the pattern consistent year over year? Consistent slow periods are plannable. Knowing them two months out changes everything. |
Which customers come back, and what do they have in common? Repeat customer patterns tell you where to point your referral and follow-up energy. | Where am I discounting most, and is it winning me the job or just costing me margin? This one usually produces a silence followed by a very specific action item. |
The prompt below structures all four into one conversation. Paste whatever you have. Answer what Claude asks back. The picture that comes out of it is more useful than anything you’d get from a month of squinting at a spreadsheet alone.
THIS WEEK’S PROMPT |
Open Claude. Paste whatever you have — a spreadsheet, a rough list, or just a paragraph describing your last six months. The fourth question in the prompt is the one that makes this a conversation, not a one-shot answer. I run a [type of business] in [city/region]. I want to understand my sales trends better. Here is my sales data. It may be rough — work with what I give you and ask me follow-up questions if you need more. [PASTE OR TYPE YOUR DATA HERE. Options:] • A list of jobs from the last 3–6 months with approximate values • Monthly revenue totals, even if estimated • A plain-English summary: “March was my best month, mostly (ex: new proposals awarded, new orders for XX, HVAC installs...)” • A copied table from Excel or QuickBooks Once you have reviewed my data, answer these four questions: 1. Which service or job type is generating the most revenue, and is that also where I’m spending the most time? 2. Do you see any seasonal patterns — slow periods or busy spikes — that I should be planning around? 3. Based on what I’ve shared, which part of my business looks like the strongest growth opportunity right now? 4. What is the one question you’d ask me that would most sharpen this analysis? Be direct. If the data is too thin to answer something confidently, say so and tell me what you’d need. |
TRY THIS |
→ Open your invoicing app, QuickBooks, or just your memory. Write down your top 10 jobs from the last six months — service type, rough value, whether the customer was new or returning. (10 minutes) → Paste that list into the prompt above. Run it. Read question 4 — the follow-up question Claude asks you — and answer it. The second pass is almost always more useful than the first. (10 minutes) → Take the growth opportunity it identifies. Write it on something physical. That’s your next quarter’s focus. (2 minutes) |
The analysis isn’t the point. The decision it makes obvious is the point. Most owners already know what they’re going to do differently once they see it clearly. They just needed someone to show them the data they already had. |