🎯 The business: you build the one-page numbers dashboard a shop owner, a two-location service business or a small nonprofit has never had. You pull sales, bookings, spending or donations out of the tools they already use, clean the data, define each number in plain words, and set it to update on its own every week. Then you sell a short monthly review to keep it right and explain what changed.
At a glance
| Money to start | ≈ $400 to $2,058 |
| Time to the first paid project | 2 to 6 weeks |
| Hours a week | 5 to 12 part time; 20 to 30 with a full client list |
| Skills that matter most | Spreadsheets, one dashboard tool, cleaning messy data, explaining numbers to people who hate numbers |
| What it can pay | About $2,600 a month before tax with five monthly clients and one new build (worked below) |
Is this for you?
It suits someone who enjoys finding out why two reports disagree, and who can sit with an owner and turn "I want to know how we’re doing" into five numbers. Business, economics, statistics and accounting students already have most of the skills.
It is a poor fit if you want pure design work or pure statistics. Most of the hours go into getting access to data and cleaning it, not into charts. And owners need you to explain the numbers patiently, more than once.
Who pays, and what
Online shops (Shopify or Square sales, ad spend, returns), restaurants and salons with two or three locations, home-service companies tracking jobs and close rates, agencies tracking hours against budgets, and nonprofits that must report outcomes to funders. Most of them have the data spread across a point-of-sale system, QuickBooks and a few spreadsheets, and no one has put it on one page.
The demand is there. In Bluevine’s 2026 survey of 942 small business owners, data analysis was the top use of AI, named by 39%. Upwork’s 2026 skills report still lists data analytics among the most-hired skills overall. AI now writes spreadsheet formulas and database queries in seconds, and the big dashboard tools are adding AI chat features, but someone still has to decide what counts as a sale, fix the duplicate customers and check that the total matches the bank. That is the job.
What they pay: research estimates put a small-business dashboard at about $500 to $3,000. Upwork’s rate guide (Nov 2026) lists data analysts at $20 to $50 an hour and data visualization specialists at $25 to $45 from a global pool; a local U.S. operator who meets the owner usually quotes above that.
What it takes to start
| LLC and city or county license | $150 to $500 |
| One paid AI account (Claude Pro, ChatGPT Plus or Gemini AI Pro) | About $200 to $240 a year |
| Power BI Pro for sharing your own demo reports (Looker Studio is free) | $0 to $168 a year |
| Data connector subscription, only if a client’s source needs one (example) | $0 to $300 first year |
| Errors and omissions insurance (an estimate; get quotes) | $0 to $700 first payment |
| Simple website, domain, business email | $50 to $150 |
| All-in | ≈ $400 to $2,058 |
Prices from the vendors’ pages in November 2026: Looker Studio is free with a Google account; Microsoft lists Power BI Pro at $14 a user a month and Premium Per User at $24, both billed yearly, and the free Power BI Desktop cannot share reports. Looker Studio Pro is a paid add-on; check the current price. Claude Pro is $200 a year paid annually; ChatGPT Plus is about $20 a month, so check it. See The AI toolkit and its costs.
Start to finish: the steps
1. Learn Sheets and one dashboard tool. Pick Looker Studio if your customers live in Google (Gmail, Sheets, Google Ads), or Power BI if they use Microsoft 365 and Excel. Learn to join two tables, remove duplicates and build a date filter. Use AI to explain formulas, then make sure you understand every one.
2. Build three practice dashboards. Use public data or a made-up shop’s sales. One for retail, one for a service business, one for a nonprofit. These become your portfolio, with no client data at risk.
3. Choose one kind of customer. "Weekly dashboards for multi-location salons" or "funder reports for food banks" is easier to sell than "data services". You will learn which systems that trade uses and reuse the setup.
4. Make the business legal. Register an LLC or start as a sole proprietor, get the local license and a business bank account, and price errors and omissions insurance. See Bookkeeping basics to keep your own numbers as clean as your clients’.
5. Run a paid numbers interview. One hour with the owner: what do you check on Monday morning, what decision does it drive, where does the number live today? Hand back a one-page plan with the five numbers, their exact definitions and the price to build it.
6. Get read-only access in the client’s name. Ask for a viewer or reporting login to their sales system and accounting app, never the owner’s password. Build the dashboard in an account the client owns, so it stays with them.
7. Clean, define, then build. Fix duplicates, refunds counted as sales and mismatched store names first. Write a definitions sheet ("Sales = paid orders minus refunds, before tax"). Check the dashboard total against the bank or the accounting report before anyone sees it.
8. Review it with the owner and cut. Show version one within a week. Remove anything they did not look at. Set the data to refresh automatically and teach one staff member to read it.
9. Sell a monthly review. A 30-minute call to walk through the month, fix anything that broke when an app changed, and add one improvement. This is where steady income comes from.
10. Grow into forecasting and alerts. Once the numbers are trusted, owners ask "what if" questions: staffing by day, reorder points, cash next month. Price those as small projects.
Pricing
| Paid numbers interview and one-page plan (credited to the build) | $250 to $400 |
| Dashboard from one or two data sources | $800 to $2,000 |
| Dashboard joining three or more sources | $2,000 to $4,000 |
| Monthly upkeep and a 30-minute review call | $150 to $400 |
| Extra analysis outside the plan | $50 to $100 an hour |
These are example prices for a U.S. operator selling locally, in line with the research range. Price on the decisions the dashboard supports, such as staffing, ad spend or reorders, not on hours. Do not promise that a dashboard will raise sales or cut costs by a set percentage; show the numbers and let the owner decide. More in Pricing AI and IT services.
The math
Take a part-time operator about a year in. Five clients on a $250 monthly review bring in $1,250. One new $1,500 dashboard a month brings the total to $2,750. Costs are about $30 for software, $20 for AI, $60 set aside for insurance and $40 for marketing, $150 in all. That leaves about $2,600 a month before tax for roughly 15 to 18 hours a week, including sales.
Expect one or two small builds in the first three months while you collect results to show. Set aside a quarter to a third of profit for income and self-employment tax.
Getting customers
Who to approach first. A business you have worked in, a family business, and a nonprofit where you volunteer. You already know their systems and their questions, and they will give you the read-only access a stranger would not get. Students: ask professors who consult for local firms, and the business school’s small-business clinic if your school has one.
The first ten customers. List 50 owner-run businesses with two or more locations or an online store, plus 15 local nonprofits. Send 30 personal emails, book 10 short coffee or video meetings, and give 1 talk at the chamber or a nonprofit network called "The five numbers to check every Monday". Expect 3 to 5 paid numbers interviews and 2 to 4 builds. Offer the first two at a fair discount in exchange for a case study.
Hi Marcus, I build simple weekly dashboards for owners with more than one location. It pulls sales from your point-of-sale system and costs from QuickBooks onto one page that updates itself, so Monday morning takes five minutes instead of an hour of spreadsheets. I’d start with a one-hour paid session to agree on the five numbers that matter to you, and you keep everything in your own accounts. Could I show you an example on a 15-minute call this week?
Channels, best to worst for this customer.
1. Referral partners: bookkeepers and accountants see owners struggling with numbers every month. 2. Nonprofit networks: your state’s nonprofit association and local funder meetings, where report deadlines create urgent demand. 3. Chamber of commerce and referral groups: owners buy from people they meet. 4. LinkedIn: post an anonymized before-and-after dashboard each week. 5. Google Business Profile: modest, but cheap. 6. Upwork: useful for practice and remote work, with global rates of $25 to $45 for visualization work. 7. Paid ads: last.
From first clients to referrals. After a month of use, ask the owner one question: what did you decide differently because of the dashboard? Their answer, with permission, is your case study. Anonymize the screenshot, add the quote, and send it to the bookkeeper who referred them. Offer each client a free month of reviews for an introduction that signs. More in the launch and marketing playbook.
Running it week to week
Early Monday, open every client’s dashboard and check that it refreshed and that the totals look sane against last week. A broken connection found by you is a small fix; one found by the owner is a lost client. Keep a log per client: data sources, logins (in a password manager), definitions and changes. Hold monthly review calls in the same week each month, and send a three-line summary afterwards.
Rules to check
No license is needed to build dashboards. The rules are about the data. Work under a written contract that covers confidentiality, who owns the work and what happens to data when you part ways; see Contracts, client data and AI rules. Do not paste customer lists or donor records into a consumer AI account; use a business plan whose terms exclude training, or describe the data shape to the AI instead of sharing the data. Watch sharing settings: Power BI’s "publish to web" and Looker Studio’s link sharing can make a report public to anyone with the link. Health data from a clinic makes you a HIPAA business associate with a signed agreement, and financial clients such as tax preparers must hold you to the FTC Safeguards Rule. Nonprofits often promise donors privacy; follow their policy. Most state privacy laws apply to larger businesses, but bigger clients will pass their duties to you in the contract.
What goes wrong
The dashboard total does not match the bank, and the owner stops trusting all of it; reconcile before you show anything. A connector breaks when the point-of-sale system updates, and nobody notices for a month; check every Monday. The owner asks for twenty tiles and reads none; start with five. A free tool changes its limits, or a client’s viewers need paid licenses you did not price in; confirm sharing costs before quoting. And a report gets shared with a public link by mistake; use named-user sharing by default.
Questions people ask
How much does a small business dashboard cost?
Research estimates put most small-business dashboards at about $500 to $3,000, depending on how many data sources must be joined and how messy they are. A monthly upkeep and review plan typically adds $150 to $400 (example range).
Looker Studio or Power BI: which should I learn?
Learn the one your customers already use. Looker Studio is free and fits businesses on Google Workspace, Sheets and Google Ads. Power BI fits Microsoft 365 offices; Desktop is free, but sharing reports needs a paid Pro or Premium Per User license.
Do I need to know SQL to build dashboards?
Not to start. Sheets, Looker Studio and Power BI let you build a lot without it, and AI can draft queries for you. SQL becomes useful once clients have databases, and you should understand any query before you run it on a client’s data.
Can AI build a dashboard for a business by itself?
It can draft charts and formulas quickly, and dashboard tools are adding AI features. It cannot decide which numbers matter to this owner, fix duplicate records it does not know are duplicates, or confirm the total matches the bank. That judgment is what clients pay for.
Can a college student build dashboards part time?
Yes. The work is deadline-based and can be done between classes, and it fits business, economics and statistics majors well. International students on F-1 visas generally cannot freelance for outside clients without work authorization; ask your school’s international student office first. See For students: visas, taxes, campus rules.
Free help
Microsoft Learn has free Power BI training paths, and Google’s Looker Studio help center has free tutorials and sample reports. SCORE mentors and your local Small Business Development Center help with pricing and contracts at no cost, and many business schools run small-business clinics where students practice on real local firms.
Checked in November 2026: Power BI prices on Microsoft’s page, Upwork rate ranges, and survey figures from Bluevine and Upwork. Looker Studio Pro and ChatGPT prices were not confirmed on the vendors’ pages; check them before you quote, and re-check every tool’s sharing and licensing rules each year.