A small business should build five tools first in Lovable: a quote request form that works out an estimate, a stock check page for the shop floor, a booking page, a read-only staff rota viewer and a customer feedback form. Each is one screen or two, uses plain data and can be tried on a free plan.
The Complete Excel Course with AI Integration
From formulas and pivot tables to Power Query, DAX, VBA and AI with Copilot, ChatGPT and Claude — hands-on from the first lesson.
By Sagnik Bhattacharya and Virginia Thorn, Primetime Emmy Award winner
Buy nowThe order matters more than the list. Start with the tool that saves the most repeated typing, build it narrowly, and use it for a fortnight before adding anything. Lovable turns descriptions into working apps quickly, which makes it easy to build too much; the businesses that get value build one small thing, check it, then build the next. The sections below give each tool a prompt, a worked example, a rough credit budget and its main way of going wrong.
How does Lovable charge credits when you build a tool?
Lovable charges credits per message: according to its documentation, a plan-mode message costs 1 credit plus any research Lovable runs while planning, and a build-mode message varies with the change. Its examples are 0.5 credits for changing a grey button, 0.9 for removing a footer, 1.2 for adding authentication and 2 for a landing page with images. Chat messages are priced per message, typically a fraction of a credit.
The Free plan carries 5 daily credits, up to 30 a month, with a monthly Cloud grant of 20 and an AI grant of 4, and Lovable describes those grants as a temporary offering subject to change. Pro is listed from $25 a month for 100 credits, or $21 a month billed yearly, with Business from $50. Cloud and AI usage in a live app is billed separately by use: database size, storage and, for built-in AI features, model calls. Unused monthly plan credits roll over while the subscription is active, expiring two months after issue on monthly plans.
A planning rule you can use for each tool below: count the messages you expect, multiply by an average of 1 credit, then add a third for corrections. Treat that as an assumption to test, not a measured figure. If your first prompt is vague, you will pay for several rounds of clarification; if it is specific, the same tool takes fewer messages. Our companion tutorial on Lovable plan prices and what credits cost covers the plan ladder in more detail.
How do you build a quote form that works out an estimate in Lovable?
Build a quote request form in Lovable by writing your pricing rules in a table and prompting for a one-page form that shows a price range straight away and saves each enquiry. A customer fills in a short form and you follow up the saved enquiry. Typical sources of repeat typing are landscaping, grooming, printing and event hire.
Main worked example. Take a garden centre that quotes for lawn laying. Today the owner answers each enquiry by phone: size, access, turf type, whether old lawn needs removing. An illustration of the rules, with made-up prices:
| Input | Rule |
|---|---|
| Lawn area in square metres | Standard turf $9 per square metre |
| Premium turf | Adds $3 per square metre |
| Old lawn removal | $4 per square metre |
| Delivery | $45 flat |
| Minimum job | $180 |
For 60 square metres of premium turf with old lawn removal: 60 x ($9 + $3 + $4) = $960, plus $45 delivery, so an estimate of $1,005. The app shows "Estimated $1,005 to $1,155" (the range is the estimate plus 15%) and says the final price follows a site visit.
Build a single-page quote form for a garden centre's lawn laying service.
Fields: name, phone, email, town or area, lawn size in square metres (number), turf type (standard or premium), remove old lawn (yes/no), preferred week.
Show a live estimate: standard $9/m2, premium adds $3/m2, removal $4/m2, delivery $45, minimum $180. Display the estimate as a range: estimate to estimate plus 15%.
Save each submission to a database table called quote_requests with a timestamp and status (new, contacted, won, lost).
Add a staff-only page, behind a login, that lists requests newest first with a status dropdown.
Do not show other customers' details to anyone not logged in as staff.
How to start. Paste that, change the numbers, then test five jobs against your own spreadsheet before touching styling. Expect 15 to 25 messages for the form, the database and the staff page.
Watch out. The estimate is only as good as the rules. If the form shows $1,005 and the real job costs $1,400 because of slopes you never asked about, you have created complaints. Add a field for the thing that most often changes your price, and keep the "final price follows a visit" line. Our quote form comparison with Bolt.new shows how another builder handles the same job.
A second example: the same form for a sports club. Swap the lawn rules for a membership enquiry. A club offers three tiers (junior $12 a month, adult $30, family $55) and a one-off joining fee of $20. The form asks who the membership is for, how many people, and which sport. The page then shows "Family membership: $55 a month plus $20 joining fee, first month $75" and saves the enquiry. The logic is simpler than the lawn quote, a lookup rather than a sum, which is why it is a good first prompt if you want a cheap trial of Lovable before a harder build. Notice that the output includes the one-off fee in the first-month figure, a detail a customer cares about and a vague prompt would miss.
How to test it in twenty minutes. Write six test enquiries on paper before you open the app, including one at the minimum job size, one with every extra ticked and one with a zero in the area field. Enter each, compare the estimate with your own sum, and record the result in a small table: input, expected, shown, pass or fail. The zero case matters most, because a form that returns "$0" or an error message invites a bad first impression. Ask Lovable to show a friendly message and no estimate when the area is below 5 square metres.
How do you build a stock check page for the shop floor in Lovable?
Build a stock check page in Lovable by importing a clean product spreadsheet and prompting for one mobile screen where staff type a product name or scan a code to see stock and shelf. It replaces "ask Priya, she knows where the lead leashes are".
Example. A pet shop holds about 400 products. The owner exports a spreadsheet with columns for product, code, shelf, quantity and reorder level, and wants staff to adjust a count with two taps when they sell or receive something.
Build a mobile-friendly stock page for a pet shop.
Import products from a CSV with columns: name, code, shelf, quantity, reorder_level.
Search by name or code as I type. Show quantity in red when it is at or below reorder_level.
Add plus and minus buttons that change quantity by 1 and record who changed it and when in a stock_log table.
Staff log in with email and password. Only the owner can import a CSV or edit reorder levels.
Credits. Roughly 20 to 30 messages, plus a few for the import step, which is where most re-prompting happens because real spreadsheets have blank rows and mixed formats. Clean the CSV first: one header row, no merged cells, codes as text.
Watch out. Two people editing the same count at once. Ask Lovable to log every change rather than overwrite a number silently, so you can reconstruct what happened. A stock page also drifts from reality unless someone counts shelves; schedule a fortnightly count of ten fast-moving lines and compare it with the app. If the numbers differ by more than a couple of units, the habit of updating needs fixing before the app does.
A second example: a gift shop's "is it in?" page. A gift shop does not need scanning. It keeps a list of 150 seasonal lines, and customers ring to ask whether the advent calendars have arrived. A public version of the stock page shows only three states per item, "in stock", "low" and "sold out", with no quantities, and updates when staff change the count. The prompt adds one line: "On the public page show only the status word, never the quantity or the shelf." That one sentence keeps your stock levels out of competitors' hands and cuts the calls.
A filled-in stock check test. Take twelve products, set the quantity of three to the reorder level, and set one to zero. Then walk the page like staff would:
Test Expected Result
Search "lead" 3 products, 1 shown in red pass
Press minus on a quantity of 0 Stays at 0, shows a warning FAIL (went to -1)
Open the page logged out Redirects to login pass
Import a CSV with a blank row Skips it, reports 1 skipped pass
Two phones press plus together Both changes logged pass
The failed row is typical, and fixing it is a single message: "Never let quantity go below zero; show a warning instead." Finding it on day one costs a credit. Finding it after three weeks of inflated negative stock costs an afternoon.
How do you build a booking page with confirmations in Lovable?
Build a booking page in Lovable by prompting for the slots, a block on double bookings, a confirmation email and an admin diary behind a login. Customers choose a slot, give contact details and receive a confirmation; you see a diary of what is booked. Typical uses are grooming, workshops, court hire and consultations.
Example. A sports club hires out two five-a-side pitches in one-hour slots from 5pm to 10pm on weekdays. Members pick a pitch, a date and an hour. The club wants no double bookings and a hold on slots for 15 minutes while someone completes the form.
Build a booking page for a sports club with two pitches (Pitch A, Pitch B).
Slots: weekdays 17:00 to 22:00 in one-hour blocks, up to 28 days ahead.
A visitor chooses pitch, date, hour and enters name, phone, email. Prevent two bookings of the same pitch and hour.
Send a confirmation email. Add an admin page, behind a login, to see the week as a grid and cancel a booking.
Never show other people's names or phone numbers on the public page.
Credits. This is the biggest of the five: 30 to 60 messages is a fair planning range because of the availability logic and email. Our walk-through of a booking portal without a developer covers it in depth, and sending confirmations from your own domain covers email.
Watch out. Email delivery. Lovable's documentation on its Resend connector says that an unverified domain delivers only to the Resend account owner, so a confirmation that reaches you will not reach customers until the domain is verified. Test with a second address that is not yours. Custom domains also need a paid plan, so the free version will carry a Lovable address.
A second example: a repeating class timetable. A sports club runs the same weekly classes: Monday juniors at 4pm, Wednesday adults at 7pm and Saturday family at 10am. Instead of one-off bookings, the page lists classes with spaces left (a maximum of 16 each) and lets a member reserve a place. The extra rule is the capacity counter: "Show spaces left; when 0, show Full and offer the waiting list." It is the same booking logic with fixed times, and it cuts the build because there is no free-choice calendar.
The confirmation message to ask for. Spell out the email text in the prompt so every confirmation reads the way you would write it. A version you can adapt:
Subject: Your pitch is booked: [pitch] on [date] at [time]
Hello [first name],
You're booked on [pitch] on [date] from [time] to [end time].
Need to change it? Reply to this email at least 24 hours before and we'll move you.
See you there, [your name]
Putting the cancellation rule in the email means you do not have to remember to say it, and the 24 hours is yours to set. If Lovable writes friendlier or longer copy than you wanted, tell it the word limit.
How do you build a staff rota viewer in Lovable?
Build a staff rota viewer in Lovable as a read-only page behind a staff login that shows each person's shifts first, then the full week, so staff stop messaging "what time am I in on Saturday?". You keep building the rota in a spreadsheet or tool you like; the app only displays it.
Example. A gift shop with seven staff keeps its rota in a spreadsheet. Each week the manager pastes a table of name, day, start and end times into the app. Staff log in and see their own shifts first, then the whole week.
Build a staff rota viewer for a small shop.
Managers can paste or upload a weekly table with columns: name, date, start, end, role.
Staff log in with email. After login they see "My shifts this week" at the top, then a full-week grid.
Highlight today. Staff cannot edit anything. Managers can edit and publish a week; unpublished weeks are hidden from staff.
Show times in 12-hour format.
Credits. The simplest of the five, around 15 to 25 messages. The login and the publish switch are the parts that take time.
Watch out. Two sources of truth. If the manager changes the spreadsheet but forgets to republish, staff see an old week and turn up at the wrong time. Add a "last published" timestamp in large type and a rule that any change made after publishing is announced in the team chat as well. Also think about privacy: a full-week grid shows everyone's hours, so confirm your staff are comfortable with that and that only logged-in staff can reach it.
A second example: a handover view for part-time staff. A garden centre has weekend staff who see the rota only on Fridays. Add a "my next shift" box on the login screen and a short notes field the manager can use for the day, such as "frost forecast, cover the seedlings by 4pm". It costs a handful of messages and replaces a group chat that people scroll past. Keep notes factual and operational; a notes field tends to collect remarks you would not want on a shared screen.
Checking that the rota is right. After publishing a test week, ask one member of staff to find their Saturday shift without any help from you and time them. If it takes more than ten seconds on a phone, move "My shifts this week" higher or enlarge the text. A tool that people need a briefing to use is not replacing the old way of asking.
If the rota or stock list starts life in a spreadsheet, The Complete Excel Course with AI Integration on Coding Liquids covers the spreadsheet side.
How do you build a customer feedback form in Lovable?
Build a customer feedback form in Lovable as a short public form that saves a score and comment, with a staff page behind a login that puts low scores at the top. Link it from receipts, a QR code or an email.
Example. A garden centre prints a QR code on receipts: "How did we do?" with a 1 to 5 score, an optional comment and an optional phone number if the customer wants a call back. Low scores should appear at the top of the owner's list.
Build a feedback form for a garden centre.
Fields: score 1 to 5 (required), comment (optional), department (plants, tools, cafe, other), phone (optional, only used for a call back).
Save to a feedback table with date. Staff page behind a login: list newest first, scores of 1 or 2 pinned to the top in red, a "called back" checkbox, and a count by department for the last 30 days.
Public page shows only the form and a thank-you message, never other people's comments.
Optional AI step. Lovable can add AI features inside an app, which draw on the monthly AI grant and then paid use. A cautious version summarises the last 30 days of comments into five themes. Try it with 40 real comments and read every theme against the source before you trust it. Do not let an AI send replies to customers unreviewed.
Credits. About 15 to 25 messages without the AI summary, perhaps 10 more with it.
Watch out. Collecting phone numbers carries responsibility. Say on the form why you ask, delete the number after the call back, and keep comments free of anything you would not want the public to see. If you have customers in the EU, data-protection law applies, and a specialist can confirm what notice you need.
A second example: reading the feedback. After a month, a garden centre has 62 responses. Scores of 5 and 4 are 48 of them, 3 is 6, and 1 or 2 is 8. Five of the eight low scores mention the cafe queue at weekends. That is a pattern the owner could not see from receipts: the average score looks healthy, while the department count shows one place with a problem. The action is cheap, a second till on Saturdays, and the next month's count shows whether it worked. The tool's value is in the department field and the filter by score, not in the thank-you page.
Which of the five tools should you build first?
Build first the tool that loses the most minutes per week, not the one that seems most interesting. Count how often the job happens, how long each occurrence takes and how many people are involved. An illustrative comparison for a pet shop:
| Tool | Times per week | Minutes each | Minutes saved if it halves the work | Build effort |
|---|---|---|---|---|
| Quote form | 12 | 10 | 60 | Medium |
| Stock check | 60 | 3 | 90 | Medium |
| Booking page | 25 | 6 | 75 | High |
| Rota viewer | 20 | 2 | 20 | Low |
| Feedback form | 10 | 5 | 25 | Low |
On those assumptions the stock check wins on minutes saved, but it also needs the cleanest data and the most discipline from staff. The quote form has a clearer payoff because it touches new business. If I were advising this shop I would build the quote form first, since a faster quote often wins the job, and the rota viewer second as an easy confidence builder. Swap in your own counts; the point is to put numbers beside the list.
A four-week build plan for a first tool
People often ask how long the first one takes. Here is a plan for a part-time builder, an hour or two on each working day:
- Week 1, decide and prepare. Pick one tool from the ranking table. Write its rules in a spreadsheet, list the fields, and write one sentence on who sees what. Create a free Lovable account and make no app yet.
- Week 2, build the narrow version. Paste one carefully written prompt, then fix what breaks one message at a time. Resist adding features. Stop when the six test cases pass.
- Week 3, use it for real alongside the old way. Run it in parallel with your phone calls or spreadsheet. Keep a tally of problems: wrong results, confusing wording, things people asked for.
- Week 4, decide. If the tally is short, switch the old way off and consider a custom domain on Pro. If it is long, fix the top three problems or decide a ready-made product does the job better.
After that month you will know what credits a tool really costs you, which is a better number than any estimate in a tutorial, including this one. Write it down next to the plan you are on so the next build starts from evidence.
Rules that apply to all five
- Write the rules in a spreadsheet first. Prices, slot lengths and reorder levels belong in a table you can read. A prompt that pastes the table beats one that describes it loosely.
- Decide who can see what before the first prompt. Customers see their own entry and nothing else; staff see operational data; the owner alone sees exports. State this in the prompt.
- Test logged out. Open a private browser window and try every address in the app. Nothing staff-only should load.
- Test on a phone. Most customers and floor staff will use one.
- Keep a change log. One line per change: date, what, why. When an app breaks, the last change is the first suspect.
- Use a paid plan only when it earns its place. Custom domains and badge removal are Pro features; user roles also start on Pro, and role-based access and single sign-on sit on Business. Most small shops never need Business. See Lovable Business vs Pro.
What is the commonest mistake when building these tools in Lovable?
The commonest mistake across all five tools is a page left public because the first prompt never said who could view it. In one illustration, a pet shop owner builds the stock page, tests it logged in as herself and shares the link with staff. Weeks later a supplier mentions that they can see the shop's shelf quantities and reorder levels. The app's stock list had been made public, since the simplest build was an open page.
Nothing was lost beyond some privacy, but the lesson is cheap to apply. Put the sentence "only logged-in staff can open this page" in every prompt, then open the link in a private window to prove it. Do the same after every later change, since a new feature can add a new public address. Add the check to your change log so it is done by habit rather than memory, and ask a second person to try the link from their own phone once a month.
When a freelancer is the better answer
Some jobs outgrow a prompt. Take-payment flows, anything touching medical or children's records, and an app that three people must rely on every day are all candidates for expert review. A cheap pattern is to build the first version yourself, then pay a freelancer for a short review of data access and email delivery rather than the whole build. Our tutorials on credits versus a freelancer and on hiring a Lovable expert to finish a half-built app describe the brief and the budget.
If you want help picking which job to automate first, that is what my AI implementation consultation covers; you can start from the contact page. Otherwise, pick one tool from this list this week, write its rules in a table, and spend your free daily credits on a first version before deciding whether to pay.
Before you start building in Lovable
Can I build these on Lovable's free plan?
Yes, you can build and test on Free: Lovable's documentation lists 5 daily credits (up to 30 a month) plus a small monthly Cloud and AI grant. Custom domains and removing the Lovable badge need Pro, which Lovable lists from $25 a month for 100 credits ($21 a month billed yearly).
Do I need to know how to code?
No, but you do need to describe the job precisely and test it like a customer would. Lovable builds from plain-English prompts. The skill that matters is writing down fields, rules and who may see what, and noticing when the result does not match.
Is it safe to put customer details in a Lovable app?
Only if you check who can read the data. Ask Lovable which pages and tables are public, test with a logged-out browser window, and keep sensitive details out of the first version. For anything involving payments, health or children's data, get a specialist to review it before launch.
How many credits will one tool use?
It varies by prompt. Lovable's documentation gives examples from 0.5 credits for a small button change to about 2 for a landing page with images, with plan-mode messages at 1 credit plus research. Budget by counting the messages you expect, then add a third for fixes.
Which five tools should a small business build first in Lovable?
A quote request form that works out an estimate, a stock check page for the shop floor, a booking page, a read-only staff rota viewer and a customer feedback form. Rank them by minutes lost per week, build the winner narrowly, and use it for a fortnight before adding another.
Further reads
- Bubble vs Lovable for a Business App: Control, Cost and Upkeep — Weigh upkeep and control against Bubble.
- Is Replit Worth It for a Non-Coder Building a Business Tool? — Replit for an owner who cannot code: what a simple internal tool costs, where it gets risky, and when a freelancer is cheaper.
Browse all AI-for-business tutorials
Sources: Lovable documentation pages on plans, credits and usage, subscription plans, and the Resend integration, read in November 2026; Lovable's pricing page did not render plan details, so prices come from its documentation.
