For a small trade business that logs jobs from the van, AppSheet is usually the better base: apps run offline and sync later, and Core costs nothing extra on most paid Google Workspace plans ($10 a user a month otherwise). Airtable has the cleaner data model and AI fields, but charges $20 a seat and does not support offline use.
The deciding question is not features, it is where your people stand when they update a job. An engineer in a basement with no signal needs a different tool from an office team that books and invoices all day. Most owners compare price lists, pick on brand, and find the real difference in the first week of use.
The setup used for every figure below
Take a six-person heating and plumbing firm, as an illustration. Two people work in the office (the owner and a part-time coordinator) and four engineers drive vans between 8 and 10 jobs a week each. That is roughly 40 jobs a week, or about 2,000 a year. Each job has a customer, a site address, a job type, a status, parts used, photos, notes and a signature at the end.
The tracker has to do five things: show today's jobs per engineer, let engineers update status and add photos on site, flag jobs waiting for parts, hand finished jobs to invoicing, and still work when a boiler cupboard blocks the signal. Everything else is a bonus.
The tables both tools need
Whichever you choose, the data is the same four tables. Getting this layout right matters more than the tool.
| Table | What a row is | Key fields |
|---|---|---|
| Customers | One account holder | Name, phone, billing contact, notes |
| Sites | One address, linked to a customer | Address, access notes, boiler make |
| Jobs | One visit or piece of work | Site, engineer, date, type, status, photos, signature, time on site |
| Parts used | One line per part on a job | Job, part name, quantity, unit cost |
Airtable handles this with linked-record fields and rollups, so a job can total its own parts cost. AppSheet does the same with Ref columns that point at another table, usually sitting on a Google Sheet or on AppSheet's own database. Both can do it. The difference is how comfortable the office finds editing it and how the engineers experience it.
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 nowWhat each platform costs this firm per month
Airtable's Team plan is $20 per seat a month on annual billing, or $24 billed monthly, and on Team a seat is charged for every collaborator with commenter, editor or creator permission. Read-only collaborators, form submissions and share links are free. AppSheet lists Starter at $5, Core at $10 and Enterprise Plus at $20 per user a month, and Google says Core is included in most paid Workspace plans.
| Scenario (six people) | Monthly | Per year |
|---|---|---|
| Airtable Team, all six edit, annual billing | 6 × $20 = $120 | $1,440 |
| Airtable Team, all six edit, billed monthly | 6 × $24 = $144 | $1,728 |
| Airtable Team, two office editors, engineers use a form | 2 × $20 = $40 | $480 |
| AppSheet Core, licence included in Workspace | $0 extra | $0 extra |
| AppSheet Core bought separately | 6 × $10 = $60 | $720 |
| AppSheet Starter bought separately | 6 × $5 = $30 | $360 |
Figures are list prices checked in November 2026 and exclude any sales tax. The row that surprises people is the third one. If engineers only create records, a form is free on Airtable and the bill drops by two thirds. But an engineer who must open yesterday's job and change its status is an editor, and that is a paid seat.
The sum that changes at ten people
Say the firm adds two engineers and a second coordinator. All-editor Airtable becomes 9 × $20 = $180 a month, $2,160 a year. AppSheet on Workspace stays at $0 extra for anyone with a Workspace account, and $90 a month if bought at $10 a user. Per-seat tools get dearer in step with the headcount, which is why the licence question deserves a proper look before you build.
One caveat for AppSheet: outside people. A subcontractor with no Workspace account in your domain is not covered by the included Core licence. Google's licensing notes say an app's users need a licence that matches its creator's (Core, or a User Pass, for a Core creator's app), and the free Core licences cover only verified users in your own Workspace domain. Count your regular subcontractors before assuming the bill is zero.
Offline use is where the two diverge most
AppSheet's pricing page lists offline running with background sync among the features of every paid plan. In practice it works like this: the engineer opens the app while online so it downloads the definition and data, then keeps using it with no signal. Changes queue on the phone and send when the connection returns.
Airtable is different. Its technical-requirements help page states that offline access is not supported and lists internet access as a requirement for the iPhone, Android, Mac and Windows apps. The workarounds it offers are CSV exports or an API integration that saves copies of your records, which help an engineer read a job sheet but not update one. Treat Airtable as online-only, with forms and interfaces that need a connection to load and to submit.
AppSheet's offline rules that catch people out
- Launch online first. The app has to download its data once before offline use works. A new engineer who installs it in a cellar will see nothing.
- Starting from the home screen offline is a setting. Turn on "The app can start when offline" before installing the shortcut, or delete the shortcut and reinstall.
- Photos are not stored by default. Switch on "Store content for offline use", or job photos will not show without a signal. Large photo sets can take several minutes to download, so engineers should sync before leaving the depot.
- Maps do not draw offline. A location can be captured with no signal, but the map will not display it. Put the address as text on the job card.
- Do not stay offline for days. Google warns that an app left offline for days or weeks falls out of date, and older offline edits can overwrite other people's changes.
Two edits to the same job
Here is a realistic collision. An engineer marks job 1042 "Complete" in a basement at 10:05 with no signal. At 10:20 the coordinator, online, changes the same job to "Awaiting parts" because the customer rang. The engineer's phone syncs at 11:00 and the status flips back to "Complete". Nothing is broken; the later sync simply won.
The fix is a design choice, not a setting. Give each person their own fields: the engineer owns "Site outcome" and "Engineer notes", the office owns "Status" and "Invoice stage". Two people then rarely edit the same column of the same row, and when they do the rule is obvious. Do the same on Airtable, where collisions are rarer because everyone is online, but a stale browser tab still causes them.
Record and row limits for a busy trade firm
This is the part most comparisons skip. With about 2,000 jobs a year and a handful of parts lines per job, the firm creates 2,000 job rows plus perhaps 6,000 parts rows annually.
| Platform and plan | Stated limit | Time to reach it (about 8,000 rows a year in one base or database) |
|---|---|---|
| Airtable Free | 1,000 records per base | About six weeks |
| Airtable Team | 50,000 records per base | About six years |
| Airtable Business | 125,000 records per base | About 15 years |
| AppSheet Starter | 5 AppSheet databases of 2,500 rows each | About four months; no such cap on a Google Sheet |
| AppSheet Core | 10 AppSheet databases of 2,500 rows each | About four months; no such cap on a Google Sheet |
| AppSheet Enterprise Plus | Up to 200 AppSheet databases of 200,000 rows each | Not a concern |
On AppSheet the 2,500-row line is the one to understand before you build, because it belongs to AppSheet's own built-in database, not to every data source. Google's limits page sets it per AppSheet database on Starter and Core, with at most 20 tables per database and 100 columns per table on every plan. Kept in one AppSheet database, this firm's jobs and parts lines together would pass 2,500 rows in about four months. A tracker whose tables sit on a Google Sheet is outside that cap; the sheet's own size and the time each phone takes to sync become the practical ceiling instead. On Starter or Core that is a good reason to build on a sheet, and to archive jobs older than a year into a separate tab so the app stays quick.
Loading old jobs into the wrong place (an illustration)
A firm builds the tracker in January on an AppSheet database, because Gemini offered one, and imports two years of old jobs as history: 3,200 rows before the first live job. Google's limits page says that once a database reaches its cap, existing data stays readable and editable but no new data can be added, so the first new booking has nowhere to go. The history was never needed in the live app. Moving it to an "Archive" Google Sheet, and keeping only open and recent jobs in the database, fixes it in an afternoon; building on a sheet from the start avoids it.
On Airtable the ceiling is the plan's per-base record count, and Airtable's plans article says that limit is cumulative across every table in the base. Check the base's total before a busy month and archive or upgrade well before you reach it.
AI features: what each platform actually offers
The AI offers are quite different in kind. Airtable puts AI inside the data: AI fields, which Airtable's help centre now calls field agents and offers on every paid plan, read other fields in a record and write a result, such as a category, a summary or a drafted message. They draw on monthly AI credits, which Airtable lists as 500 per editor on Free, 15,000 per billable collaborator on Team and 20,000 per paid user on Business, shared across the workspace rather than held by each person. Airtable's billing article puts a simple categorisation at about 1 credit a run; a short rewrite like the one below may cost more, so run it on 20 jobs and read the credit meter before switching it on for every job. For the credit arithmetic, see the tutorial on estimating Airtable AI credit costs.
AppSheet's AI shows up in two places. Gemini for App Creation lets you describe an app in plain words and returns a structure you review before creating it; Google says it is available on paid AppSheet accounts, including Workspace customers on Core. The Smart Assistant view, available from Core upwards, is a conversational way for users to ask questions of their data. Machine learning models and text recognition sit on Enterprise Plus only.
Airtable AI field on engineer notes
A concrete use: engineers type rough notes, and an AI field turns them into a customer-ready line for the invoice. An illustrative prompt for the field:
Read {Engineer notes} and {Job type}. Write one plain sentence a customer would understand describing the work done. No jargon, no part numbers, no promises about future faults. If the notes are empty, return "No notes".Illustrative input: "swapped diverter vlv, flushed sys, pressure back to 1.2 bar, advised cust re magnetic filter". Illustrative output: "We replaced the diverter valve, flushed the heating system and restored normal pressure, and recommended fitting a magnetic filter." What you would fix: if the notes had said "customer declined filter", a model might smooth that over. Add a rule such as "Never state that advice was accepted unless the notes say so."
AppSheet's Gemini prompt for the first draft
For AppSheet the useful AI moment is the first build. An illustrative description to paste in:
Build a job tracker for a small heating and plumbing firm. Tables: Customers, Sites (linked to a customer), Jobs (linked to a site and an engineer), Parts used (linked to a job). Jobs need a status of Booked, On site, Awaiting parts, Complete or Invoiced; a photo; a customer signature; and the time the engineer arrived. Engineers should see only their own jobs for today.Expect a workable skeleton with sample data that may be inaccurate, as Google's own help page warns. What you would fix: Gemini tends to pick sensible column types but may set the "engineer sees only own jobs" rule loosely. Open the slice (a filtered view of a table) and check the filter by logging in as an engineer, not as yourself.
Building and living with each one
Airtable is quicker to start. You can import a spreadsheet, add linked fields, build a calendar view and a form in an hour, and the office will understand it immediately. Its interface designer gives engineers a tidier screen than the raw grid. The trade-off is that field-capture quality depends on a connection, and every editing person is a seat.
AppSheet is slower to shape and rewards patience. Every column has a type, every behaviour is an action or a format rule, and the editor can feel fussy. In return you get native-feeling phone features such as signature capture, location, photos and, from Core, barcode scanning. The app lives on a spreadsheet the office can open at any time, which is also its safety net: if AppSheet disappeared tomorrow, the data is in a sheet you own.
Who maintains it when you are busy
Ask a blunt question: who changes this when a new job type appears in March? On Airtable, the coordinator can add a field and a view without help. On AppSheet, the same change often means editing a column type, a slice and a form view, which is manageable for someone willing to learn but a real task for someone who is not. Plan for one named owner on either platform, and an hour a month of upkeep.
Scoring the two for this firm
A simple weighted score keeps the choice honest. Give each platform 1 to 5 on each row, multiply by the weight, and add up. These are illustrative scores for the trade-firm setup, with the weighted result in brackets; change them to match yours.
| Criterion | Weight | Airtable | AppSheet |
|---|---|---|---|
| Works with no signal | 5 | 1 (5) | 4 (20) |
| Cost for six people | 4 | 2 (8) | 5 (20) |
| Office finds it easy to change | 3 | 5 (15) | 3 (9) |
| AI inside the data | 2 | 5 (10) | 3 (6) |
| Room to grow in rows | 3 | 4 (12) | 3 (9) |
| Total | 50 | 64 |
The result flips only if you change the weights. A firm of office staff who take bookings by phone and rarely work off-site should weight offline at 1, which narrows the gap to 48 against 46; raise "office finds it easy to change" to 5 as well and Airtable edges ahead, 56 to 54. A firm whose work is mostly in cellars, lofts and remote sites should weight it at 5 and probably not consider Airtable for the field side at all.
A mixed setup that some firms choose
You do not have to choose one. A common compromise is Airtable for the office (customers, quotes, scheduling, AI-written summaries) and a form or small app for engineers that feeds it. Airtable forms are free for submitters but need a signal to load and to send. The cost is a second thing to maintain, and the engineer cannot open yesterday's job to continue it.
The other mix is AppSheet for the field plus a Google Sheet that the office reads directly. Because AppSheet writes to the sheet, the owner can build reports with normal spreadsheet formulas, pivot tables and charts, without touching the app. If your customers also need to see job progress, compare the portal options in Airtable client portal: Portals add-on, Softr or Glide.
A one-week pilot that settles it
Do not decide from a feature list. Run both with the same 20 real jobs for a week and measure four things.
- Day 1, load 20 jobs into each tool, including three at sites you know have poor signal.
- Days 2 to 4, engineers update status, add photos and a signature on every job, in whichever tool you assign them for the day.
- Day 5, count failures. How many updates were lost, delayed or duplicated? How many took more than 60 seconds on site?
- Day 6, the office tries to change something. Add a new job type and a new status in each tool. Time it.
- Day 7, price it. Count the seats you would really need, including subcontractors, and put them in the table above.
Whichever loses the field test usually cannot win on price alone, because lost updates cost more than seats. A single missed "awaiting parts" flag that delays a return visit by a day costs more than a month of licences.
Moving off a spreadsheet you already have
If the tracker already lives in a spreadsheet, the migration is the cheap part. Clean the sheet first: one row per job, one date format, status words from a fixed list. The tidy-up is covered in cleaning messy data with AI, and it matters more than the platform. A messy sheet imported into Airtable or AppSheet produces a messy tracker with nicer buttons.
For an app you would rather describe than build, Glide's AI builder is a third route worth ten minutes of comparison, and Airtable Team versus Business explains when the higher plan pays for itself.
What I would choose for this firm, and when I would switch
For the six-person firm above I would pick AppSheet on the existing Workspace licence, keep the Jobs table lean with a yearly archive, and give engineers their own columns to avoid overwrites. I would choose Airtable if the engineers were rarely off-grid, the office wanted to change the structure often, or AI fields summarising notes were a priority from day one.
Switch signals to watch for: AppSheet becoming a burden on one person, or Airtable seat costs passing what a ready-made trade system such as those compared in Jobber, Housecall Pro and ServiceTitan would cost. At that point you have outgrown a home-built tracker, and buying a finished product may be cheaper than maintaining one.
Questions owners ask before picking a job-tracker platform
Can AppSheet really be free if I already use Google Workspace?
Google says most paid Workspace editions include an AppSheet Core licence for each verified user in your domain, so the licence itself costs nothing extra. Check your edition in the admin console. Outside contractors and customers are different: they need their own licence or a User Pass, so count them before assuming zero.
Does Airtable work offline on a phone?
No. Airtable's technical-requirements help page says offline access is not supported and lists internet access as a requirement for its iPhone and Android apps. The workarounds it suggests are CSV exports or an API integration that saves copies of your records, which help for reference on site but not for updating a job with no signal.
Does AppSheet's 2,500-row limit apply to a Google Sheet?
No. Google sets that cap on AppSheet's own built-in database: each Starter or Core user gets a fixed number of databases of up to 2,500 rows each, and Enterprise Plus raises it to 200,000. A Google Sheet behind the app is limited by the sheet itself and by sync speed, so archiving finished jobs each year is still a good habit.
Which one is easier to hand to a non-technical office manager?
Airtable, usually. Its grid looks like a spreadsheet, forms are quick to build and a mistake is easy to see. AppSheet has more to configure, because every column has a type and every view is a setting, but it gives engineers a proper phone app. The office manager mostly lives in the sheet behind it.
Can I start on Airtable's free plan and move later?
You can, but the free plan caps a base at 1,000 records, which a busy firm can reach within months. Export to CSV is simple, yet linked records, forms and automations have to be rebuilt by hand on any other platform. Pilot with real jobs on the platform you expect to keep.
Further reads
- Airtable vs Google Sheets for Business Data AI Can Use — If your tracker could stay in a plain sheet.
- Airtable for Startups: Build an AI-Assisted Operations Base — A fuller Airtable operations base with AI fields.
- Turning On-Site Voice Notes Into Job Sheets and Quotes With AI — Turning engineers' voice notes into job sheets.
- Airtable CRM Setup for a Service Business: Tables, Views and Alerts — Tables, views and alerts if the tracker grows into a CRM.
- Glide Mobile App Builder: A Snag List Photo App for a Building Firm — Which Glide to pick, how to build a sheet-backed snag list with camera photos and status filters, how to show clients only their plot, and the plan cost.
- Airtable vs Notion as a Business Database: Which Handles AI Better? — Airtable's AI fields and credits against Notion's agents and Business plan, compared on price, structure and upkeep for a small team.
- AI Tools and AI Development: The Complete 2026 Guide — the AI hub, including every tutorial in the AI-for-business series.
Sources: Airtable pricing page, plans help article and technical-requirements help page; Google AppSheet pricing page, AppSheet database limits page, offline and sync help page, Gemini for App Creation help page and Workspace licensing article; vendor pages read November 2026.