Skip to content
Atami Labs — custom websites, CRM and ERP software

Free tool

Spreadsheets, software
or something
built for you?

Twelve questions about how the business actually runs: people, growth, how many places the same data lives, the re-typing, how standard the process is, who needs it on a phone, the budget. A clear recommendation with every reason shown, the risks of each path, and what to build now versus later.

  • Free, no sign-up. The headline result is always open.
  • Your numbers stay yours. Nothing is converted or guessed behind the scenes.
  • Last updated October 11, 2026.

12 of 12 answered

  1. How many people work in the business?
  2. How fast do you expect to grow over the next two years?
  3. How many systems hold the same customer or job data today?

    Count the spreadsheet, the accounting system, the job app, the email tool.

  4. How many hours a week does someone re-type data from one place into another?
  5. Is the way you run jobs standard for your industry, or particular to you?

    Standard: a vendor could have written it. Particular: it is how you win.

  6. How many hand-offs does a job go through, from enquiry to invoice?
  7. How often does the process itself change?
  8. How many hours a week go into building reports by hand?
  9. Do field, shop or floor staff need to use it, on a phone?
  10. What does it need to connect to?
  11. What could you put into this in the first year?

    In your own currency; software subscriptions and a build both count.

  12. How do you feel about a vendor holding your data and setting the price each year?

Used only for the “what it costs you” line.

Currency

Changes the labels only. Enter your numbers in your own currency.

Our read

Buy software

Your process is close enough to standard that someone has already built it. Pick the tool that holds the customer and the job in one place and connect it to the books.

The current setup costs about $13,700 a year in re-typing and hand-built reports.

How the columns scored

  • Spreadsheets0 points
  • SaaS17 points
  • Custom5 points

The rules that fired

  • People: Five to fifteen people: shared software stops the “which version is current” problem.
  • Growth: Steady growth: per-seat software grows with you without a project.
  • Copies of the data: Three or more copies of the same data: no single tool fits; a system of record is the fix.
  • Re-typing: Two to eight hours a week of re-typing: connected software usually removes most of it.
  • How standard: Mostly standard: buy, and live with the twists or automate around them.
  • Hand-offs: Three to five hand-offs: a workflow tool keeps jobs from falling between people.
  • Rate of change: Yearly changes: either path works if someone owns the configuration.
  • Manual reporting: A few hours of reporting a week: software with dashboards pays for itself.
  • Field and shop staff: The whole crew on phones every day: it has to be built for the field, bought or custom.
  • Integrations: One integration: packaged software almost always has it.
  • Budget: $5,000 to $20,000: a year of good software, or a small integration.
  • Owning the data: Prefer to own it: a point for custom, not a decision on its own.

Build or buy now

  • One record for customers and jobs: pick the system of record and move the data once.
  • Connect what you keep so nobody re-types: jobs to the books, customers to the email tool.
  • Whatever the crew uses on a phone every day: that is the first screen to get right.

Later

  • Reports that build themselves from the record, instead of from a Friday spreadsheet.

The risk of each path

  • Spreadsheets: The risk: it works until the day it doesn’t. Copies multiply, one person understands the formulas, and a hire or a growth spurt turns it into the bottleneck.
  • SaaS: The risk: the software decides how you work. The twists that make you better get flattened, the price rises every year, and the data lives on someone else’s terms.
  • Custom: The risk: a build is a project, and projects need an owner, a scope and a budget. Built badly, it is a spreadsheet with a login. Built well, it is the business.

Not sure which service this points to? The free diagnostic maps it for you.

What to fix first

In order of money

Each line is a year’s money it addresses, from your answers and your hourly cost. The first one is where to start.

  1. Stop the re-typing

    $9,100 a year

    Every hour spent moving data between tools is an hour the business pays for twice. Sync jobs to the books and customers to the email tool, or move them into one system.

    Automation & QuickBooks

  2. One record for customers and jobs

    $6,850 a year

    When the same data lives in three places, nobody trusts any of them. A system of record ends the reconciling and is the foundation the other fixes sit on.

    CRM & ERP

  3. Reports that build themselves

    $4,550 a year

    A Friday spreadsheet is a sign the data is in the wrong shape. Put the numbers where the work happens and the report is a view, not a job.

    Automation & QuickBooks

What it does

What it tells you

Every growing business hits the day the spreadsheet stops being enough: two people editing the same file, a customer typed in three places, a Friday afternoon lost to a report. The question that follows (“buy something, or build our own?”) usually gets answered by whoever shouts loudest or by the first vendor demo. This tool answers it from how the business actually runs, with every reason written out.

Twelve questions, each scoring points to three columns: stay on spreadsheets, buy software, build your own. The read is the leading column, with three override rules (a tie, a close call, a small budget) that are also on this page. You see the columns, every rule that fired, the risk of each path, and what to build or buy now versus later. The headline is never gated; an email unlocks the full scoring table. It never calls an AI model: the scoring is arithmetic you can check.

Who it is for

Who it is for

Owners and office managers of businesses between about five and fifty people who run on spreadsheets, a job app and the accounting system, and feel the seams. Trades and home services first, where the job is the unit of everything, and just as much clinics, distributors, agencies and small manufacturers.

It is not a software comparison, and it does not pick a product. It tells you which kind of answer fits, so the next conversation (with a vendor or with us) starts from the right question. The free diagnostic is where that conversation happens; this tool links to it and does not repeat it.

Method

How it scores

Each answer adds points to the three columns. The leading column is the read, unless an override rule applies:

  1. A tie with spreadsheets goes to software. Spreadsheets only win outright.
  2. Custom leads but the first-year budget is under $20,000: buy now, build later.
  3. Custom and SaaS within 2 points of each other (either way): buy now, build later, with the build planned for the parts the software fights.

The “what it costs you” line uses the middle of the re-typing and reporting bands you picked (under-2: 1 h, 2-8: 5 h, over-8: 10 h; under-1: 0.5 h, 1-4: 2.5 h, over-4: 6 h) times your hourly cost, times 52. The three fixes are the re-typing cost, the reporting cost and, when the data lives in two or three places, half or a quarter of both for a system of record, ranked by money.

Assumptions

The rules and assumptions

Every rule, with its points for spreadsheets / SaaS / custom. These are our judgement as operators, not a statistic; disagree with one and the result shows what changes.

QuestionAnswerSpreadsheetsSaaSCustomRule
People1 to 42Up to four people: a spreadsheet everyone can see is usually enough.
5 to 152Five to fifteen people: shared software stops the “which version is current” problem.
16 to 5021Sixteen to fifty people: software is a given; the question is whose.
More than 5012Over fifty people: the process is the business, and a system built around it starts to pay.
GrowthAbout the same size1No growth planned: don’t buy for a size you won’t be.
Steadily, a few people a year1Steady growth: per-seat software grows with you without a project.
Fast, doubling or more11Fast growth: whatever you pick has to survive twice the volume; spreadsheets rarely do.
Copies of the dataOne1One copy of the data: nothing to reconcile yet.
Two1Two copies: one tool that holds both, or a sync between them.
Three or more12Three or more copies of the same data: no single tool fits; a system of record is the fix.
Re-typingUnder 2 hours1Little re-typing: the current setup is not costing much in labour.
2 to 8 hours2Two to eight hours a week of re-typing: connected software usually removes most of it.
More than 8 hours12Over eight hours a week: that is a part-time job spent moving data; it pays for a system.
How standardStandard, most competitors do it the same way3A standard process: someone has already built software for it; buy it.
Mostly standard, a few of our own twists2Mostly standard: buy, and live with the twists or automate around them.
Particular to us3A process particular to you: off-the-shelf software will fight it; build around it.
Hand-offs1 or 21One or two hand-offs: a list is enough to track.
3 to 51Three to five hand-offs: a workflow tool keeps jobs from falling between people.
6 or more2Six or more hand-offs: the workflow is the product; it deserves its own system.
Rate of changeRarely1A stable process: a fixed tool is fine.
About once a year11Yearly changes: either path works if someone owns the configuration.
Every few months11Frequent changes: spreadsheets bend; so does a system you control. Packaged software does not.
Manual reportingUnder an hour1Little manual reporting: not a reason to change.
1 to 4 hours1A few hours of reporting a week: software with dashboards pays for itself.
More than 4 hours12Over four hours a week of reporting by hand: the data is in the wrong shape; a system that reports itself is the fix.
Field and shop staffNo, office only1Office only: no mobile requirement.
A few people, occasionally1A few mobile users: most packaged apps cover it.
Most of the crew, every day21The whole crew on phones every day: it has to be built for the field, bought or custom.
IntegrationsNothing1No integrations: a spreadsheet is not yet a liability.
One thing (usually accounting)2One integration: packaged software almost always has it.
Several: accounting, e-commerce, payments, email12Several integrations: the glue between tools is where custom earns its keep.
BudgetUnder $5,00031Under $5,000: a build is out of reach; a tidy spreadsheet or an entry-level tool.
$5,000 to $20,0002$5,000 to $20,000: a year of good software, or a small integration.
$20,000 to $60,00012$20,000 to $60,000: a custom system is in range, if the other answers point there.
More than $60,0003Over $60,000: budget is not the constraint; fit is.
Owning the dataFine, that is what software is2Comfortable with a vendor: SaaS has no hidden cost for you.
I would rather own it1Prefer to own it: a point for custom, not a decision on its own.
I have to own it (contracts, exits, compliance)3Must own the data: that rules out most SaaS for the system of record.
  • Budget bands are in your own currency. Nothing converts; the bands read the same on both sites by design, because the step between them matters more than the exact dollar.
  • Hours per band for the cost line are the middle of each band, listed above. Set the hourly cost yourself.
  • No AI in the scoring. A free-text “describe your process” step that drafts a process map was specified and may come later; it would never score or recommend.
  • Rounding. Figures are rounded so the cost line doesn’t pretend to be exact.

Worked example

A worked example

The answers the questionnaire opens with: a 5 to 15-person trades business growing steadily, a few people a year, with the same customer and job data in three or more places, 2 to 8 hours a week of re-typing, a process that is mostly standard, a few of our own twists, 3 to 5 hand-offs a job, changes about once a year, 1 to 4 hours a week of hand-built reports, most of the crew, every day on phones, needing to connect to one thing (usually accounting), a first-year budget of $5,000 to $20,000, and an owner who would rather own the data. Office time costs $35 an hour.

The columns score Spreadsheets 0, SaaS 17, Custom 5. SaaS leads The read is buy software: Your process is close enough to standard that someone has already built it. Pick the tool that holds the customer and the job in one place and connect it to the books.

The current setup costs about $13,700 a year (5 hours of re-typing and 2.5 of reporting a week). First: one record for customers and jobs: pick the system of record and move the data once. The biggest fix is stop the re-typing, worth $9,100 a year. Change the budget to over $60,000, or the process to “particular to us”, and watch the read move.

Questions

Questions people ask

Will this just tell me to build custom software?

No. The scoring is rule-based and every rule is on the page. A small team with one copy of its data and little re-typing comes out as “stay on spreadsheets”, and custom only wins outright when the process is particular to you, the data lives in several places and the budget is there. A small budget turns a custom lead into “buy now, build later”.

Where do the points come from?

From us, as operators who have run a business on spreadsheets, then on packaged software, then on a system we built. They are judgement, not a statistic, which is why every rule is written out next to the answer so you can disagree with one and see what changes.

What does “SaaS” mean here?

Software you subscribe to and configure rather than build: a job management app, a CRM, an accounting system. “Custom” means a system built around your process, usually on top of standard pieces for the standard parts.

Is “buy now, build later” a cop-out?

It is the most common right answer. Packaged software covers the standard three-quarters of most businesses; the build is for the quarter that is yours, and it is cheaper and safer to build it once the volume and the data are real. The result says which parts are which.

How is this different from the free diagnostic?

This tool answers one question, from your answers alone: spreadsheets, software or custom. The free diagnostic is a conversation with us about your whole operation and which service to start with. The result links to it; it does not repeat it.

Do you store what I enter?

The answers and the result are stored without any name or contact details, so we can see how the tool is used. If you ask for the full report, your email is stored with the time, the page and the consent you gave. See the privacy policy.

Next step

Turn the read into a plan.

A free diagnostic looks at the systems you run today and tells you which move pays first: tidy, buy, connect or build. If the answer is “stay on spreadsheets”, we will say so.

We reply within one business day.

Call (905) 650-0230Free diagnostic