The minimal software stack for a small team
Every small team's software stack grows the same way: a tool gets adopted for one job, does it fine, and stays forever. Nobody subtracts. Five years later the stack is ten subscriptions deep, half of them overlapping, and the monthly total is a number nobody has written down. The minimal stack is not a shopping list — it is a subtraction exercise with a floor.
The four jobs every team actually has
Strip away category names and every service business runs on four functions: winning work (pipeline, proposals), doing work (tasks, files, coordination), getting paid (invoicing, payments), and staying in touch (email, scheduling). Anything outside these four is a candidate for deletion, not addition.
The reason stacks bloat is that vendors sell by category, and categories multiply. "CRM" and "project management" and "workspace" and "docs" are four products for what a three-person team experiences as one continuous flow of work. The stack is minimal when each job has exactly one owner — and when one tool can own two jobs without straining, it should.
The overlaps that do the subtracting
- Bookings ride on the calendar. A scheduling tool's job is a public page in front of your calendar. A solo consultant's scheduling need is real but small — our breakdown of what a scheduling tool actually needs to do is in Calendly pricing explained — and it should be justified by volume of bookings, not by category completeness.
- Money rides on anything that holds a record. Invoicing is the function you cannot skimp on — it is regulated, it touches tax, and the processing fee economics are their own subject (where the fee on an $800 invoice goes). But client tracking at three people does not need its own CRM subscription; a sheet or a light workspace holds the pipeline until deal flow earns the upgrade. The graduation path is in spreadsheet to CRM.
- Doing work absorbs "workspace" and "docs". One project tool with good views beats two tools with overlapping ones; the cost of a second tool is not its price, it is the split attention — the tax we quantified in the hidden costs of cheap software.
- Email needs one ruler, chosen deliberately. A newsletter to clients and a newsletter to subscribers are different products priced by different meters; before adding an email subscription, decide which ruler you are buying — the three rulers of email pricing.
What the minimum looks like, in money
Built from verified prices across our coverage, a three-person team's floor looks like this:
- $0 tier: scheduling on a free plan (one event type — the real constraint, per Calendly's free tier), invoicing on Wave Starter (free; you pay only the card fee), project tracking on ClickUp's free plan, client list on Kit's free tier to 10,000 subscribers. Total subscription: $0. Cost is entirely variable — the card fee on money collected.
- ~$60–90/month tier: the first paid steps — a real CRM (Pipedrive's cheapest tier per seat, noted with its seat rules in our Pipedrive breakdown), or a project tool's first paid tier when the free plan's structure runs out (ClickUp pricing explained).
- ~$150–250/month tier: the point where a five-person team adds automation capacity and admin seats — and where per-seat minimums start to matter more than list prices, per seat minimums nobody budgets for.
The trajectory matters more than any row: growth should push you up within a tool (the vendor's own ladder) before it pushes you across tools (the stack). Every cross-tool addition multiplies integration, billing, and exit surfaces.
The two rules that keep a stack minimal
- One job, one owner. When a second tool claims a job the first already covers, the question is never "which is better" but "which one is leaving". Run that decision with the four-hour method in how to compare two tools in an afternoon.
- A tool must pay for its own complexity. Each subscription costs more than its invoice — an account, a renewal, an export, a seat audit. If a $15 tool saves less than an hour a month, it is priced wrong for you. The audit that finds these is in the one-hour software spend audit.
What we cannot verify
The plan references above link to our own breakdowns, each of which carries its verification date and caveats; none of them constitutes a recommendation, and the tier amounts are illustrative bands built from verified list prices — not quotes for your situation. Regional pricing, seat minimums and promotions change all of them; the cart wins. The four-jobs framework is our synthesis, not a standard. See How We Test.