A software budget for a brand-new agency: month one, month six, year two
New agencies overspend in month one and underspend in month twelve. The pattern is predictable: a full stack is bought before the first client, then nothing is upgraded when the first real project arrives and the free tools start failing. A better sequence spends almost nothing until revenue exists, then upgrades in the order that protects delivery.
Month one: $0, deliberately
| Job | Tool | Cost |
|---|---|---|
| Pipeline and contacts | HubSpot Free (2 users) | $0 |
| Client work | ClickUp Free (unlimited members) | $0 |
| Documents and notes | Notion Free | $0 |
| Email and invoicing | Existing Workspace / accounting tool | $0 marginal |
| Total | $0 |
Nothing here is a compromise for a two-person agency. HubSpot's free CRM covers contacts, deals, tasks, tracking and dashboards. ClickUp's free plan has unlimited members and time tracking, which is more than most paid tiers offer. The only real constraint is 60 MB of storage in ClickUp — solved by keeping files in Drive and linking them, a habit worth forming anyway.
What not to buy in month one: a paid project tool, a paid CRM, a website builder subscription, or an email marketing platform. None of them do anything for you until there is a pipeline to manage and an audience to email.
Month six: the first real invoices
By month six you should be hitting two specific walls: a third person (HubSpot's 2-user cap) and billable time you cannot report on. Those are the only two upgrades that pay for themselves at this stage.
- CRM: Pipedrive Lite at €14/seat/month — €42/month at three seats, €504/year. Move when the free CRM's user cap blocks a hire, not before. If you sell in staged deals, this is the right first paid line.
- Or the flat alternative: systeme.io Startup at $17/month — $204/year regardless of seats, and it also covers funnels and email. Weaker CRM, dramatically cheaper at four-plus people.
- Project tool: still ClickUp Free, or Core at $9/seat — Core buys storage headroom, custom fields and recurring automations. At three seats, $27/month. Do not buy monday at this stage; the 3-seat minimum and the 250-action automation cap make Basic a tier you outgrow immediately.
Realistic month-six total: $250–$600/year, depending on which CRM route you take.
Year two: the delivery upgrades
In year two the constraint moves from sales to delivery. This is when time tracking, utilisation and client visibility start to matter — and it is where agencies overspend on the wrong axis.
- Time tracking. ClickUp includes it on Core ($9/seat). monday puts it on Pro at $19/seat — for a billable agency that is the real entry price, not Standard's $12. Compare $135/month against $285 at fifteen seats before deciding that monday's nicer boards are worth it.
- CRM at scale. Pipedrive Growth at €39/seat is €585/month at fifteen seats — by far the largest line. Only justified if deals are forecast and reviewed weekly; otherwise it is the most avoidable cost in the budget.
- Email list. Kit Free covers 10,000 subscribers. Do not pay $33/month for Creator until sequences are actually needed.
The three costs new agencies forget
- Add-ons. Pipedrive's LeadBooster is €32.50/month — at five seats that is a 46% increase on a Lite subscription. Budget one add-on per tool by default.
- Currency. A euro-priced CRM is a foreign-currency exposure on your largest software line. It is small, and it is real.
- Administration. Twenty minutes a week keeping a pipeline honest is 17 hours a year. At $75/hour that exceeds most subscription lines. The tool is not the expensive part.
The sequence in one line
Free until a hire blocks you, paid CRM at the first hire, paid delivery tooling when you start billing hourly, everything else only when a specific failure demands it. The inverse — buying the full stack first — costs $3,000–$5,000 in year one and produces no capability you would have used.
What we cannot verify
All prices come from vendor pricing pages verified October 23, 2026, on annual billing, and we re-check quarterly. The staging reflects how agency constraints typically arrive, not a measured cohort. Our hands-on testing cycle is in progress — we have not run an agency through this budget sequence, so the timings (month one, month six, year two) are structural estimates rather than observed. Findings will be added per our How We Test process. For a full itemised stack at three sizes, see what an agency tech stack really costs.