The martech stack at every ARR band
Five stacks from pre revenue to $100M ARR, each with monthly cost, the tool to buy next, who runs it, and the three tools most teams overbuy early.
On this page 10 sections
- What a martech stack should cost at each ARR band
- The pre revenue stack: $150 a month, run by a founder
- The $0 to $1M stack: first real CRM, first real email tool
- The $1M to $5M stack: the band where overbuying starts
- The $5M to $20M stack: the RevOps hire changes everything
- The $20M to $100M stack: security reviews and a real bill
- What each band costs per customer acquired
- The downgrade path nobody plans for
- What we would buy next at each band
- What to do this week
- Frequently asked questions
The short answer
A SaaS martech stack should track ARR, not ambition. Pre revenue teams run on roughly $150 a month of free and near free tools with the founder operating them. At $1M to $5M ARR the stack lands near $2,500 a month and needs one part time marketing ops owner. At $20M to $100M ARR it runs $25,000 to $60,000 a month with two to four dedicated operators. Buying a tool more than one band ahead adds cost and admin without adding output.
Key points before you start
Most martech advice is written as a shopping list with no price and no operator attached. That is why so many Series A companies own a $30,000 platform nobody logs into. The stack below is organised by ARR band, and every band states three things: what it costs per month, who runs it, and the event that justifies moving up.
Read it as a buying order, not a wish list. If you are at $2M ARR and something in the $20M column looks appealing, that is the tax talking.
What a martech stack should cost at each ARR band
Tool spend scales faster than headcount but slower than revenue. Under $5M ARR you should be at 8 to 12 percent of marketing budget. Above $20M ARR, 20 to 30 percent is normal once warehousing, enrichment and attribution arrive. Our fuller breakdown of what SaaS companies actually spend on marketing tools has the distribution by band.
| ARR band | Monthly tool spend | Annual | Share of marketing budget | Operator |
|---|---|---|---|---|
| Pre revenue | $120 to $250 | $1,500 to $3,000 | n/a | Founder, 2 hrs/week |
| $0 to $1M | $600 to $1,200 | $7,000 to $14,000 | 8 to 10% | First marketer, 4 hrs/week |
| $1M to $5M | $2,000 to $3,500 | $24,000 to $42,000 | 10 to 14% | Marketing ops, 0.5 FTE |
| $5M to $20M | $7,000 to $15,000 | $84,000 to $180,000 | 15 to 22% | Marketing ops, 1 FTE |
| $20M to $100M | $25,000 to $60,000 | $300,000 to $720,000 | 20 to 30% | RevOps team, 2 to 4 FTE |
Source: aggregated practitioner reports, saas-marketing.net estimate. Treat the ranges as a sanity check, not a target.
20 to 30%
Share of marketing budget spent on tools above $20M ARR
Aggregated practitioner reports, saas-marketing.net estimate
The pre revenue stack: $150 a month, run by a founder
At pre revenue the correct answer is almost entirely free tiers. You need a site, a way to collect emails, a way to see what people do, and a place to keep conversations.
Webflow or a Next.js site on Vercel for the marketing pages. PostHog free tier for product and web analytics, which handles the first million events a month without a bill. A Google Workspace inbox. Attio’s free tier or a single Notion database for the CRM, because you have under 200 contacts and nothing to automate yet. Calendly free for demos.
The only line item worth paying for is a domain and a transactional email service, usually Resend or Postmark at $15 to $20 a month, because deliverability from a personal Gmail will burn the domain you have to live with for a decade. Our list of free and near free SaaS marketing tools covers the tiers that are genuinely usable rather than trial bait.
The classic pre revenue mistake
Buying an SEO platform before you have a page worth ranking. Ahrefs and Semrush both start near $129 a month. At pre revenue you can answer every keyword question you actually have with Google Search Console, free Ahrefs Webmaster Tools, and an afternoon.
The $0 to $1M stack: first real CRM, first real email tool
The trigger for this band is the first repeatable inbound flow. Once more than five people a week hit a form, spreadsheets stop tracking them.
Buy a CRM with email sequencing. Attio at $34 per seat per month or Pipedrive at $39 is the right scale. Add an email marketing tool: Loops, Customer.io’s starter tier or Mailchimp, at $50 to $200 a month depending on list size. Keep PostHog. Add Ahrefs or Semrush at $129 to $149 if content is a real channel, and skip it if it is not.
Total lands near $600 to $1,200 a month. One marketer runs all of it in about four hours a week. The martech stack audit template is worth running once here, mostly to catch the trial subscriptions nobody cancelled.
Editable working copy
Download this template
Save an editable working copy of the framework on this page. Add your own owners, evidence and decisions.
The $1M to $5M stack: the band where overbuying starts
Something changes psychologically after a Series A. There is budget, there are vendor calls, and nobody has yet been burned by a platform migration. This is the band where teams buy two years ahead.
The honest stack here: HubSpot Starter or Professional ($800 to $1,600 a month with contacts), Customer.io for lifecycle ($150 to $500), PostHog or Mixpanel for product analytics ($200 to $600), Ahrefs or Semrush ($200 to $450 across seats), Webflow ($50), Zapier ($70 to $300), and a scheduling and routing layer like Chili Piper once inbound demos pass about 40 a month ($30 per seat).
Around $2,000 to $3,500 a month, operated by someone who spends half their week on it. That half person is the real constraint. A stack without an owner degrades in about one quarter: fields stop being filled, automations silently break, and reports stop matching.
Three tools most teams buy one band too early
An ABM platform like 6sense or Demandbase, at $30,000 to $100,000 a year, bought before a named account list exists. A CDP like Segment, bought before there are two systems worth joining. A multi touch attribution tool, bought at a volume where no model is statistically stable. Each needs an operator you have not hired.
The $5M to $20M stack: the RevOps hire changes everything
The trigger here is not revenue, it is the first dedicated RevOps or marketing ops hire. Before that person exists, every tool you add is overhead on someone doing three other jobs. After they exist, the stack can carry real complexity.
What gets added: a warehouse (BigQuery or Snowflake, $500 to $3,000 a month), a reverse ETL or CDP layer (Segment or Hightouch, $1,000 to $3,000), enrichment (Clearbit or Clay, $800 to $2,500), sales engagement (Outreach or Salesloft, $100 to $140 per rep), and conversation intelligence (Gong, at roughly $1,400 per seat per year plus platform fee).
What gets swapped: PostHog to Amplitude when three or more teams need self serve analytics and the query load stops being a marketing question. Mailchimp to Customer.io or Braze when lifecycle depends on product events rather than list membership. Pipedrive to HubSpot when the handoff between marketing and sales needs a shared object model.
| Transition | Swap | Trigger event | Migration cost |
|---|---|---|---|
| $1M to $5M | PostHog to Amplitude | Three teams querying independently | 2 to 4 weeks |
| $1M to $5M | Mailchimp to Customer.io | Lifecycle keyed on product events | 3 to 6 weeks |
| $5M to $20M | Pipedrive or Attio to HubSpot | First sales manager, shared pipeline | 4 to 8 weeks |
| $20M+ | HubSpot to Salesforce | Second product line or 25+ reps | 6 to 12 weeks |
| $20M+ | Spreadsheet reporting to warehouse | Board asks a question CRM cannot answer | 8 to 16 weeks |
Do at most one of these per band. A CRM migration costs six to twelve weeks of ops time and always lands on the quarter you needed that person for something else.
The $20M to $100M stack: security reviews and a real bill
Two events define this band. The first enterprise security review, which forces SOC 2 and a vendor inventory (Vanta or Drata, $12,000 to $30,000 a year). The second is the first multi product line, which breaks every report built on a single funnel assumption.
The stack reaches $25,000 to $60,000 a month: Salesforce or HubSpot Enterprise, Marketo or HubSpot Marketing Enterprise, Snowflake, Segment, Gong, Outreach, 6sense or Demandbase, an attribution layer, plus a content and SEO suite. Two to four people run it, and at least one of them writes SQL daily. The eight SaaS marketing stacks we tore down all sit in this band and none of them look identical, which is the point.
Editable CSV worksheet
SaaS benchmark evaluation worksheet
Record the source, date, cohort and metric definition before comparing your numbers with a benchmark.
What each band costs per customer acquired
The number that makes the argument to a CFO is tool cost per new customer, not tool cost per month. It falls sharply with scale, which is the honest case for spending more later and less now.
| ARR band | Monthly tool spend | New customers/month | Tool cost per customer |
|---|---|---|---|
| $0 to $1M | $900 | 12 | $75 |
| $1M to $5M | $2,800 | 45 | $62 |
| $5M to $20M | $11,000 | 130 | $85 |
| $20M to $100M | $40,000 | 400 | $100 |
saas-marketing.net model, method shown on the page: monthly spend divided by new logos. Note the curve is not monotonic. Cost per customer usually rises at the $5M to $20M transition because the warehouse, CDP and enrichment layers all land before the volume that justifies them. That bump is normal, and it is also where the martech cost per customer calculator earns its keep, because you can see whether your bump is six months or two years long.
The downgrade path nobody plans for
Every stack is built during a good quarter and paid for during a bad one. Plan the reverse order before you need it.
Cutting the stack without breaking the go to market
- Inventory with renewal dates
List every tool, seat count, monthly cost and renewal date. You will find two tools nobody has opened in 60 days. That is the test: last login, not last invoice.
- Cut enrichment credits first
Clay, Clearbit and Apollo credits are usage based and reversible. Halving them costs you list coverage, not systems. Expect $800 to $2,500 a month back within one cycle.
- Collapse seats before tools
Move from ten Gong seats to four, from eight Ahrefs seats to two. Seat reduction preserves the integration and the history, which a cancellation does not.
- Drop the ABM platform
It is the largest single line above $20M ARR and the hardest to attribute. If pipeline sourced from it is under 15 percent, it goes. Give 90 days notice.
- Replace the CDP with native integrations
Most CRM to product analytics syncs exist natively now. You lose flexibility and gain $1,000 to $3,000 a month. Budget four weeks of rebuild.
- Never cut the CRM, email or product analytics
These hold history you cannot rebuild. Downgrade the tier instead. You can tell it worked when reporting still matches last quarter's numbers after the cuts.
The failure mode is cutting a tool whose data lives nowhere else. Export before you cancel, always, and check that the export includes activity history and not just records.
What we would buy next at each band
Straight answer per band, no hedging. Pre revenue: nothing, spend the money on a writer. $0 to $1M: the CRM, before anything else. $1M to $5M: lifecycle email keyed on product events, because it is the only tool in that band that changes revenue rather than reporting. $5M to $20M: the warehouse, but only after the ops hire. $20M to $100M: conversation intelligence, because at that headcount the cheapest research you can buy is a searchable archive of your own sales calls.
The pattern in the full SaaS marketing stack overview holds here: buy for the constraint you have this quarter. If you cannot name the report that is missing or the workflow that is manual, you are buying a feeling.
The one exception to buying ahead
AI tooling is the current exception, because seat prices are low and the learning curve is the real cost. A $20 to $60 per seat spend on writing, research and transcription tools pays back at almost any band. The rules for picking them are in our guide to AI marketing tools for SaaS teams.
What to do this week
Pull the last three months of card statements and list every martech charge against last login date. Map your current stack to the band above that matches your ARR, and mark anything that belongs to a higher band. Then pick one thing to cancel and one thing to buy, and write down the trigger event for the next purchase so the decision is made before the vendor call.
If you want the vocabulary straight before the internal debate, the martech stack definition and the B2B SaaS marketing tech stack reference are short and will settle most arguments about what counts as core versus optional.
Editable CSV worksheet
SaaS Marketing Tools planning worksheet
A practical tools planning worksheet: decisions, owners, evidence and next actions.
Frequently asked questions
How much should a SaaS company spend on marketing technology?
Under $5M ARR, budget 8 to 12 percent of total marketing spend on tools. Above $20M ARR, expect 20 to 30 percent as data warehousing, enrichment, ABM platforms and attribution enter the stack. In absolute terms that is roughly $150 a month pre revenue, $2,500 a month at $3M ARR, and $25,000 to $60,000 a month at $50M ARR.
What is the first paid marketing tool a SaaS startup should buy?
A CRM with email, usually Attio or Pipedrive at $30 to $60 per seat per month. Analytics and a site builder both have credible free tiers, but the record of who talked to whom does not survive in spreadsheets past about 200 contacts. Buy the CRM before the SEO tool, the scheduling tool, or anything with the word platform in its name.
When should a SaaS company move from HubSpot to Salesforce?
When you have a dedicated RevOps person and either a second product line or a sales team above roughly 25 quota carrying reps. Salesforce needs an administrator to be useful. Without one you pay enterprise licence costs for a worse version of what HubSpot gave you. Many $50M ARR companies stay on HubSpot Enterprise on purpose.
Which martech tools do early stage SaaS teams overbuy most?
Three repeat offenders: an ABM platform bought before the account list exists, a customer data platform bought before there are two systems worth joining, and an attribution tool bought before there is enough volume for any model to be stable. Each costs $2,000 a month or more and each needs an operator you have not hired.
Do you need a data warehouse for SaaS marketing?
Not until roughly $10M ARR or the point where product usage, billing and CRM data have to be joined for routing or reporting. Below that, the native integrations between your CRM, product analytics and billing system cover the questions you are actually asking. A warehouse without an analyst is a storage bill.
What happens to the martech stack when funding tightens?
Cut in this order: enrichment credits, second seats on specialist tools, the ABM platform, then the CDP. Keep the CRM, the email platform, the site and the product analytics. The practical target is to reach the previous ARR band's stack within one renewal cycle, which usually means 90 to 120 days of notice periods to manage.
The saas-marketing.net editorial team Research and editorial
We research, write and maintain every page on this site. The library explains marketing decisions through practical frameworks, explicit assumptions and references. Corrections can be requested through the contact page.
Published September 11, 2026. Last updated .