44% of SaaS licenses go unused in a typical month. Companies waste 25-30% of their software spend on licenses nobody opens. For a 10-person startup spending $5,000/month on tools, that's $1,500 a month walking out the door. This framework recovers most of it in 90 minutes.

The bad news: most founders don't know what they're paying for. Spend lives on 3 different credit cards, subscriptions auto-renew in email folders nobody reads, and new tools get added faster than old ones get cancelled. "We should audit the stack" has been on your backlog for 8 months.

The good news: an audit isn't hard. It's tedious, and 90 minutes of discipline will pay back more than most things you'll do this quarter. Here's the framework that actually works.

30%
The typical amount of SaaS spend that can be cut in a single audit, without touching anything the team actually uses.

Before you start: the mindset

The biggest mistake is treating the audit as finance work. It isn't. It's a stack-hygiene exercise that happens to save money.

The goal isn't to cut costs. The goal is to re-align your stack with your operation. Cost savings are a byproduct. If you frame it as "how do I get cheaper?" you'll make penny-wise-pound-foolish trades. Frame it as "what is each of these tools actually doing, and do I still need it?" and the savings come out clean.

Block 90 minutes on the calendar. Close Slack. One pass, no interruptions.

The 90-minute framework

1
~15 minutes

Inventory every tool

Open all three of these and make one flat list:

  • Your billing dashboards - Stripe, the company credit card, any personal cards reimbursed for tools.
  • Your password manager - 1Password, Bitwarden, LastPass. Every tool you've logged into in the last 12 months is here.
  • Your email search - search "receipt" and "invoice" for the last 12 months. You'll find subscriptions your password manager missed.

Put everything in a spreadsheet with columns: Tool | Monthly cost | Billed to | Primary user. Don't rank or judge yet. Just list.

Most startups find 30-50% more tools than they thought they had. The first honest surprise is always the size of the list.

2
~15 minutes

Tag by actual usage

Add one column: Last used. For each tool, quickly answer - not when you last logged in, but when anyone on your team actually did real work in it.

Tag each tool with one of four labels:

  • DAILY - someone uses this every working day.
  • WEEKLY - used, but not daily.
  • DORMANT - logged in at some point this year but not this month.
  • DEAD - nobody has opened it in 3+ months.

This alone exposes the easiest wins. Every DEAD tool is a guaranteed cancel. Every DORMANT tool is a candidate.

3
~15 minutes

Identify overlap

Group the list by what each tool actually does. Write the job, not the brand:

Email marketing:    Mailchimp, ConvertKit, Beehiiv
Analytics:          GA4, Plausible, PostHog, Mixpanel
Docs:               Notion, Coda, Google Docs, Confluence
Project tracking:   Linear, Trello, Notion, ClickUp
Automation:         Zapier, Make, n8n
Chat / messaging:   Slack, Discord, Microsoft Teams

Categories with more than one tool are red flags. You're either paying for overlap, or you have a silo problem where different teams bought different tools for the same job. Neither is good.

Mark the canonical tool in each category. The others are candidates for consolidation.

4
~20 minutes

Make the kill / keep / consolidate decision

For each tool, answer five questions. Be ruthless. Be fast.

  1. What's the job-to-be-done? State it in one sentence. If you can't, it's a candidate to kill.
  2. Who uses it and how often? "Everyone, weekly" is a different answer from "one person, sporadically."
  3. What happens if we cancel it tomorrow? If "nothing urgent," cancel it.
  4. Is there another tool we pay for that could do this job at 80%? 80% is the threshold. Don't hold out for perfect.
  5. What's the true cost? License + admin time + integration maintenance + cognitive load of another login. Always 3-5x the subscription fee.

Tag each tool KEEP, KILL, or CONSOLIDATE. A healthy audit kills 20-30% of the list and consolidates another 10-15%.

5
~15 minutes

Document and schedule the cutover

Write decisions into a doc (Notion page or Google Doc - one place). For each KILL or CONSOLIDATE:

  • Owner - the person who will actually cancel it.
  • Cutover date - realistic, usually 1-2 weeks.
  • Replacement (if consolidating) - which kept tool now does this job.
  • Data export - does anything need to be saved before canceling?

Add a single line in the decision log: "Why we killed this." Your future self will try to resubscribe in 6 months - the note will stop you.

6
~10 minutes

Set the quarterly recurring review

Add this audit as a recurring 90-minute block in your calendar, every 90 days. Q1, Q2, Q3, Q4. Without the recurring block, you'll end up back here in 18 months with the same mess.

Tie it to a specific trigger: "first Monday of the quarter" or "day after board reporting." Triggers beat intent.

What a "good" audit result looks like

For a typical early-stage startup (2-10 people), a first-time audit usually finds:

  • 3-5 tools to kill outright (DEAD, unused, or cheaper free alternatives exist).
  • 2-3 tools to consolidate into a tool you already pay for.
  • 1-2 tools to downgrade from Pro to Free, or from Team to Solo.
  • 1 renegotiation - the biggest annual contract, negotiated at renewal.

Realistic savings: 25-35% of monthly SaaS bill, within 2 weeks of the audit. For a company spending $3,000/month on software, that's $9,000-12,000 a year - usually equivalent to one additional month of runway.

The cancellation mechanics

The fastest way to actually cancel is to batch the work, not spread it across weeks. Block 60 minutes a day after the audit and just do it. For each KILL:

  1. Export anything worth saving - data, templates, integrations. Archive to Drive or S3.
  2. Cancel in-app - most tools let you cancel without contacting support. The exceptions are always the most annoying ones to use.
  3. Confirm the cancellation email - if you don't get one, the cancellation didn't take. Many vendors use dark patterns to keep you subscribed past the date you thought you cancelled.
  4. Reset the password to a random string you don't store. Prevents accidental resubscription.
  5. Delete from password manager. Clean slate.
Every tool you cancel is a tax you stop paying, a context you stop switching into, and a subscription you stop forgetting about.

Common mistakes

Auditing too often. Quarterly is right. Monthly creates audit fatigue and doesn't give you enough data between audits to decide.

Letting one person own the "keep" decision. If one team member says "we need it" the tool stays. The right filter is "who will own the work this tool replaces?" If nobody volunteers, that's a kill.

Downgrading without killing. "Let's just move to the cheaper tier" is usually a half-measure. If a tool is on the chopping block, ask whether you need it at all before moving to Free.

Skipping the quarterly cadence. Tool sprawl is gradual. Each individual signup felt rational. Without the recurring check, you'll be back to a bloated stack within 12 months.

The audit template

Copy this spreadsheet structure and use it directly:

| Tool | Monthly cost | Billed to | Primary user | Last used | Category | Decision | Replacement | Owner | Cutover date |
|------|--------------|-----------|--------------|-----------|----------|----------|-------------|-------|--------------|
| ...  | ...          | ...       | ...          | ...       | ...      | ...      | ...         | ...   | ...          |

Fill it in order. Don't over-design. A working audit in 90 minutes beats a beautiful audit in two weeks.

Want us to run the audit for you?

We'll inventory every tool, find the 30% you can cancel this week, and rebuild what's left into a stack that actually holds. Typically saves 10x the engagement fee in the first year.

Get a Stack Audit →

Further reading