How to Audit Your SaaS Stack Before You Rebuild

This guide is for owners about to rebuild or replace part of their software stack. It walks through five steps: list every tool, find its real cost, map where the data flows, score each tool keep, consolidate, cut or rebuild, and sketch the system you want, so you only rebuild what is worth rebuilding.

The printable checklist and the scoring rules live in this guide, because they are steps four and five, not separate methods. Run them in order whether you rebuild with us or on your own.

01 · Why audit first

The problem you can’t see is the one you’re paying for

Most teams underestimate their stack by a wide margin. Not because they’re careless, but because SaaS accumulates quietly: a tool here for a campaign, a subscription there a contractor set up, a “free” integration that quietly moved to a paid tier. The result is a portfolio nobody fully owns.

The numbers back this up. The average company now runs 106 SaaS applications, and roughly half of all software licenses go unused: the highest waste rate ever recorded. Somewhere between a quarter and a half of the average software budget is spent on tools that overlap, duplicate each other, or sit dormant.

There’s a second reason to audit now specifically: the economics of building have flipped. In Retool’s 2026 Build vs. Buy report, 35% of enterprises had already replaced at least one SaaS tool with custom software, and 78% expected to build more in 2026. Custom is no longer the expensive, slow option it was five years ago. But “we could just build it” is a trap without an audit. You have to know what you’re replacing before you can replace it well.

02 · The framework

Five steps, in order

The sequence matters. Each step feeds the next, and skipping ahead is how audits produce a spreadsheet nobody acts on. Inventory before cost, cost before data flows, data flows before you dare score anything for the chopping block.

Step 01

Inventory everything, including what’s hiding

You cannot rebuild what you can’t see. Start by listing every tool that touches your business, not just the ones in your bookmarks bar. The fastest way to surface the full picture is to pull from where the money and the logins actually live:

• Card and bank statements for the last 12 months, recurring charges are the truth serum. Annual renewals hide from monthly reviews, so look across a full year.
• Your SSO / Google Workspace app list: every third-party app anyone has authorized.
• Browser password managers and shared vaults: logins reveal tools that never hit a company card.
• A five-minute survey to the team: “What do you open to do your job?” Shadow IT accounts for a third of most stacks, and this is the only way to find it.

For each tool, capture the same columns: name, owner (a real person), category, what it’s used for in one sentence, number of seats, billing cadence, and renewal date. Resist the urge to analyze yet. Right now you’re just building the map.

Step 02

Follow the money, the real number, not the sticker

Now attach cost to every row, and be honest about the full cost. The subscription line is only the visible part. The real monthly cost of a tool includes per-seat fees that scale as you hire, usage tiers that jump when you cross a threshold, mandatory onboarding fees, paid add-ons that quietly became load-bearing, and the contractor or agency retainer that exists only to keep the tool running.

Total it up per tool, then per category. This is the step that produces the “wait, how much?” moment, the one that gets a rebuild approved. When you find two tools doing 80% of the same job, or a plan tier you outgrew the reason for, write it down. You’ll score it in Step 4.

Step 03

Map the data flows and integrations

This is the step most audits skip, and it’s the one that predicts whether your rebuild succeeds. A stack isn’t a list of tools. It’s a web of data moving between them. Your form fills a CRM, which triggers an email tool, which syncs back a status, which a spreadsheet reads, which a dashboard charts. Every arrow in that diagram is a place things silently break.

Draw it. For each tool, note what data comes in, what goes out, and how the connection is held together: native integration, a paid automation tool, a webhook a developer wrote once and left, or a human copy-pasting between tabs. Mark anything held together by a single automation tool or a single person; those are your fragile seams. The hours that person spends copying across the seams are middleware labor, and that argument is made in full in You’re paying for middleware. When you rebuild, these connections are the actual product. The tools are just endpoints.

The tools on your invoice are the nouns. The integrations between them are the verbs, and the verbs are where your team actually loses its hours.

Step 04

Score every tool: keep, consolidate, cut, or rebuild

With inventory, cost, and data flows in hand, you can finally make decisions. Score each tool on two axes: how much real value it delivers (not how much you pay: how much you’d miss it), and how much control and ownership you have over the tool, its data, and its cost. Those two axes give you four quadrants and four clear verbs.

The rebuild quadrant is the one to sit with. These are the high-value tools you depend on but don’t own: the marketing platform whose bill climbs with your contact count, the automation layer holding your business together, the site builder you can’t leave without losing your content. These are where owning the system pays off the most, because the value is high enough to justify the build and the lack of control is what’s costing you.

Rebuild does not mean switch to a cheaper subscription. It means replace a rented capability with software you own, so the cost does not scale against you and canceling a vendor does not take the business offline.

Ask two questions for each tool, independently. How much would the business miss it tomorrow, and how much control do you have over its cost, schema, and data? High miss and low control is rebuild. Low miss and any ongoing cost is cut. High miss and high control is keep.

Then act in this order. Cut true zombies first, the seats and automations that free budget without a migration. Consolidate overlaps into the tool you are keeping, before you rebuild, so the data is cleaner. Sequence the rebuilds by dependency, usually the site, then the CRM, then email on the same record. Re-run the Stack Sprawl Index after each major cut. A score that does not move means something marked keep should have been cut.

A few calls warp the score. Scoring before the data-flow map is the expensive one: a tool that looks redundant may be feeding three automations. Sticker price is the other. A $25/month builder with a $2,000/month agency keeping it alive is not a $25 tool. High spend is not automatically rebuild, and fear of change is not automatically keep. A tool unused for 90 days is a cut candidate regardless of what it cost to start.

Step 05

Design the target system before you touch a tool

The final step of the audit is a one-page sketch of where you’re going. Not a spec (a direction. It answers: what does the consolidated stack look like? Which tools survive as “keep”? What gets folded together? And critically, for everything in the “rebuild” pile) what would it mean to own that instead of rent it?

This is where the audit stops being a cost-cutting exercise and becomes a design brief. The goal isn’t the smallest possible number of tools; it’s a system where you control your code, your content, and your data, where the seams between tools stop being fragile, and where your cost doesn’t climb every time you add a contact or a seat.

Checklist

The printable pass

Use this beside the interactive scorecard. Finish each block before the next.

Inventory. Twelve months of card and bank recurring charges. SSO or Workspace third-party apps. Password vault and shared logins. Agency and freelancer tools running on your behalf.

Cost. Sticker, seats, and meters per tool. Onboarding fees amortized. Weekly admin hours estimated. Totals entered in the stack cost calculator.

Data flows. The top five lead and customer paths drawn. An owner named for each integration. A source of truth marked for contacts, deals, and content.

Score. Keep, consolidate, cut, or rebuild on every row, using the table above.

Target. A one-page sketch of site to CRM to email, plus the migration risks: SEO, open deals, and live lists.

Pitfalls

Four ways audits go wrong

Having run this process across dozens of stacks, the failure modes are predictable:

• Auditing tools but not integrations. You cancel a “redundant” app and discover it was quietly feeding three others. Always finish Step 3 before Step 4.
• Using sticker price instead of loaded cost. The $25/month tool with a $2,000/month agency keeping it alive is not a $25/month tool.
• Cutting for savings, then stopping. Cancelling dead weight feels great and banks quick wins, but if you never address the “rebuild” quadrant, the sprawl regrows within a year.
• Rebuilding the mess faithfully. A migration that recreates every quirk of the old stack just relocates the problem. The audit is your chance to decide what not to carry over.

Conclusion

The audit is the rebuild

It’s tempting to treat the audit as paperwork you push through to get to the fun part. But the audit is the strategy. It’s where you decide what your business actually depends on, what it’s overpaying for, and what it’s worth owning outright. A rebuild without it is a gamble; a rebuild guided by it is just execution.

So before you migrate a single tool, spend the week. Pull the statements, draw the data flows, score the quadrants, sketch the target. Whether you build it yourself or bring in a partner, you’ll be making decisions from a map instead of a memory, and that’s the whole difference between a stack you rent and a system you own.