SaaS Triage: A 4-Step Playbook for Deciding What to Cancel and What to Build Yourself
What's Changed
Small businesses now have access to the same infrastructure that powers the SaaS tools they're paying for. CRMs handle automation. Cron jobs handle monitoring. APIs handle data flow. The $20-200/month tools that filled gaps five years ago are increasingly replaceable with infrastructure the business already owns.
The playbook isn't "cancel everything." It's a triage framework for deciding which SaaS is backbone and which is fat.
Why This Matters
Every $45/month SaaS you cancel is $540/year back in your operating budget. Three cancellations fund a meaningful API credit pool. Five fund a part-time operator. The money is already in your business — it's just routed to other people's platforms.
The Playbook: 4 Steps
Step 1 — Audit one SaaS.
Pick the smallest one. Not your CRM. Not your email. The one that costs $20-50/month and does something you barely think about.
Ask: what does it actually do for the business right now? Not what the landing page promises. What functions does it perform?
Write these down. Be specific. "Hosts 7 pages. Processes Stripe payments. Sends 3 onboarding emails. Shows me a morning summary."
Step 2 — Map to what you already own.
You probably already have a CRM, a website host, a payment processor, a cron or automation tool. For each function from Step 1, ask: can something I already pay for absorb this?
Don't ask "can I buy something better?" Ask "can something I already have do this?"
Step 3 — Migrate the live pieces only.
Most SaaS tools have features you never use. Dead pages. Abandoned workflows. Settings you set once and forgot. Don't rebuild these.
Migrate only what's live. The homepage. The checkout. The email sequence. The form. If it wasn't active last week, it doesn't need to move.
Step 4 — Replace the operations layer.
After migration, how will you know if something breaks? The SaaS gave you a dashboard. You need a replacement.
One cron job. One morning check. One alert when something's wrong. Silent when everything's green. The same pattern for every SaaS you replace.
Real Example: Polsia → Internal Infrastructure
Polsia was a $45/month AI business surface. It hosted the How2OS homepage, product pages, Stripe checkout, onboarding emails, and a daily operations summary.
The migration took 3 hours:
- 7 pages → GHL blog (infrastructure already owned)
- 2 order forms → GHL + existing Stripe connection
- 3-email sequence → GHL workflow
- Daily summary → COSMIC OS cron job ($0/month). The mission —
how2os_ops_brief— runs every morning at 9am, checks surface health and pipeline activity, and alerts only when something needs attention.
5 dead pages were not migrated. 3 unused features were not rebuilt. The result: $540/year redirected to API budget. The surface runs on infrastructure the business already pays for.
Your Turn
Open your SaaS subscriptions. Pick the smallest one. Run the 4 steps:
- Audit: What does it actually do?
- Map: What do I already own that could do this?
- Migrate: Move the live pieces. Leave the dead ones.
- Operate: One cron job. One alert. Done.
Start with the one that costs $20-50/month. Let it run for two weeks. Then cancel.
This is how How2OS runs its own stack: fewer rented surfaces, more governed infrastructure. The playbooks we publish are patterns we use ourselves.
]]>