All posts

How Magic of Me cancelled €500/month in SaaS licenses

6 tools. €500/month. It worked fine, but it was complex. When something broke, debugging took hours. This is the story of how we consolidated everything into one custom system.

TL;DR

Magic of Me ran a children's-book pipeline across 6 SaaS tools costing ~€500/month. We replaced the stack with a single custom platform. Bills dropped to ~€50-100/month, and the process finally just works.

That was the situation at Magic of Me, a platform for personalized children's books. AI writes the story, generates the illustrations, builds the PDF, and after a human quality check it goes to the printer.

Fun product. But behind the scenes? A chain of 6 tools, each doing its own thing.

The old setup

To make a single children's book, data had to flow through 6 different systems:

Tool What it did Cost/month
BannerBear Overlay text on images €150
CloudConvert File conversions (JPEG→PDF) €30
Airtable Data storage €45
Cloudinary Image hosting €50
n8n Automation flows €40
Misc Smaller tools ~€50

Total: ~€365-500/month

Each tool did its job well. But the whole?

  • Data had to be copied back and forth
  • If one tool failed, everything stopped
  • Debugging was a nightmare: which tool broke?
  • Adding new features meant: does this fit in all 6 tools?

What we changed

We built one custom system that does everything.

No more 6 separate tools. One platform that:

  • Pulls in orders
  • Lets AI write the story
  • Generates the illustrations
  • Places text in the right spot
  • Produces the PDF
  • Queues it up for human review
  • After approval → straight to the printer

One flow. No data copying. No "which tool broke?"

The result

Licenses cancelled:

  • BannerBear: cancelled
  • CloudConvert: cancelled
  • Airtable: replaced with Supabase (free tier)
  • Cloudinary: no longer needed
  • n8n: no longer needed

What we still pay for:

  • Supabase: €0 (free tier)
  • Vercel hosting: €0 (free tier)
  • Shopify: still needed for the storefront
  • AI API costs: ~€50-100/month (volume-dependent)
From ~€500/month to ~€50-100/month.

But that's not even the main point

The cost savings are nice. But the real win?

The process just works now.

  • No more manual data copying
  • One place to debug when something breaks
  • Adding new features is easy, it's our own system
  • AI is baked in, tuned for exactly this process

You can't get that from 6 separate SaaS tools. They're generic. They don't understand your specific flow.

When does this work for you?

You might recognize this pattern:

  • You pay for several tools that "almost" work together
  • You copy data by hand from A to B
  • When something breaks, you don't know where to look
  • You want to build something new, but it doesn't fit in your current stack

If you're spending more than €300-500/month on tools that don't play nicely together, consolidating into custom is worth a look.

Frequently asked

When does consolidating SaaS tools into one custom system make sense?

If you're paying more than €300-500/month for several tools that don't play nicely together, you're manually copying data between them, and debugging means hunting through 6 dashboards to find the one that broke, consolidation is worth exploring.

Isn't custom software more expensive than SaaS?

Not anymore. With AI coding tools, custom builds that used to take months now take days. Magic of Me went from ~€500/month in SaaS to ~€50-100/month in infrastructure and AI API costs.

What was the biggest win: cost or process?

The process. Cost savings are nice, but the real upside is one place to debug, no manual data copying, and new features that actually fit the flow because the system was built for it.

Paying too much for tools that don't fit?

Book a no-strings 30-minute call to see what's already possible today.

Talk to us

Read next

Engineering

Why custom software is now cheaper than SaaS

Strategy

Too many tools? The answer isn't another tool.

Practical

How much does an app cost to build in 2026?

AI & Software

Software isn't what it used to be

Talk to us

Let's find the 20% of work that shouldn't be manual.

A 30-minute call. No pitch deck, no pressure. We listen first. If AI isn't the right answer, we say so.

Reply within 24 hours
No pitch deck, no pressure
Honest if AI isn't the fit