Every finance team in the world has now asked the same question: how many software subscriptions do we actually have? The answer is always higher than the list, and the gap between the two is the project.
Reported consolidation is running in the 20% to 30% range, and it is not purely a cost exercise. Something structural is pushing it.
Why Now
Three forces arrived together. Budgets tightened. AI features started shipping inside the large platforms companies already pay for, making the standalone tool that did one thing look expensive. And fragmented tools became an active liability rather than a mild annoyance, because scattered data is exactly what makes an AI assistant useless.
That third point is the one people underestimate. When your customer information lives across nine systems, no assistant can answer a question about a customer. Consolidation is now an enabler, not just a saving.
What Gets Cut First
| Category | Risk of being cut | Why |
|---|---|---|
| Single-feature point tools | Very high | The platform now does it, adequately |
| Duplicate tools across teams | Very high | Two departments, same job, two invoices |
| Low-adoption seats | High | Easy win, no argument |
| Systems of record | Very low | Migration cost exceeds any saving |
| Regulated or audited tools | Low | Compliance beats budget |
| Anything with your data history | Low | The data is the lock-in |
How to Run the Cut Without Breaking Things
- Find the real inventory. Expense reports, identity provider OAuth grants and card statements. The official list is always wrong.
- Map tools to workflows, not to departments. A tool with four users may sit in the middle of your billing process.
- Check the integration graph before cancelling. The cheapest tool is often the one holding two expensive ones together.
- Talk to the actual users first. Low seat count is not low importance, and the person who will tell you why is not in the finance meeting.
- Cancel in stages with a rollback window. Disable access, wait a month, then terminate. Some pain only shows at month end.
The Trap: Consolidating Into a Worse Tool
The bundled feature inside your big platform is usually adequate and occasionally terrible. Adequate is fine for most things and a disaster for the workflow your revenue depends on.
Judge each replacement on the job it does rather than on the invoice it removes. Saving twelve thousand a year while making your support team 20% slower is not a saving, it is a cost moved somewhere finance does not measure.
If You Sell Software
You are either the consolidator or the consolidated, and the deciding factors are unglamorous. Depth of integration into your customer other systems. Whether your data would be painful to lose. Whether an agent can reach your product without a human.
Point solutions with shallow integrations and no data gravity are the first line in the spreadsheet. If that describes you, deepen the integration story this quarter rather than adding features.
Conclusion
Build the real inventory from expense reports and OAuth grants, map tools to workflows rather than owners, and check what breaks before you cancel anything. Cut duplicates and single-feature tools first, protect systems of record, and stage every cancellation with a rollback window. The consolidation wave is genuine, and the organisations that get hurt are the ones who treated it as a spreadsheet exercise rather than an operations one.
Frequently Asked Questions
How much can a mid-size company realistically save?
Reported consolidation in the 20% to 30% range of tool count is common, though the spend saving is usually smaller, because the tools you cut are rarely the expensive ones.
Does consolidation hurt productivity?
It can, when you consolidate into a weaker tool for a critical workflow. Cut breadth, protect the two or three workflows your business actually runs on.
What about shadow tools people expense personally?
Those are your best signal of an unmet need. Find them, understand why, and either approve them properly or fix the gap that created them.