How many apps does a small business use, and when does SaaS sprawl become a cost problem?

Ask the managing partner of a 20 person firm how many software subscriptions the business pays for and you will usually hear a number between 10 and 20. Go through the card statements and expense claims and the real count is often two or three times that. Nobody chose to run that many tools; each one was a sensible purchase on the day someone needed it.

The direct answer: there is no reliable single number for small businesses, because every published average comes from a vendor's own customer base. What matters more is when the count turns into a cost problem. For most 10 to 50 person firms that happens when three things appear together: tools that do the same job, seats paid for people who no longer use them, and staff copying data between apps by hand. This article shows how to spot each one and count your own stack in an afternoon.

What the published numbers say

The figures you see quoted are vendor research, collected from each company's own customers, so treat them as context rather than benchmarks.

• Okta's Businesses at Work 2025 report, based on anonymized data from its customers, put the average at 101 apps per company. Okta's customers skew toward mid-size and large organizations, so a 20 person firm will usually sit well below that.
• Zylo, which sells SaaS management software, states on its license management page that organizations leave 46% of licenses unused on average. Again, that is drawn from its customer base, which is mostly larger companies.

The useful lesson from these reports is not the headline count. It is that app numbers grow without anyone deciding they should, and that a large share of paid seats go unused.

Why the count grows in small firms

In a 10 to 50 person law, accounting, consulting or clinic practice, sprawl rarely comes from IT. It comes from good intentions:

• Free trials that convert. Someone signs up to try a tool, adds a card, and the trial quietly becomes a subscription.
• Team-by-team purchases. Billing picks one form builder, intake picks another, marketing picks a third.
• Bundles you already own. Microsoft 365 and Google Workspace both include file storage, chat, forms and video calls, yet many firms pay separately for standalone versions of the same things.
• Leavers' tools. A departing employee's personal subscriptions keep billing to the firm card.
• Point fixes for gaps. A tool bought to move data between two others, because nobody connected them directly.

None of these is a mistake on its own. Together they create a stack that nobody has a full picture of.

When sprawl becomes a cost problem

The number of apps is not the problem. Thirty well-chosen, well-connected tools can run a firm smoothly. Sprawl becomes a cost problem when one or more of these signs appear.

1. Overlap: two tools doing one job

The most common overlaps in professional firms are project or task management, file storage, forms, e-signature, chat and scheduling. Each duplicate costs its subscription, but the bigger cost is that information now lives in two places, and staff waste time checking both.

2. Shelfware: seats nobody uses

Per-seat pricing punishes firms that never remove people. Look for seats assigned to former staff, people who have not logged in for 60 days or more, and premium tiers bought for one feature that nobody uses.

3. Re-keying: people acting as the integration

When a new client is typed into the CRM, then the practice management system, then the accounting tool, your staff are doing the job an integration should do. This cost does not appear on any invoice, which is why it is usually the largest. Our worked example on what manual data entry costs a 20 person firm shows how to price it.

4. Renewals nobody owns

Annual contracts renew automatically. If nobody owns the renewal date, nobody reviews the tool before the next year is committed.

5. Security and offboarding gaps

Every app is another login to remove when someone leaves and another place client data sits. If you cannot list where client data lives, check with your compliance lead before it becomes a problem.

How to count your own apps in an afternoon

You do not need software to do this. A spreadsheet and three sources are enough.

1. Pull 12 months of card statements and bank debits. Filter for software vendors. Annual subscriptions only show up once a year, so a single month is not enough.
2. Collect expense claims. Tools bought on personal cards and claimed back are the ones most often missed.
3. Check sign-in records. If staff use "Sign in with Microsoft" or "Sign in with Google", your admin console shows which third-party apps have been granted access.

For each tool, record:

Then sort by "what job it does". Any job with two tools against it is a consolidation candidate.

What to do with the list

Give every tool one of four verdicts:

• Keep: it does its job and people use it.
• Configure: it is the right tool but set up badly, connected to nothing, or on the wrong plan.
• Consolidate: another tool you already pay for does the same job.
• Replace: it does not fit how your firm works.

Act at renewal where you can. Cancelling mid-contract rarely saves money, but a renewal date three months away is enough time to migrate. Before cancelling anything, check which automations depend on it. A Zap or Make scenario that points at a cancelled tool fails quietly, and the first sign is often a missing client record. Our comparison of Zapier and Make for a 20 person firm covers how to keep those automations owned and monitored.

Finally, put one person in charge of new purchases. Not a heavy approval process; simply a rule that new tools are added to the list with an owner and a renewal date.

When cutting apps is the wrong move

Consolidation is not always the answer. Specialist tools often exist for good reasons: a practice management system built for law firms, or a client portal an accounting firm's clients already know. Forcing everything into one general-purpose platform can cost more in lost productivity than the subscriptions it saves. The goal is fewer duplicates and better connections, not the smallest possible number of logos.

How Flow Mesh helps

Our stack audit does this exercise for you: it lists every tool you pay for, gives each one a verdict (keep, configure, consolidate or replace) and writes a scope for the connections worth building. It starts with a first call, and any implementation is quoted in writing after the audit. See what a workflow automation audit includes for the detail.

Key takeaways

• Published app counts are vendor research from mostly larger companies; count your own stack instead.
• The app count is not the problem. Overlap, unused seats, re-keying and unowned renewals are.
• Twelve months of statements, expense claims and sign-in records give you a full list in an afternoon.
• Give every tool a verdict and act at renewal, after checking which automations depend on it.
• Fewer duplicates and better connections matter more than the smallest possible stack.