Every week, someone pitches me on a new AI tool that’s going to transform their business. And almost every time, my first question is the same: “What are you currently paying for that you’re not using properly?”

The awkward silence that follows tells me everything. Most small businesses are already sitting on technology they’ve barely explored, paying monthly subscriptions for features they didn’t know existed, and duplicating capabilities across multiple tools. Adding another AI product on top of this mess doesn’t fix anything — it makes the mess more expensive.

Before you spend another dollar on AI, do a proper audit of what you’ve got. Here’s how.

Step 1: Build a Complete Inventory

This sounds basic, but almost nobody has a current, accurate list of every software tool their business pays for. Start with three sources:

Check your bank and credit card statements. Go back 12 months and flag every recurring software charge. You’ll find subscriptions you forgot about, tools that auto-renewed without anyone noticing, and pricing tiers that crept up since you last checked. For a 20-person business, it’s not unusual to find $500-$1,000 per month in subscriptions that nobody actively uses.

Survey your team. Send a simple form to every employee asking them to list the tools they use daily, weekly, and rarely. Ask them what they wish they could do but can’t. You’ll discover shadow IT — tools people signed up for with personal credit cards because the official tool was too slow or didn’t do what they needed.

Check your identity provider. If you’re using Microsoft 365, Google Workspace, or an SSO provider like Okta, look at the list of connected applications. This captures tools that authenticate through your corporate identity but may not show up on financial statements (free tiers, trials that never converted).

Build a spreadsheet. For each tool, record: name, category (CRM, accounting, communications, etc.), monthly cost, number of licensed users, number of active users, and primary purpose.

Step 2: Map Overlapping Capabilities

With your inventory complete, look for overlap. This is where the waste lives.

Common overlaps I see constantly in Australian SMBs:

• Communication tools: Paying for Slack AND Microsoft Teams AND a separate project management chat tool. Pick one.
• File storage: OneDrive, Google Drive, Dropbox, and a shared NAS server, with files scattered across all four. Consolidate.
• Project management: Asana for some teams, Monday.com for others, Trello for a third group. This usually happened because different managers chose different tools and nobody standardised.
• CRM and sales: A full CRM subscription alongside sales tracking spreadsheets, because the CRM was poorly configured and people reverted to what they knew.
• Accounting add-ons: Multiple plugins doing expense tracking, invoicing, and reporting that duplicate what the core Xero or MYOB installation already does.

For each overlap, make a decision: consolidate onto one tool, or intentionally keep both with clear role separation. There’s no wrong answer, but the decision needs to be deliberate, not accidental.

Step 3: Assess Usage Depth

Most businesses use about 20% of the features in the tools they pay for. This isn’t an exaggeration — Microsoft’s own research suggests that average enterprise users actively use fewer than a quarter of available Microsoft 365 features.

For each tool in your inventory, ask:

• Are we using the automation features? Most modern SaaS tools include workflow automation. If your CRM can automatically send follow-up emails after meetings but your team is doing it manually, you have a training problem, not a technology gap.
• Are we using the reporting and analytics? The data you need might already be sitting in your existing tool, unqueried.
• Are we on the right pricing tier? You might be paying for an enterprise plan with features your team doesn’t use, when a cheaper tier would do. Or conversely, you might be on a basic plan and paying for a third-party add-on that’s included in the next tier up.
• Have we customised it for our workflows? Many tools ship with generic defaults. Thirty minutes of configuration — custom fields, automated triggers, tailored dashboards — can transform a tool from “not quite right” to “exactly what we need.”

Step 4: Identify the Actual Gaps

Now — and only now — identify what’s genuinely missing. Be specific. “We need AI” is not a gap statement. These are gap statements:

• “We spend 15 hours per week manually categorising support tickets that could be auto-routed.”
• “Our inventory forecasting is based on gut feel, and we’re consistently over-ordering by 20%.”
• “Our sales team can’t tell which leads are most likely to convert, so they treat them all equally.”
• “We have 10,000 customer reviews and no way to extract trends from them.”

Each of these is a specific problem with a measurable cost. Some of them will have solutions in tools you already own. Others genuinely need a new tool or AI capability. The difference between an informed purchase and a wasted subscription is this specificity.

Step 5: Check What Your Existing Tools Already Added

Here’s one that catches people off guard. Many SaaS tools have added AI features in the past 12 months that existing subscribers get for free or at minimal additional cost.

HubSpot added AI-powered content assistants, lead scoring, and conversation intelligence. Xero added AI-driven cash flow predictions and invoice categorisation. Microsoft 365 Copilot is rolling out across the entire suite. Google Workspace has Gemini integration throughout Docs, Sheets, and Gmail. Notion added AI summarisation and writing assistance. Even Canva has AI image generation built in now.

Before subscribing to a standalone AI tool, check whether your existing subscriptions already include the capability you’re looking for. You might be a settings change away from the functionality you thought required a new purchase.

Step 6: Calculate Before You Commit

For any new AI tool that makes it through this process, do the maths:

Cost per outcome, not cost per seat. A $50/month AI tool that saves 10 hours of manual work per month at $40/hour is delivering $400 of value for $50. That’s a clear win. A $200/month AI tool that produces slightly better email subject lines? Harder to justify.

Implementation cost. Every new tool takes time to set up, configure, learn, and integrate. Budget 2-4 weeks of reduced productivity during the adoption period. If you’re adding a tool during your busiest season, that cost is amplified.

Integration requirements. Does the new tool connect to your existing systems? If it needs data from your CRM but doesn’t have a native integration, you’ll need middleware (Zapier, Make, or custom API work). That’s additional cost and complexity.

Exit cost. What happens if the tool doesn’t work out or the vendor raises prices? How easily can you extract your data and switch to an alternative? Vendor lock-in is a real risk, especially with AI tools that learn from your data over time.

The Audit Cadence

Do this exercise quarterly. Not because your tech stack changes that fast, but because your business does. The tool that was perfect six months ago might be underused now because your team’s workflow has shifted. The capability you needed last quarter might now be built into a tool you already own.

Set a calendar reminder. Spend two hours every quarter reviewing the inventory, checking usage, and making deliberate decisions about what stays, what goes, and what gets added.

The Point

The businesses that get the most value from technology aren’t the ones that buy the most tools. They’re the ones that get the most out of fewer tools, chosen deliberately and configured properly. An audit isn’t glamorous. It doesn’t feel like innovation. But it’s the single best thing you can do before your next AI purchase.

Clean up the foundation first. Then build on it.