Your Client Pays for Twelve Tools. They Use Three.

Most small businesses rent a dozen overlapping subscriptions, use a fifth of the features, and own nothing. One afternoon with a frontier model maps the whole mess. Then the real question starts: keep renting, or rebuild what you actually use.

I walk into small operations and find the same drawer full of chargers. A CRM plus a spreadsheet tracking the same customers. Three tools that all send email. Two that manage tasks. Seats nobody remembers buying, renewing every year on autopilot. Nobody planned this stack. It accumulated one signup at a time, and now it costs real money every month while the team lives in maybe three tabs.

I am Brian Marvin, an AI-native fractional CTO. Auditing that stack used to take me a week of interviews and invoice hunting. Now I can map it in an afternoon with a frontier model pointed at each tool. This article is the playbook I run, and the decision I help owners make at the end of it: keep the rent, or own the infrastructure.

Step one: inventory everything with a price tag

List every subscription, every seat, every annual contract. CRM, project management, chat, file storage, billing, support desk, marketing automation, scheduling, e-signature, the lot. For each one, write the annual cost, the primary users, and what the business thinks it does.

This sounds trivial and it never is. I have never done this audit without finding at least one tool nobody uses and one seat count that drifted upward. Pull the actual invoices, not the admin dashboard. The invoices tell the truth about what leaves the bank. Expect the list to land near ten to fifteen tools for a team under twenty people. That count surprises every owner exactly once.

Step two: document used versus paid

Point the model at each tool and document which features the team actually touches against what the license covers. Read the help docs, click through the real workflows, ask three users what they open daily. Then write the gap down in plain words.

The pattern repeats so reliably I now warn owners up front. Most teams use about a fifth of what they pay for. The CRM is a contact list with a Ferrari engine. The project tool is a shared to-do list billed as a portfolio suite. The marketing platform sends one newsletter a month at enterprise tier. None of this is the team's fault. The tools were bought for a future that never arrived, then renewed out of habit.

This is the same discipline I describe in where to start and what to skip. Name the actual job being done before you judge the tool. If the job is one newsletter a month, say so. The honest sentence makes the next steps obvious.

Step three: map every overlap

Now lay the tools side by side and mark the redundancies. Three tools sending email. Two managing tasks. A CRM and a spreadsheet holding the same customer list. A chat app plus a support desk answering the same questions. Document every duplicate, every overlap, every pair that never talks to each other.

I draw this as a simple grid. Tools down the side, jobs across the top, a mark where they collide. The visual does the persuading. Owners who defended each subscription in isolation go quiet when they see four marks in the email column. Nobody chose that redundancy. It grew one free trial at a time.

Watch for the quietest overlap of all: the tool versus the spreadsheet. Almost every small business has a shadow system in sheets that duplicates the official tool because the official tool was too slow or too confusing. That spreadsheet is telling you what the team actually needs. Listen to it.

Step four: keep, consolidate, or rebuild

Not everything gets replaced. Some tools earn their license honestly, and I say so in every audit. The rule I use is simple. If the team lives in it daily, it integrates well, and the price fits the value, it stays. Everything else becomes a rebuild candidate.

Consolidate first where you can. Two task tools become one. The newsletter moves into the tool you already pay for. Seats get cut to the real headcount. This pass alone usually recovers a meaningful slice of the annual spend with zero new software. Do it before you build anything.

Then look at what is left with fresh eyes. A scheduling flow the team uses twice a day but pays enterprise rates for. A client portal that is really four forms and a dashboard. A reporting pack assembled by hand every Friday from three exports. These are the rebuild targets. A frontier model can now deconstruct a live application, map its features, and document how it fits the business, which used to be the hardest part of a rewrite. That capability is what makes this whole audit end in a build decision instead of a shrug.

If you are weighing rebuild versus buy in general, I laid out the decision in build, buy, or AI-generate. The short version is that rebuild wins when the job is narrow, stable, and core to how you operate. Rent wins when the domain shifts under you and someone else tracks it.

Why owning beats renting for the core pieces

Here is the argument I make to owners at the end of every audit. Rent is fine for commodities. Nobody should self-host email. But the pieces that run your actual operation should be infrastructure you own, not subscriptions that can reprice, deprecate, or hold your data hostage.

Owned infrastructure compounds. The first rebuild is the most expensive because you are learning. The second reuses the patterns. By the third, you have a small platform shaped exactly like your business, with no per-seat tax on growth. Every new hire stops costing you five license increments. Every workflow change ships in an afternoon instead of waiting on a vendor roadmap.

There is also a quieter win. When you own the core tools, your data lives in one place under your control. Reporting stops being Friday export archaeology. Support questions get answered from one system. Procurement stops asking how customer data flows through six vendors. I cover the cost side of running that platform in orchestration and spend. The numbers are about agents, but the discipline is identical.

Run the audit in one afternoon

The whole exercise fits in a single focused afternoon if you scope it right. Morning: pull invoices and list every tool with its cost and users. Midday: point the model at each tool and write used-versus-paid in plain sentences. Afternoon: grid the overlaps, then sort every tool into keep, consolidate, or rebuild. End the day with a one-page recommendation matrix the owner can act on.

Bring the owner into the room for the last hour. Walk the grid, name the redundancies out loud, and agree on the first consolidation and the first rebuild candidate. One of each. Ship those two before you touch the rest. That is the same prove-it-small discipline from the one person company playbook. One win funds the next.

FAQ: How many tools does a typical small business actually have

In my audits, a team under twenty people usually pays for ten to fifteen subscriptions. They open three or four daily. The rest are renewals running on habit. The invoice pull in step one almost always surprises the owner, which is why I insist on invoices rather than the admin dashboard.

FAQ: Should we replace everything with custom builds

No. Keep what the team lives in daily, what integrates well, and what is priced fairly. Consolidate the easy duplicates first. Rebuild only the narrow, stable, core workflows where rent exceeds value. Commodities like email stay rented. Your client portal, your scheduling flow, your reporting pack: those are ownership candidates.

FAQ: What does the one-page matrix look like

One row per tool. Columns for annual cost, primary users, features actually used, overlaps with other tools, and the verdict: keep, consolidate, or rebuild. One page, plain words, no jargon. The owner should be able to act on it without you in the room.

FAQ: How do we keep the stack from rotting again

Quarterly subscription review, same grid, thirty minutes. New tool requests answer three questions first: what job, what does it replace, who owns the renewal. Kill the spreadsheet shadow systems by fixing the official tool or owning the workflow. The audit is a habit, not a project.

About the Author

I am Brian Marvin, an AI-native Fractional CTO with 30 years in technical leadership. At empowered.guru, I help startups and small businesses build MVPs, shape roadmaps, and make AI-powered technology decisions that scale.