How many tools are you actually running?

I ask the question in every first meeting: "how many tools do you have?" The answer always starts as a confident list, slows around the fifth, and almost always ends with the same sentence: "I must be forgetting some."

That's not a criticism. It's the normal symptom of a business that solved its problems one at a time, over years — which is exactly what it should have done each time.

Why nobody has the list#

Every purchase was justified#

Each tool was bought for a good reason, by a competent person, to solve a real problem. Accounts took their collections tool. HR took their people system. Sales have their CRM. Treasury has its own. Every one of those decisions is defensible in isolation.

Why does nobody know the full list?#

1. Nobody owns the whole. Each department knows its own tools; none knows the others'.
2. Small tools leave no trace. A local database built by one person to track one process appears in no budget. Yet it's critical the day that person is away.
3. Nothing ever leaves. You add a tool when a need appears; you almost never remove the one it partly replaces.

The inventory, in half a day#

How do you take an inventory of a company's software?#

This isn't an audit. It's a table, and it only needs three columns to become useful.

Talk to the managers — five to eight people is enough in a smaller company — with the same questions each time, and fill in:

The third column does all the work. It reveals re-keying, and re-keying is always a recurring cost nobody has put a figure on.

The two columns that help you sort#

Add two to sort by afterward: annual cost (license plus human time) and how many people actually use it. A tool at eight thousand a year used by two people isn't necessarily one to kill — but the question deserves asking once.

Oh, and there's also the little file I use for recharging. But that's not really software.

It is, though. A file that holds information nobody else has, that's used every month and that one person alone knows how to run, is software — with the particularity of having no support, no backup and no successor.

What the table almost always shows#

Three things, in this order.

Re-keying#

The same data entered two or three times: the customer in the sales tool then in invoicing, hours on paper then in payroll, the supplier invoice in the approval tool then in the ledger. Each re-keying has an hourly cost and an error rate. These are the first euros to recover, and the easiest.

Areas with no tool#

There is always an important process that lives nowhere: recharging between companies, tracking technicians' certifications, payroll data. Those areas are held up by one person, and that's the real risk.

Overlaps#

Two tools doing partly the same thing, each held by a different department, with two versions of the truth. Nobody wants to give theirs up, and both are right, because each has a function the other lacks.

What to conclude — and what not to#

The tempting conclusion is "we need to unify everything". It's almost always the wrong one.

A tool that works, that teams have mastered, and whose cost is known, has no reason to be replaced. Replacing it means paying to get back what you already had. Accounting, payroll, HR: leave them alone.

What to aim for, in order:

1. Remove the re-keying between the tools that stay. That's plumbing, not replacement, and it takes weeks.

2. Give a tool to the areas that have none. That's where custom development pays best, because there's nothing to take over and nothing to migrate.

3. Settle the overlaps, last, once the first two points have made the question less sensitive.

The question to ask before buying the next one#

Just one, and it's usually enough:

"What information will this tool hold that I don't already have somewhere?"

If there isn't any, you're buying an interface, not a capability. That happens, and it isn't illegitimate — but it's better to know before signing than after a year.

And a second one, for the cases where the first has a good answer: "who will re-key, and how many times a month?" If the answer is "nobody", excellent. If the answer starts with a first name, add that cost to the subscription before comparing.

What we do with that inventory#

The inventory is not there to delete tools. It is there to find where the same data is entered twice, because that is where the invisible work sits.

Our answer is almost never "let's replace everything". It is more often: your ERP stays the system of record, your spreadsheets stay for simulations, and we build the layer missing between the two — the one where information moves without being retyped. The page describing that layer explains what it does and what it does not.

Tool inventory: what to remember#

Almost no business knows its full list, and that isn't an organizational failing: it's the normal consequence of having solved its problems well, one at a time.

Take the inventory in three columns, the most important being "how data gets in and out". The re-keying it reveals is your first recoverable money.

Then resist unification. Connect what works, build for what doesn't exist, and keep replacement for last — or for never.