Before you buy another tool — five questions about the one you have

A conversation about a new tool usually starts with a symptom: something does not work, somebody cannot see something, something gets lost. The conclusion follows quickly and sounds like “we need a system for…”.

In most cases I see, the company already has a tool that does it. It was simply never finished: switched on in trial mode so as not to block anything, and left that way. Or half the features were configured because the rest required decisions nobody took.

A new tool in such an environment lands in exactly the same place.

Five questions

1. Is the tool we have finished?

The most common case: a security product running in monitor-only mode. It was switched on “as a trial” so as not to disrupt work, and nobody came back to switch it to blocking.

Formally the company has protection. In reality it has notifications nobody reads.

Checking takes fifteen minutes: open the console and see which mode it is in and when anyone last acted on a detection.

2. Does anybody actually look at it?

A tool reporting into a mailbox nobody opens is not a safeguard — it is a cost.

The practical question: who specifically is obliged to react, and within what time. If the answer is “well, everyone can see it”, that means nobody.

3. Is the problem technological or organisational?

This question saves the most money.

“Tickets get lost” is sometimes a missing ticket system — but far more often it is that tickets arrive through five channels at once: phone, email, chat, in the corridor and via a third person. A system will not fix a channel that people route around.

Similarly, “we do not know who has what” is sometimes a missing inventory, but more often it is that equipment is bought by somebody other than whoever configures it, and nobody passes the information on.

4. Is this already in something we pay for?

Office suites today include far more than companies take from them: multi-factor authentication, basic data protection rules, device management, mail archiving, simple document workflows.

Before another subscription joins the list, check whether the feature is already paid for. It happens more often than is comfortable to admit to whoever is selling.

5. Who will administer this in a year?

Every tool has a maintenance cost: updates, accounts, reviews, reacting to alerts. If there is no answer to who does that, the tool joins the list of things running half-configured within six months.

This question eliminates more purchases than the previous four combined.

When a purchase is the right call

The point is not to avoid buying. The point is the order. A purchase makes sense when:

• what exists has been finished and something specific is still missing,
• the missing thing is named as an action rather than a product category (“we want to see who downloaded a file from the share” rather than “we want DLP”),
• it is known who will use it and who will administer it,
• you can say how you will know the tool is working.

The last point is the test for the whole thing. If success cannot be described, neither can failure be noticed — and the tool will run half-configured for another three years.

How this goes in practice

A typical conversation ends in one of three ways:

The purchase is dropped, because the feature is in something already owned — just switched off.

The purchase is postponed, because what exists has to be finished first, and only then is it clear what is genuinely missing.

The purchase goes ahead, but in a different scope from the original — usually smaller, sometimes an entirely different product.

Each of those three outcomes beats buying something that in six months will be running in trial mode next to three other things running in trial mode.