Skip to content

Where a mid-sized company’s technology money goes (and how to cut it)

If your technology bill grows every month and you are not quite sure why, start here: what to review, where the usual leaks are and how to cut without touching what works.

6 min read

Month-end comes round and someone in finance asks about a charge nobody recognises. And the cloud bill is higher than last month’s again, even though you have not sold more or hired anyone. You pay it, because switching off something that might be in use feels riskier than paying for it.

This is not bad management. It is what happens when technology has been bought piece by piece over the years: each department signed up for what it needed, each supplier left something running, and nobody has the full picture. The good news is that there is almost always room to cut without unnecessary risk.

What actually goes into your technology bill

A mid-sized company’s technology spending almost always falls into five groups, and the first step to cutting it is getting all of them onto one list:

  • Servers and cloud. Where your website, online shop, ERP (the system you use for orders, stock and invoicing) or backups live.
  • Per-user licences. Email, office software, accounting, the CRM (where your sales team tracks customers and deals). Paid per person.
  • Stand-alone subscriptions. Design tools, e-signature, surveys, project management. Many are paid by someone on a company card.
  • Domains and certificates. Small amounts each, but often spread across several accounts and renewing automatically.
  • Support and maintenance contracts. What you pay suppliers to be there when something breaks.

For the inventory, a spreadsheet with five columns will do: what it is, who uses it, what it costs per month, when it renews and who is responsible for it. Pull the data from the last year of bank and card statements, not just from the invoices that reach finance.

Where the money usually leaks

Leaks are rarely one big hidden cost; they are lots of small payments nobody checks:

  • Licences for people who have left. Someone left months ago and their account is still being billed.
  • Duplicated tools. Three departments, three different tools for sharing files.
  • Plans bigger than you need. You pay for the top tier of a product when the team only uses the basics.
  • Oversized servers running around the clock. A machine sized for your busiest season that runs exactly the same at four on a Sunday morning.
  • Forgotten test environments. Copies of your system that a supplier set up to try out a change and never switched off.
  • Storage nobody reads. Old backups and files piling up in the most expensive kind of storage.
  • Auto-renewing contracts. Annual contracts that renew themselves because nobody put a review date in the diary.

How cloud billing works, in plain words

The cloud charges for what you have switched on and reserved, not for what you actually make use of. When you rent servers from a provider such as Amazon, Microsoft or Google, you mainly pay for how powerful the machines are and how many hours they run, for the space your data takes up and, often, for data leaving their network.

Think of the office lights: leave them on overnight and you pay even though nobody is there. A server used only during office hours but running 24 hours a day is exactly that.

Two more details. Providers usually lower the price if you commit to a minimum level of use for one or more years, which only makes sense for things you know you will keep using. And data that is hardly ever opened can move to archive storage, which is cheaper and slower.

Quick wins and structural savings

Quick wins take days and carry almost no risk; structural savings take more work, but they are what changes the bill for good.

Quick wins:

  • Remove licences for people who have left.
  • Cancel subscriptions nobody recognises, after asking around.
  • Switch off test environments that are not being used.
  • Move tools to a cheaper plan when you are paying for features nobody uses.

Structural savings:

  • Resize servers to what they really use and schedule them to switch off when they are not needed.
  • Consolidate duplicated tools and help the team through the change.
  • Renegotiate support contracts with evidence: how many incidents there were and how long they took to fix.

An honest warning: not every saving is worth it. Switching tools to save a little each year, at the cost of weeks of work and disruption for the team, is rarely worth doing.

The risk of cutting blindly

Cutting without knowing what each item does can cost far more than the bill you were trying to reduce. The classic case: someone cancels a service with an odd name, and it turns out to be the backups or the link that sends orders from your online shop to the warehouse.

Three things should never be touched without checking first: backups, anything security-related and anything that connects two systems together. Before cancelling, suspend for a few weeks if the supplier allows it. Before deleting data, make sure you have a copy and are not legally required to keep it.

How to keep costs under control

Costs stay under control when one person owns them and there is a short monthly review, not a big clean-up every few years. Without an owner, you will be back where you started within a year.

The owner does not need to be technical: just someone organised who keeps the inventory up to date, approves any new sign-up and receives the renewal notices. The monthly review can take half an hour: compare with last month and ask about every unexplained increase.

What you can do this week:

  • Get the last year of bank and card statements and mark everything that is technology.
  • Move that list into a spreadsheet with cost, owner and renewal date.
  • Check the licences against your current staff list.
  • Ask each department which tools they genuinely use.
  • Put the renewals for the next six months in the diary.
  • Name one person as owner of the inventory.

How we approach it at vitamina.dev

First, we understand what each item does. We build the inventory with you, review how your infrastructure is set up and separate what can be cut now from what needs a plan. Our team has experience with critical systems and infrastructure cost optimisation, so we know which pieces are delicate.

Our rule is simple: bring down what you pay without touching what works. If something is not worth changing, we will tell you.

Frequently asked questions

How do I know if I am overpaying for the cloud?

The clearest sign is a bill that keeps rising while the business has not grown at the same pace. Another clue: servers running 24 hours a day that are only used in office hours. Ask for a breakdown by service and question every line you do not understand.

Is it cheaper to run our own servers than to use the cloud?

It depends on how you use your systems. With steady, predictable use, your own server or one rented at a fixed price can cost less, but you take on maintenance, hardware replacement and the responsibility for keeping it running. The cloud is more flexible, although without control it is easy to overpay.

Can I cancel licences and subscriptions without breaking anything?

In most cases, yes, as long as you first check who uses them and for what. The safe approach is to suspend first and only cancel if nobody has missed it after a few weeks. For backups, security and anything that connects systems together, check first with someone who knows what they do.

Does this sound like your company?

Tell us about your case and we’ll tell you where we would start.

Tell us about your company

Keep reading

All articles