Print twelve months of statements and mark every recurring charge as keep, cancel, or unknown. The unknowns are usually where the money is.

The method is the whole thing

Everybody agrees subscriptions accumulate and almost nobody does the exercise, because it sounds tedious and vague.

It is neither if done in one specific way: twelve months of statements, printed or exported, read line by line with a pen.

Not from memory, not from a list of what you think you pay for, and not from the software's own billing page.

The statements are the only complete record, because they include the things you have forgotten and the things nobody told you about.

Four marks per line

The fourth is the interesting one. Every business doing this finds two or three charges nobody can identify from the name on the statement, and those almost always turn out to be cancellable.

Work in annual figures

Which is what makes the decisions obvious.

A charge of nineteen dollars a month reads as trivial and two hundred and twenty-eight a year reads as a decision.

Convert everything to an annual figure in a second column, including the things billed annually already.

Then total it.

Most small businesses are surprised by the total and not by any individual line, which is precisely why the exercise works and why monthly review does not.

The number at the bottom is what makes somebody actually cancel things.

Nobody has ever cancelled a subscription because it was nineteen dollars a month.

A worked example

A business exported twelve months of card statements and found thirty-one recurring charges.

Nineteen were kept, four downgraded to smaller plans, five cancelled outright, and three were unidentifiable.

Of the three unknowns, one was a design tool from a project in 2021, one was a scheduling product they had trialled and forgotten, and one was a duplicate of something they already paid for under a different name.

The annual saving was a little over two thousand dollars.

The exercise took about two hours including the cancellations.

That works out to a better hourly rate than almost anything else available to them in December.

Look for the duplicates

Which is the category people find most surprising.

Two tools doing the same job, bought at different times by different people for different reasons.

Storage in three places, two products that both send email, or a feature included in something you already pay for and also bought separately.

That last one is common, since bundled features get added to products you already have and nobody re-examines what they were separately buying.

Read what your main tools now include before renewing any of the smaller ones around them.

A surprising share of what most businesses pay for separately is already included somewhere else.

Cancel on the day rather than diarising it

Which is where these audits usually fail.

A list of things to cancel, written in December and actioned in February, mostly does not get actioned.

So cancel each one at the moment you mark it, in the same sitting, rather than making a list.

Where cancelling requires exporting something first, do the export immediately and cancel the same day.

Where a subscription runs to a date you have already paid for, set the cancellation to take effect then rather than losing what you paid.

The difference between an audit that saves money and one that does not is entirely whether the cancelling happened in the same session.

Check what is charged to whom

Which is the structural half.

Business subscriptions on a personal card, personal subscriptions on the business card, and anything on a card belonging to somebody who has left.

That last is how a service silently stops working nine months later.

Move everything business-related onto one business card while you have the list in front of you.

It also makes next year's exercise take twenty minutes rather than two hours.

Keep the list for next December

Which compounds the value.

Save the marked-up list with the date, and next year compare rather than starting again.

New lines that appeared during the year are the ones to examine, and everything already marked keep can be skimmed.

Three years of those lists shows what your actual software cost is and whether it is growing faster than the business.

That is a number very few small businesses can produce, and it is worth having.

Software cost growing faster than revenue is a finding, and one nobody notices monthly.

Ask what each one is actually for

Which is the question that decides the borderline cases.

For every line marked keep, say in one sentence what job it does and what would happen if it stopped tomorrow.

Anything you cannot answer in a sentence is a candidate regardless of whether somebody logs into it, since paying for a tool nobody can justify is the same as paying for one nobody uses.

That question also finds the tools bought to solve a problem that has since gone away, which are the hardest to spot because they were genuinely useful once.

Write the sentence beside each line, since next year it saves you making the same judgement again from scratch.

The counter-case

Cancelling can cost more than it saves.

A tool used occasionally but critically, or one holding data you would have to migrate, is worth its cost even at low usage.

Downgrading also occasionally removes the one feature that mattered, which is discovered at the worst moment.

And a business with only six subscriptions has very little to audit and should spend the two hours elsewhere.

The exercise scales with how many people in the business have ever bought software.

One person buying carefully produces a much shorter list than three buying independently.

Export twelve months of statements, mark every line, convert to annual figures, and cancel in the same sitting.

The two hours

  1. Export twelve months of statements.
  2. Mark every recurring line.
  3. Convert to annual figures.
  4. Total it.
  5. Investigate the unknowns.
  6. Cancel in the same sitting.
  7. Save the list for next year.

Step six is the difference between an audit that saves money and one that produces a list, since a cancellation deferred to February mostly does not happen at all.

The prior version of this is at subscriptions you are still paying for.


Frequently asked questions

What is the method?

Twelve months of statements, exported or printed, read line by line with a pen. Not from memory and not from a list of what you think you pay for.

Why statements rather than a list?

Because they are the only complete record, including the things you have forgotten and the things nobody told you about.

How should I mark each line?

Keep, cancel, downgrade, or unknown. Every business finds two or three charges nobody can identify, and those are almost always cancellable.

Why convert to annual figures?

Nineteen dollars a month reads as trivial and two hundred and twenty-eight a year reads as a decision. The total at the bottom is what makes people cancel.

What surprises people most?

Duplicates. Two tools doing the same job, or a feature included in something you already pay for and also bought separately.

When should I cancel?

In the same sitting, at the moment you mark it. A list written in December and actioned in February mostly does not get actioned.

West Coast Media Solutions Inc. provides web design, web development, hosting, digital marketing, and business consulting to organisations across Canada, drawing on more than twenty-five years in the field.

Never audited your recurring charges?

Export twelve months of statements and read them with a pen. The total is the part that surprises people.

Start a Conversation