What Actually Affected Your Business

The test is one question

For anything you were told mattered this year, ask which number in your business changed, and in which month.

Enquiries, orders, traffic, average order value, conversion rate, or time spent on something.

If you can name the number and the month, it affected you and is worth understanding.

If you cannot, it was news rather than a consequence, however important it may have been to the industry.

Why the distinction is hard to hold

• Suppliers benefit from urgency.
• Publications need a story weekly.
• Peers repeat what they read.
• Anxiety feels like information.
• Nobody publishes the nothing-happened follow-up.

The last is the structural problem. Every prediction gets an article and almost none gets a correction, so the record you remember is the one made in March rather than what actually occurred by November.

Work backwards from your own numbers

Which is the reliable direction.

Instead of asking whether a given change affected you, look at your monthly figures for the year and find any month that moved unexpectedly.

Then ask what happened around it.

That produces a much shorter list than working forwards from a year of announcements, and every item on it is real by definition.

Most years that exercise produces two or three moved months, and the causes are usually local: a competitor opened, a large customer left, a listing changed, or something on your own site broke.

Industry-wide events rarely appear in it.

A worked example

A business worried all spring about how the new tools would affect enquiries.

Reviewing the year in December, enquiries were slightly up and the only two unusual months were August, which was down, and October, which was up.

August turned out to be a form that had silently stopped sending for eleven days.

October was a mention in a local publication.

Neither had anything to do with the thing they had spent six months worrying about.

The broken form had cost more than every industry change that year combined.

Local causes dominate

Which is the pattern across most small businesses.

A competitor's new location, a road closure, a review, a listing edit, a supplier change, a member of staff leaving, or a technical failure on your own site.

Each of those moves a number more than any platform announcement will.

So the attention a business can spare is better spent on what is happening within a few kilometres than on what is happening in the industry.

That is not a claim that industry changes never matter; it is an observation about relative frequency.

Across four years of writing these reviews, the local causes have won almost every single time.

That consistency is itself worth something.

Separate the three responses

Which keeps this from becoming complacency.

Some things require action now, because they have a date and a loss attached.

Some require measurement, so that you would notice if they began to matter.

And some require nothing, which is a legitimate category and the one most things belong in.

Deciding which of the three a change belongs to, once, and writing it down, is faster than revisiting the question every time somebody mentions it.

The middle category is where most of the real caution belongs, and it is the smallest of the three.

Set up the measurement instead of worrying

Which converts anxiety into something useful.

If you are concerned that search traffic may fall, record what it is now and check quarterly.

If you are concerned about a platform, record what share of enquiries it produces.

That takes an hour and replaces an ongoing worry with a number you will see move if it moves.

It also means that when something does change, you can say when it started, which is the difference between responding and guessing.

Most businesses cannot currently answer either of those two questions.

Review it in December, once

Which is the discipline this depends on.

Once a year, look at what you were told to worry about, and mark each item against what actually happened.

Keep the marked list, because the value compounds: after three years you have a record of how often the warnings were right.

That record is the most useful filter you can build, and nobody can supply it to you.

It also makes next year's conversations considerably shorter.

Somebody who has the record can answer in a sentence rather than debating.

Ask the people who answer the telephone

Which is a source of evidence that no analytics package holds.

Whoever takes the calls knows what customers asked about this year, what they mentioned having seen, and what they complained about repeatedly.

That is direct evidence of what changed in your market, arriving months before it shows in any figure, and it is usually never written down.

Ask them once in December what was different about this year, and write the answers beside your own list.

In a small business that conversation frequently identifies the real cause of a moved month faster than the analytics does.

The counter-case

Some effects are slow.

A change that moves nothing this year can move a great deal over three, and a test that only accepts an immediate number will miss it.

Small businesses also lack the volume to detect modest effects, so an absent signal is not proof of an absent cause.

And some industries genuinely were reshaped this year, and for those the correct response was substantial rather than watchful.

The test still applies perfectly well there; it simply returns a different answer.

That is the point of using a test rather than a rule.

Ask which number moved and when, work backwards from your own figures, sort changes into act, measure, or ignore, and keep the record.

What to do

1. List what you were told mattered.
2. Ask which number moved.
3. Work backwards from your figures.
4. Find the months that moved.
5. Look for local causes first.
6. Sort into act, measure, ignore.
7. Keep the marked list.

Step three is the direction that works, since a year of announcements produces a long list of maybes and a year of your own figures produces two or three months that genuinely moved.

A prior version of this argument is at the update that did not affect you.

Frequently asked questions

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.