Printed website analytics charts and a notebook on a desk

Website Analytics for Small Businesses: The Four Numbers Worth Watching

Most small-business owners open Google Analytics once, see a wall of acronyms, and close it. Four numbers answer the only question that matters: is this website working?

Share this with the world

Website analytics is the practice of measuring what people do on your site — where they arrived from, which pages they read, and whether they took the action the site was built to produce. For a small business, almost all of the value sits in four numbers. The rest of the dashboard is there for people running paid media at scale.

Most business owners I work with have Google Analytics installed and have opened it maybe twice. It showed a line chart, some acronyms, and a number called engagement rate, and nothing in it answered the only question they actually had: is this website working?

That question is answerable. It just needs to be asked before you open the dashboard, not after.

Decide what counts as a win before you look at anything

Analytics cannot tell you whether your site succeeded. It can only tell you how often a thing happened. You have to name the thing.

For most small businesses it is one of four: a contact form submission, a phone call, a booking, or a purchase. Write yours down in a sentence before you touch the settings. If you cannot finish the sentence “this website is doing its job when someone ___,” no amount of measurement will rescue it.

The answer is rarely one thing. When I built the site for Trout Unlimited New Mexico, the statewide council coordinating chapters and conservation work across the state, a visitor could reasonably donate, join a chapter, sign up for an event, or take an advocacy action — four different outcomes, all of them real. The site’s structure came out of naming those four first. So did what we measured.

In Google Analytics 4 these are called key events, and you mark them yourself. Nothing is a conversion until you say it is.

Number one: how many people did the thing

Key events per month. That is the top number, and it is the only one that maps directly to revenue.

Watch it as a trend, not a score. Eleven form fills in September means nothing on its own. Eleven in September against four a month all summer means something. Eleven against thirty last September means something too, and probably something urgent.

One practical warning: a contact form that quietly stops sending email will keep recording key events perfectly. The tracking fires on submission, not on delivery. Test your own form from an outside address once a quarter.

Number two: where those people came from

Traffic acquisition, grouped by channel: organic search, direct, referral, organic social, paid. This is the number that tells you which of your efforts is earning its keep.

Direct traffic is the one people misread. It is not “people who typed my address from memory.” It is the bucket for every visit whose source could not be determined — clicks out of email apps, some messaging apps, links from documents, and any visitor whose privacy settings stripped the referrer. Treat a large direct share as missing information rather than a loyalty signal.

Number three: which pages are doing the work

Landing pages, ranked. Not all pages viewed — the page each session started on. That distinction matters more than it sounds.

Owners assume the homepage is the front door. Often it is not. A single service page or one old blog post routinely brings in more first-time visitors than the homepage does, which changes what you should be improving. If the page most people meet you on is a service page written three years ago in a hurry, that is your next project, and it is a cheaper project than a redesign.

Number four: what people typed to find you

This one is not in Google Analytics at all. Search queries live in Google Search Console, a separate free tool, and it is the more useful of the two for most small businesses.

Search Console shows the actual phrases people searched before your site appeared, how often it appeared, and how often it was clicked. It keeps sixteen months of that history, so you can compare this September against last September honestly. Analytics tells you what happened after the click; Search Console tells you what happened before it.

The useful pattern to hunt for: queries where you show up often and get clicked rarely. That gap is usually a page title and description problem, and it is a thirty-minute fix rather than a strategy.

What to stop looking at

  • Total pageviews. Traffic without outcomes is a number that goes up. Bots, your own visits, and a link shared once in a large Facebook group all inflate it.
  • Bounce rate, as a grade. In GA4 it is simply the inverse of engagement rate, and a session counts as engaged if it lasts ten seconds or more, views a second page, or triggers a key event. A visitor who reads your hours and drives over is a bounce.
  • Average session duration. Long sessions can mean people are absorbed or that they cannot find anything.
  • Real-time. It is the most watched report in Analytics and the least useful. Nothing you learn there survives the day.

How often should a small business check website analytics?

Once a month, for about fifteen minutes, is right for most small businesses. Look at key events against the previous month and the same month last year, check the top five landing pages, glance at the channel mix, and open Search Console for queries that gained or lost clicks. Anything more frequent turns normal weekly variation into a false alarm; anything less and you will not notice a broken form until a quarter has gone by. The exception is the two weeks after a site launch or a major change, where a quick daily look catches tracking that never got installed.

If Google Analytics is more tool than you need

It is free, it is thorough, and for a five-page business site it is heavier than the job. Two things to know before you commit to it: on the free tier, the detailed exploration data is kept for fourteen months at most, and because it sets cookies, it is the reason many sites carry a consent banner.

Lighter, privacy-focused alternatives exist. Plausible starts around $9 a month for ten thousand pageviews and Fathom around $14 a month, both cookie-free, both delivering a single screen you can read in a minute. For a business that wants four numbers rather than four hundred, paying a little to get only the four is a reasonable trade. Whichever you pick, keep Search Console — it is free and it has no substitute.

Start with the sentence

Finish “this website is doing its job when someone ___,” make sure that action is tracked, and check it monthly against the same month last year. Four numbers, fifteen minutes. That is a complete analytics practice for a small business, and it will tell you more than a dashboard you scroll through and close.

Patrick Iverson is a brand strategist and web designer in Santa Fe, New Mexico, running an independent practice here since 2001. If your site is collecting visitors but you cannot tell whether it is collecting customers, that is usually a structure problem rather than a measurement one, and it is worth a conversation.

Share this with the world
Patrick Iverson

Patrick Iverson

Brand strategist and custom WordPress developer, born and raised in Santa Fe. I've run an independent practice here since 2001, working with founders, marketing leads, and creative directors on rebrands, brand strategy, and websites built to last. I write here about the parts of that work clients rarely get to see.

Patrick's design sense opened up new possibilities for our website. His keen sense of matching client needs and skills to the available tools produced a site that we love working with. And his relaxed and responsive personal style made the whole project a pleasant experience.

Cecile LaBore, Recovery Systems Institute