How to Use Google Analytics to Track Local SEO Performance
Start with the awkward part. This is the weakest of the three tools for local work. Its location data is approximate, the map results are invisible to it and the telephone call most local enquiries end in cannot be recorded. What it does tell you is worth having once you know that.
This Tool Is Poor At Local
Three limitations. None of them is a fault you can configure away. Location is inferred rather than known. Map and local pack activity does not appear here at all. And the outcome a local business actually cares about is usually a telephone call, which nothing on your website can see.
Location is worked out, not told. Approximate by design.
The tool estimates where somebody was from network information. On a mobile connection that can place a customer several towns from where they were standing.
The map results are absent. Completely.
Nothing about the local pack, the maps application or your listing appears in analytics. For many trades that is where the majority of local visibility sits, so a substantial half is simply missing.
The call is invisible. The big one.
Somebody reads your page, picks up the phone and books a job. Nothing in this tool records that. Block six covers what can and cannot be done about it.
Why we say this first. Credibility.
A guide claiming this tool measures local performance would be wrong in its opening line. Naming the limits is what makes the useful half worth reading.
What it is genuinely good at. After the click.
Which pages local visitors land on, what they do next and whether they reach the point of getting in touch. That is real information and no other tool provides it.
Where The Local Picture Actually Lives
Local visibility is reported in three places and no single one of them is the whole answer. Knowing which question belongs where saves a great deal of looking in the wrong report.
Your business listing. The map half.
How many people saw you in the map results, how many rang from there and how many asked for directions. That reporting has its own guide.
Search reporting. The place searches.
Which town and service combinations your website appeared for, plus whether anybody chose it. Our guide to using search reporting for local covers that, including why it also excludes the map results.
Analytics. What happened next.
The part this page is about. Once somebody has arrived, which pages they read and whether they got as far as contacting you.
Why they cannot be added together. Different counting.
Each measures a different event at a different point. Putting the three sets of numbers in one column produces a total that describes nothing.
What to do with all three. Assign the questions.
Am I visible on the map, am I visible in ordinary results and does my site convert the people it gets. One tool each. The business profile series covers the first properly.
Segmenting Local Traffic
You can isolate visitors the tool believes were in the areas you serve, then look at that group on its own. It is worth doing and it is worth doing with the coarseness of the underlying data firmly in mind.
What the segment does. Cuts everything else.
Once applied, every other figure describes only those visitors. Which pages they saw, what they did and whether they enquired, without the rest of the country diluting it.
Why it is worth the trouble. Local behaviour differs.
Somebody twenty minutes away behaves differently from somebody researching from two hundred miles off. Averaging them together produces a picture of neither.
The caveat that must travel with it. It is inferred.
The tool did not ask anybody where they were. It estimated, from connection information. Mobile networks in particular can be some distance out.
What that means practically. Region, not town.
Treat a county or broad region as reasonably reliable and a specific town as indicative. Building a service area plan on town level readings is building on something soft.
The one filter everybody should have. Your own country.
Before any of this, filter out other countries entirely. It costs nothing and it makes every other figure on the page more trustworthy.
Landing Pages Are The Better Signal
Which location pages people actually arrive on tells you more than any report of where visitors appeared to be. It is precise, it is not inferred and it describes what the market associates you with rather than where somebody's phone was.
Why it beats geography. It is a choice.
Somebody landing on your page about one town searched for something that led there. That is intent rather than an estimate of position.
What you usually find. Two pages working.
A business with a dozen town pages typically discovers that two receive nearly all the arrivals and the rest almost none. That gap is the finding.
What the empty ones are telling you. They are thin.
Pages differing only by a swapped place name do not attract arrivals, because nothing on them is about that place. Our material on risky practices covers why a set of those is a doorway arrangement rather than merely ineffective.
What to do about them. Fewer and better.
Consolidate the empty ones into the pages that work and write those properly, with the jobs you have done there and the conditions specific to the area.
The second reading worth taking. Did they act.
A town page receiving arrivals and producing no enquiries is a different problem from one receiving nothing. Block five is that distinction.
Measure The Action, Not The Visit
For a local business the visit is not the outcome. Nobody books a roofer by reading a page. The measurable thing is whether somebody took the step that leads to work. A property not configured to record that is answering the wrong question.
What counts as an action here. Four things.
A form sent, a number tapped, an email address clicked and a request for directions where your site offers one. Those are the moments a visitor became an enquiry.
Why visits mislead a local business. Small numbers.
A trade may get modest traffic and a healthy business from it. Judged on visits that looks like failure. Judged on enquiries it looks like what it is.
What configuring this changes. The question.
Rather than asking how many people came, you can ask how many enquired and from which pages. Setting up conversions covers doing it.
What most properties look like. Unconfigured.
The commonest state we inherit is analytics recording visits faithfully and nothing else. Every report from it describes activity and none describes results.
The half hour that matters most. This one.
Of everything in this cluster, defining what counts as an enquiry produces more than any amount of reading. It is also the thing nobody does.
The Phone Call Problem
A click on a telephone number can be recorded. An actual call cannot. That single sentence is the largest measurement gap in local marketing and it affects the trades more than anybody else.
What can be captured. The tap.
Somebody tapping a number on a phone can be recorded as an action, since it happens on your page. That is genuinely useful and it is not the same as a call.
What cannot. Everything after it.
Whether it connected, whether anybody answered, how long it lasted and whether it became work. All of that happens outside the website entirely.
The invisible group. The readers.
Anybody who reads the number off the screen and dials it manually. Anybody who writes it down. Anybody who rings later from a different phone. None of them exist in this reporting.
Why that matters for the trades. It is most of them.
For a plumber or an electrician the majority of enquiries arrive by telephone. So the tool is blind to the majority of the results while reporting confidently on everything else.
What can be done about it. Ask every caller.
Ask every caller how they found you and record it, which is unglamorous and works. Dedicated tracking arrangements exist and we do not name products. The business profile cluster carries our caution about the effect of using a different number in different places.
Comparing Areas
Businesses want to know which towns are working. This tool can support that question partially. The answers it gives need reading with the limits from block one attached to every one of them.
What you can conclude. Relative direction.
That arrivals to one town page are growing while another is flat is a reasonable reading, because you are comparing page data rather than inferred positions.
What you cannot conclude. Market share.
Nothing here tells you how many people in a town searched, how many chose a competitor or what proportion of the available work you are getting.
The comparison that misleads. Town against town by geography.
Reported location is too coarse for that, particularly between neighbouring towns. Two areas five miles apart are not reliably distinguishable.
The comparison that works. Page against page, over time.
Same page, this quarter against last, with a note of what changed. That is a clean test and it does not depend on the soft data at all.
Where the search side sits. The other tool.
Which towns you actually appear for is a search reporting question and a considerably better guide to expansion. Everything on this tool sits on the Google Analytics guide.
Your analytics
cannot see the
call that won
you the job.
A tap on a number can be recorded. The call itself cannot. Neither can anybody who read the number off the screen and dialled it. For a plumber or an electrician that is the majority of the work, which is why reading analytics alone makes a healthy local business look quiet.
How we measure a local business properly:
If your twelve town pages produce arrivals on two of them, the answer is fewer pages written properly rather than more.
Every guide.
One tool.
Getting it installed and configured properly, reading the reports without being misled, measuring outcomes rather than activity, the advanced reporting and what to do when the numbers look wrong.
