The question I get most
"How do I know if this is working?"
It's the right question, and the answer is less exotic than people expect. You measure geofencing the same way you measure any digital advertising, plus one thing you can't do anywhere else.
Geofencing is targeting, not a separate universe
Geofencing is the capture of mobile device IDs at physical locations. That's all it is. Once those devices are captured you've got an audience, and you're serving that audience traditional display, video, and social ads.
So the campaign you're measuring is a display campaign and a social campaign. The only unusual part is how the audience was built. There's no special geofencing metric you've been missing. There's standard measurement, done properly, and then one bonus.
Impressions, clicks, engagement. A health check.
Traffic, form fills, and calls.
Conversion zones. Only possible with geofencing.
Layer one: the ads themselves
Impressions, clicks, click-through rate, ad engagement. The normal set.
This layer tells you delivery is happening. The audience was large enough, the creative is being served, the flight is pacing. It's a health check, not a performance measure, and that's where a lot of people get stuck.
Why click-through rate is the wrong number to judge this on
Display click-through rates are low. Everywhere, for everyone, always. That's the medium, not your campaign.
More importantly, clicks aren't how display works. Somebody sees your ad on a health site while reading about their mother's diagnosis. They don't click. They see it again four days later inside a game, then a third time on social. Two weeks after that, when they're ready, they type your company name into Google and go straight to your site.
Zero clicks on your display campaign. One conversion that the display campaign caused.
If click-through rate is your scoreboard, that campaign reads as a failure. It wasn't. Judge the medium on whether business happened, not on whether somebody tapped a banner.
Layer two: what happens on your site
This is the layer that actually matters for most businesses. Did the traffic show up. Did they fill out a form. How many pages did they see. Did they call.
One serious caveat, and it trips up almost everyone. A lot of mobile traffic from display campaigns arrives at your site without a clean source. Google Analytics doesn't always know where it came from, so it dumps it into the "direct" bucket, where a lot of unrelated traffic also lands. If that's all you're looking at, a working geofencing campaign can look like it produced nothing. The traffic is there. It's sitting in a bucket with everything else.
Two fixes, and neither is optional.
- UTM parameters on every geofencing ad. This tracks the traffic down to which fence or which creative produced it, instead of watching it disappear into "direct."
- Call tracking. If your business gets business by phone, and most local businesses do, the calls are the conversions. Without tracking numbers you're guessing.
Set both up before launch. Retrofitting attribution to a campaign that already ran is how you end up with a month of data nobody can defend.
Layer three: the one you only get with geofencing
Because we're targeting mobile devices tied to physical locations, we can put a conversion zone around a brick-and-mortar location. Your store, your dealership, your community, your office. Then we track how many devices from your fences later showed up inside that zone.
Read that carefully, because there's a detail in it that surprises people: this includes devices we never served an ad to. If we saw a device in a fence, we can track it into the conversion zone whether or not we won an impression on it.
For a retail business this answers the only question they actually care about. Is this bringing people into my store? Not clicks. Not impressions. Feet through the door.
It also produces market intelligence as a side effect. Fence competitor locations, put conversion zones on your own, and you learn how people move between businesses in your market during a shopping process. I did exactly that for a jewelry retailer across thirty-plus markets.
Month one versus month three
Different questions at different points, and confusing them is the most common way a good campaign gets killed early.
Are we hitting the forecasted impressions. Are the fences producing the device volume the strategy deck projected. Is the traffic arriving and tagged properly. Are calls being tracked. Don't judge results here. You're confirming the machine is running.
Which fences produce traffic and which don't. Which creative gets engagement. Whether form fills and calls have moved off baseline. This is where you cut the fences that aren't producing and shift impressions to the ones that are.
Enough frequency has accumulated for people to recognize you, and enough time has passed for a considered purchase to move through its cycle. Now you can compare conversions, calls, and walk-ins against where you were before.
For long-consideration categories like senior living, even three months is early. A year-long decision cycle doesn't resolve on a quarterly reporting schedule.
What a monthly report should actually show you
If a report you're getting doesn't contain these, ask why.
- Delivery. Impressions served against forecast. Devices reached.
- Performance by fence. Which locations produced traffic. The single most actionable thing in the report, because it tells you where to move next month's budget.
- Site behavior. Sessions from the campaign, pages per session, time on site.
- Conversions. Form fills and tracked calls, by source.
- Walk-ins, if you've got a physical location. Conversion zone visits.
- What changed and what's changing next. A report with no decisions in it is a receipt, not a report.
What geofencing can't tell you
Being honest about the limits is how you know the rest of the numbers are honest too.
- Who anybody is. Anonymized device IDs. No names, no phone numbers, no personally identifiable information. I can tell you 400 devices from your fences visited your location. I can't tell you which 400 people.
- Perfect single-source attribution. No channel gives you this, and anybody who claims it is measuring badly. Somebody who saw your display ad, then your social ad, then searched your name, then called, gets credited to "phone call" by most systems. The display ad started it.
- Immediate results in a long-consideration category. If your sales cycle is nine months, a 30-day campaign hasn't had the chance to produce a sale yet. That isn't a measurement failure, it's a timeline.
The real test
Strip it back and there's one question: is the phone ringing more than it was.
Traffic and impressions and click-through rate are diagnostics. They help you figure out what to fix. They aren't the point. If the calls, forms, and walk-ins are up and holding, the campaign is working. If they're flat after three months of clean delivery and proper tracking, something's wrong and it's worth finding out what. Here's where campaigns usually break.
What would be measurable in your situation?
Find out before you spend anything. About 15 minutes, no pitch, and the strategy deck is yours either way.