How a UTM parameter works
A web address can carry extra instructions after a question mark. A UTM parameter is one of those extras, and it exists purely so the receiving website can be told where the visitor came from. The page itself ignores it; the analytics code on the page reads it, stores the values against that visit, and the visitor sees no difference at all.
There are five recognised parameters. Source names the origin, medium names the kind of traffic, campaign names the activity, and term and content are optional extras for keyword and creative detail. A sixth, utm_id, exists for joining analytics data to cost data imported from elsewhere. Nothing enforces the values — analytics records exactly what you write.
The name comes from Urchin, the analytics product Google bought and turned into Google Analytics, which is why the letters no longer stand for anything useful.
Why UTM parameters matter
Without them, analytics has to guess. A click from an email client, a QR code, a messaging app or a PDF often arrives with no referrer at all, so it lands in direct traffic next to people who typed your address from memory. Everything you spent effort on becomes indistinguishable from everything that happened by itself.
Tagging removes the guesswork for the channels you control. It is the difference between a report that says most visits came from somewhere unknown and a report that tells you which newsletter, which post and which partner actually sent people. For a business in Nepal where a large share of links are shared inside Facebook and messaging apps, that gap is usually the biggest single hole in the data.
Common mistakes with UTM parameters
The first is inconsistency. The values are case sensitive, so Facebook and facebook are stored as different sources and split one row into two. Agreeing lower case and hyphens costs nothing and prevents most of the mess.
The second is tagging internal links. Putting a UTM on a link between two pages of your own site restarts the visit and rewrites where it came from, so your own navigation starts stealing credit from the channel that really sent the person.
The third is doubling up with automatic tagging. Google Ads and several other platforms already pass their own data; adding manual tags on top can override it. The fourth is putting anything personal in the values, since a tagged URL gets copied, shared and written into server logs.
How to act on it
Write a short naming convention before you tag anything: fixed values for medium, a predictable pattern for campaign names, lower case throughout. Keep it in a shared sheet so everyone builds links the same way, and generate every link through the UTM link builder rather than by editing a URL by hand.
Then check the results after a week. Open your traffic reports, look for two rows that clearly mean the same thing, and correct the links that created them. Tagging is not a one-off task; it is a habit that quietly decays unless someone reviews it, which is a large part of what ongoing analytics and tracking work involves.