Category: Analytics

This looks like a small thing. It isn’t.

When a newsroom posts a link, it is making a claim. The claim is: this goes where we say it goes, and we stand behind what’s on the other end. That claim is the entire reason anyone clicks a news link rather than ignoring it.

A generic short link quietly removes the evidence for the claim. The reader sees a domain that belongs to a shortening company, a slash, and six random characters. Nothing in it says which newsroom sent it. Nothing says where it lands.

Ten years ago that was fine, because nobody was worried. It’s less fine now.

The provenance problem

Readers have spent the last several years being told, correctly, to be careful about what they click. Phishing training at work. Warnings about misinformation. A general sense that a link from an unfamiliar source is a small risk.

All of that training points the same direction: look at the domain before you click.

So a news organization that distributes through a generic shortener has put itself in an odd spot. It has asked readers to be careful about provenance, then handed them a link with no provenance in it. The reader who follows the advice is the reader least likely to click.

This lands harder on newsrooms than on most businesses. A retailer’s link that looks slightly sketchy costs a sale. A newsroom’s link that looks slightly sketchy costs credibility, which is the whole product.

What a branded domain actually does

A branded short domain is your own domain doing the shortening. The Seattle Times uses st.news. WBUR uses wbur.fm. WLRN uses wlrn.us.

Look at those for a second. Each one answers the provenance question before the reader has to ask it. You know who sent it. You know it’s a news organization. You know it isn’t a stranger.

The link is the same length. The click is the same click. The difference is entirely in what the reader can verify at a glance, and that turns out to be most of the decision.

There’s a second benefit that matters more over time. You own the domain, so you own the redirect. If you ever leave your shortener, the links keep working because you point them somewhere else. Newsrooms have archives measured in decades. A link that dies when a vendor contract ends is a small piece of the archive going dark.

What to look for

Most shorteners treat a custom domain as a premium feature, which is worth knowing before you start comparing. A few things separate the useful setups from the ones that will annoy you in six months.

Check whether the plan actually includes a custom domain, and at what tier. This is the most common place people get surprised, because the entry-level paid plan often doesn’t include it.

Check what happens to your links if you stop paying. Ask directly. The answer should be that you keep the domain and can redirect it yourself.

Check whether anything gets inserted between the click and the destination. Interstitial pages and ad units on a click path are common on cheaper tiers, and on a newsroom’s link they read as an endorsement you didn’t make.

Check what the analytics actually report. Country and city are standard. If your coverage area is smaller than a city, and for most local newsrooms it is, city-level data won’t answer the questions you care about.

The part most people skip

Getting a branded domain is the easy half. Most newsrooms that set one up stop there, and then use it for years without ever looking at what it’s telling them.

The domain solves trust. It doesn’t solve measurement. You still end up with a click count and no sense of where those clicks came from, which for a newsroom accountable to a specific place is the more interesting number by a wide margin.

If you’re going to do the work of setting up a branded domain, it’s worth picking a tool that reports geography underneath it. Otherwise you’ve improved how the link looks and learned nothing new about who’s on the other end of it.

blrb.ai supports custom aliases and reports every click at the zip code level with Census data attached. Built for people who need to know where their audience actually is.

Every newsroom I’ve talked to can tell me how many people clicked a story. Almost none of them can tell me where those people were.

That gap sounds small. It isn’t. For a local newsroom, geography is not a detail attached to the audience; geography is the audience. A station licensed to serve a metro area, a paper that covers two counties, a statewide network answering to a board: all of them are measured on reach across a place. And the number they have in hand describes volume, not place.

So you end up in a strange position. You can prove a story did well. You cannot prove it did well in the part of your coverage area you were worried about.

The question a click count can’t answer

Say you’re a station in Boston. You run an investigation into housing conditions. It does 4,000 clicks, which is a good day.

Now answer this: did it reach Dorchester?

You don’t know. The 4,000 is one number covering an area with wildly different neighborhoods, incomes, and housing situations. It could be 4,000 clicks from Cambridge and Brookline, which would mean a housing story reached the people least affected by it. That is not a small distinction. That is the difference between covering a community and covering a community’s problems for a different audience.

The same question shows up everywhere, just wearing local clothes. In San Diego it’s North County versus South Bay. On Long Island it’s Nassau versus Suffolk. In South Dakota it’s Sioux Falls and Rapid City versus the other sixty-plus counties. A statewide network can post a strong click number while reaching almost none of the state, and nothing in the dashboard would tell anyone.

Why the standard tools stop short

Most link shorteners report country, sometimes city. City is where it usually ends, and city is too coarse for a newsroom. “Boston” is not one audience. It is dozens.

Web analytics has the same ceiling for a different reason. It tells you about traffic that already arrived on your site. It’s quieter about the link you put in a newsletter, a push alert, or a social post, which is exactly where most newsrooms distribute now.

Neither is a bad tool. They were built for marketers who care about conversion volume, and for that they work. A newsroom is asking a different question, and it’s a question about place.

What zip-code-level data actually shows you

Zip codes are the useful unit here for a boring reason: the Census publishes detailed data at that level, and it’s free and public. So once you know the zip, you know a lot more than the zip.

Attach Census figures to a click and you can see median household income, homeownership versus renting, median home value, and urban, suburban or rural share for the places your stories actually land.

Three things fall out of that, and they’re worth separating.

The first is coverage equity, which is the one most newsrooms care about immediately. If your housing coverage reaches homeowners at three times the rate it reaches renters, that’s a finding. It may change what you cover, or how you distribute it, or which platforms you use. You cannot act on it until you can see it.

The second is distribution, which is more practical. If a story performed well in one county and vanished in the next one over, that’s usually not an audience problem. It’s a distribution problem, and distribution problems are fixable in an afternoon.

The third is the reporting you owe other people. Boards, funders, underwriters, and grant officers all ask some version of “who are you reaching.” A map is a better answer than a total. It’s also a harder answer to argue with.

A caution about precision

Two things are worth saying plainly, because the vendors in this space are not always careful about them.

Click geography comes from IP address, which is accurate at the zip and metro level and unreliable below it. Anyone selling you household-level precision from a click log is selling you something that doesn’t exist. Treat this as neighborhood-scale data, because that’s what it is.

And Census figures describe an area, not a person. If a click comes from a zip with a median income of $80,000, that tells you about the place. It does not tell you the reader earns $80,000. The distinction matters when you’re writing up findings for a board, and it’s the kind of thing a careful analyst says out loud before someone else says it for them.

Used with those two caveats, the data is still far more useful than a click count. It just isn’t magic, and nobody should pretend otherwise.

Where to start

Pick one story. Ideally one where you already have a suspicion about who did or didn’t see it.

Put the link through a shortener that reports at the zip level, distribute it the way you normally would, and look at the map a week later. You’ll usually learn one of two things: your reach matches your coverage area, which is a good thing to be able to prove, or it doesn’t, which is more useful.

Either way you’ll know something you didn’t know before, and you’ll have spent about ten minutes finding out.

blrb.ai reports click data at the zip code level with Census figures attached, including income, homeownership, home value, and urban or rural share. It was built for exactly this question.