Bad news for people who hate using the phone: Just 41% of U.S. news publishers “make it easy” for subscribers to cancel their subscriptions online, according to a new survey from the American Press Institute.

News organizations also vary widely in how and whether they identify or assist subscribers who are at risk of canceling. The blunt, “we’ll-just-make-it-hard-for-you-to-cancel” method remains more common.

Jeff Sonderman and Gwen Vargo set out to see which subscriber retention techniques are, and aren’t, in use at 526 news organizations across the U.S. (“they include publishers big and small, newspapers and digital-only, nonprofit and for-profit”). They found some retention tactics that were extremely common: 90% of publishers surveyed encouraged new subscribers to sign up for their newsletters, 89% use analytics to track what subscribers as a whole are reading, and 86% “track data about which digital content online users engage with.”

When it comes to cancellations, though, trends are less consistent. According to the survey, only 41% of those surveyed “make it easy for people to cancel subscriptions online, so they have an improved customer experience.” (It also helps if you live in California, or, at least, if you temporarily change your billing address to California.)

As a news consumer, you are less likely to be able to cancel online than you are to get a please-stay pitch from a customer service rep on the phone: 60% of news organizations surveyed “customer service reps [who] are trained in tactics for ‘saving’ renewals when customers ask to cancel” (60%).

It’s probably easier to make it hard to cancel than it is to figure out why people want to cancel and what you might have to do to try to get them back. Less than a third (28%) of publishers surveyed “segment [their] subscribers based on their risk of cancellation.” But when publishers do segment readers based on likeliness to unsubscribe, they have the opportunity to intervene:

The Arizona Republic realized that almost half of its paid digital subscribers were not visiting their website at all in a given month. They found that group accounted for 50% subscription stops each month. They used analytics to guide content changes that cut the share of unengaged subscribers from 42% to 26%, increasing retention as a result.

In our study, this strategy of identifying at-risk subscribers was one of the biggest areas where publishers could improve their tactics.

Publishers tended to rate their proficiency at identifying at-risk subscribers as “not very” or only “somewhat” proficient. More than half (52%) of publishers are “not very” or “not at all” proficient at identifying at-risk subscribers.

But 84% say it would be at least “fairly valuable” and 38% say doing so could be “very valuable.”

As you might expect, large newspapers, which generally have more resources, are better at identifying at-risk subscribers than smaller ones.

Large newspapers (with a circulation of 200,000 or more) all felt at least “somewhat proficient” at identifying subscribers at risk of cancellation. This difference supports the idea that access to technology and resources is a driver of this gap.

We asked about two specific tactics that publishers might use to learn about subscribers — identifying subscribers who are not engaged with your online content, and identifying those who are not engaged with your other products and platforms.

Fewer than half of all news organizations said they do either of these. …

The smallest newspapers (below 50,000 circulation) were even less likely to do these things — fewer than 20% of them employ either tactic of identifying disengaged subscribers.

An outlier in the other direction is the large-metropolitan newspapers, at which 100% said they identify subscribers who are not engaging online.

The full report is here.