UK Election Email Newsletters Rated

Since our earliest user research into email newsletters 8 years ago, we've known that newsletters are a superb mechanism for growing a relationship with customers. Indeed, when we recently asked users why they were visiting particular websites, the most common response was, "I was reminded to do so because I received an email newsletter from the site."

Beyond being good for business, newsletters are an effective means for political candidates to stay in touch with their supporters during an election campaign. (Twitter and Facebook campaigning might get more press coverage, but representatives from all 3 parties told The Economist that the old-fashioned newsletter is actually the more effective campaigning tool.)

I decided to check the state of newsletter design for the current UK elections. To do so, I evaluated newsletters from all 3 main parties: the Conservatives, Labour, and the Liberal Democrats.

I evaluated the subscription process on April 8, 2010, the unsubscribe process on April 21, 2010, and the newsletter content over the two-week interim (April 8–21).

I didn't evaluate the websites themselves, only the parties' use of off-web internet communication. Still, I can't help but point out that the Lib Dems' site violates a key navigation design guideline by using tabs that don't look like tabs:

Score Card

Following are the newsletters' average compliance ratings in the 4 guideline categories:

The overall usability rating is determined by averaging all our guidelines for newsletter usability (149 for this round). Thus, it's not simply an average of the 4 category scores, because there are different numbers of guidelines in each category.

Signup Process

When you're trying to build a mailing list, the worst mistake is to violate the guideline to remove distractions from the signup page. You almost captured the user's email — but they decided to go off and gawk at some of your Facebook fans instead.

Also, while consistency is usually good, it's actually bad to include the standard navigation's newsletter subscriptions element on the page that users land on when they click its Sign Up button. Doing so violates a standard application design guideline to avoid "no-op" commands that simply return users to the same state they're already in.

(Similarly, even though all other website pages should link to the homepage, the homepage itself shouldn't have a homepage link. Homepages that link to themselves were mistake #10 on my list of top homepage mistakes. Sadly, Labour and the Lib Dems both make this classic mistake — 7 years after I warned against it — but at least the Conservatives get it right.)

To its credit, Labour makes a half-hearted attempt to explain its privacy policy right on the signup page. Of course, this is one of the weakest privacy policies in history, simply saying that they "may use the data you have supplied." But the other two parties don't even bother to mention how they'll use your private data. (And no, it's not enough to have a full privacy policy elsewhere on the site. Users need to be told the policy on the spot when you ask for sensitive info.)

The header is Sign up for Email News — but that's what I just did! The header is followed by a slogan and some blah-blah text. Only at the very bottom do we get a modest line confirming that "your details have been submitted." Even here, they don't say that I've been subscribed or inform me of when to expect the first newsletter. Maybe they want to approve new subscribers before spending the 0.1 penny it'll cost to send those emails for the duration of the campaign.

Besides violating all the guidelines for newsletter confirmations, this page also violates several guidelines for writing for the web, including elementary mistakes like punctuation inconsistencies. Also, anyone who appreciates online typography is unlikely to be fond of the sudden switch to a serif typeface for that final line, even though it does make it stand out a bit on an all-sans-serif site.

Although it has the worst confirmation page, the Lib Dems won the confirmation email race, being the first party to send me a welcome message (beating Labour by 4 minutes). The Conservative party never bothered to acknowledge my subscription, even though its weekly newsletter would benefit more than daily ones from a good welcome message. After all, people might easily go 6–7 days without hearing a thing; at that point, some users will have forgotten they subscribed and consider the newsletter spam.

The Lib Dems' welcome message also has by far the best From field and Subject line, as this comparison shows:

• From: Liberal Democrats — Subject: Thank you for signing up for Liberal Democrat email news
• From: [email protected] — Subject: Thank you for signing up

Given that users are drowning in spam, it's important that your subject lines be as precise and descriptive as possible to encourage users to open your messages. For the same reason, it's also better to have a human-readable From field than to rely solely on a geeky-looking (and less readable) email address. Liberal Democrats is better than labourparty — at least typographically.

Publication Frequency

Following are the number of emails (not counting confirmation messages) that I received from each party during my 14-day evaluation period:

• Labour: 3
• Conservatives: 4
• Liberal Democrats: 9

The Conservative homepage asks users to sign up for "David's weekly email," so you'd think messages would be sent weekly. Not so. I received 4 Tory emails over 2 weeks. I don't think this constitutes overloading supporters during an election campaign, but it's certainly not weekly.

In contrast, I got only 3 Labour newsletters during the same period. This is definitely too little communication; people sign up for a party's newsletter to be kept informed about the election.

The Lib Dem party sent 9 emails, which borders on information pollution. This was especially true on the day it sent 3 emails within an hour, rather than consolidating the barrage into a single message.

In general, 2–3 emails per week is appropriate during a campaign's early phases, with the frequency increasing to daily updates during the final hectic days before the election, when subscriber interest intensifies.

American politicians make the mistake of courting supporter fatigue by spamming subscribers with far too many emails. British politicians make the opposite mistake: they communicate too little during a period when supporters want frequent updates.

All 3 parties in my analysis are making the (safe) choice to carefully limit how much email they send to keep people from unsubscribing. This is generally good, but I think they're perhaps overly concerned with the less-committed segments of their audience. This will disappoint the more-committed supporters, but perhaps the parties assume that their Twitter feeds serve that group.

Even though the parties comply with more social networking usability guidelines than newsletter guidelines, Twitter can't replace a good email newsletter. For one, the parties tend to tweet more or less hourly, which is overkill for anybody except political wonks. As a result, they have only 14,000 to 27,000 followers, which is nothing in a country with 61 million people.

The recommended solution to this dilemma is to offer 2 newsletters with different publication frequencies, but none of the parties do this.

Receiving Newsletters

The first newsletter I received from the Conservatives was sent with a From field reading "George Osborne," even though the website advertised the newsletter as "David's weekly email." (David Cameron is the Conservative candidate for Prime Minister; Osborne is the party's candidate for Chancellor.)

The general guideline is for newsletters to be sent from either a recognized institutional name or from a celebrity in his/her own name. In this case, "David's email" ought to be from David, not George. It's fine for political newsletters to feature multiple party leaders, but if they do so, they shouldn't be promoted in the name of a single celebrity.

Of the 3 newsletters, 2 feature prominent links to the parties' Facebook pages and Twitter feeds. This is recommended. When we did user testing of social network publishing, we found that users had great difficulty locating the official pages for companies and organizations. Of major online services, only the Apple App Store has more atrocious findability.

The Lib Dems newsletter is the most scannable and the only one truly designed for today's time-pressed readers. The Tory newsletter does follow writing guidelines, like the use of highlighted keywords, and also uses stills from (linked) video clips to break up the text, drawing users' eyes down the screen. In contrast to these rival offerings, Labour's newsletter is one big wall of gray, undifferentiated text.

To target a broad consumer audience, our recommendation is to write at an 8th-grade reading level. (Meaning that the text should have a complexity level suitable for a well-schooled 15-year-old.) Only the Lib Dems actually follow this recommendation, as the following readability statistics for sample newsletters show:

• Conservatives: 10th grade
• Labour: 9th grade
• Liberal Democrats: 8th grade

Does it matter that content from the Conservatives and Labour is too complicated for 43% of the population? Probably not much; people who follow politics so closely that they'll subscribe to a party newsletter probably belong to the elite 57% of the population who can read the text easily enough. At least none of these newsletters were written at the university level — a style characteristic of many government and big-company websites (and even my own articles, which tend to be written at a reading level equivalent to undergraduate university texts).

Subject lines are uniformly bad. Here's a sample. Try to guess (a) which party wrote which headline, and (b) which (if any) of the following emails you'd open:

• One simple word
• Answer time
• Can you help make it a fair fight?
• Jakob, my take on week two of the campaign
• State of the Race memo 3

At least the last two (from the Tories and Labour, respectively) have some small information scent to indicate that they're about the election. Still, it would have been better to add some actual content to summarize the analysis and enhance the "open-me" attractiveness of these subject lines.

Following are two of the better subject lines, both from the Lib Dems:

• Spread the word — The tax cut you can believe in
• Help elect a Lib Dem MP

The first of these contains specific content about a topic of interest (taxes). Sadly, however, it defers the information-carrying words to the headline's second part, leading instead with off-putting blather. (I'll decide for myself whether to spread the word.)

The second subject line was on the only personalized email I received during the entire 2-week study and referred to a specific, closely fought race of interest to the user profile provided during the sign-up process. (In contrast, it's not true personalization to just mail-merge a user's name into a headline; that's presumptuous, not personal.)

Who Will Win?

Last time I scored the usability of campaign newsletters was in 2004, when George W. Bush won with a usability score of 58% compared to John Kerry's 57%. In addition to winning my review, Bush also won the election.

(You can't directly compare those 2004 usability scores with the scores in this article, because I relied on an earlier version of our newsletter research to conduct the 2004 review. We now know much more about email usability, so a good rating is more difficult to achieve. Considering that high marks are harder to get now, it's a nice sign of usability progress that the winning score has increased from 58% in 2004 to 63% in 2010.)

In 1996, I evaluated Bill Clinton vs. Bob Dole for usability. Even though my initial score was the same for both candidates, Clinton followed all of my recommendations within 2 weeks, whereas Dole never bothered to do what I told him. Thus, for most of the campaign, Clinton's site had superior usability. And he won that election.

Given this track record of usability scores predicting election outcomes, will the Conservatives win the UK election on May 6? Not necessarily, because much can happen from April to May. Also, the internet probably influences only 1–2% of an election's outcome. In a close race (such as Bush vs. Kerry), good usability might move the needle enough to determine the winner.

If the race is tight in May, then maybe the Conservatives will win because of their superior usability. But if Gordon Brown pulls a last-minute rabbit out of his hat or if David Cameron goofs in the last television debate, the outcome could easily be reversed, because such events can move 5% or more of the electorate. Old media rules in the end. But that's no reason to ignore email, especially since it's a much cheaper medium.

[Update: The Conservative party did win the 2010 UK election, but not by enough to gain a majority in Parliament. They formed a coalition government with the Liberal Democrat party. Thus, the two parties in the new government were those with high usability scores, whereas the losing party was the one with a low score. I still don't claim that UX won the election, but it probably did influence the result.]