This is a post made of 4 parts:
1. Outreach elements we test (and recommend others test in their own campaigns)
2. Citation Labs experiment examples
3. The Citation Labs Best Practices Record
4. A webinar conversation between Garrett French and Megan Hannay on Citation Labs Outreach Hacks.
The first rule of outreach is that there are no rules.
Except for one:
Every outreach effort needs to be accompanied by a value proposition. This goes back to quality content creation and audience understanding: provide something of value to the right people. (Rand Fishkin did a great Whiteboard Friday on this topic recently as well.)
We’re calling this piece a “record” instead of a “guide” because we’ve learned that most outreach advice is only effective in a vacuum. Change one variable and results may alter with it. Instead, we recommend that you take all results with a grain of salt and experiment on your own.
The variability of this area of link building is also why we’re including a list of our experiments within this post. Sometimes the best way to teach is by example. Hopefully, our examples will teach your team to think of your own (and please share them in the comments!).
Outreach is a Moving Target
Outreach is a continuous process; it ain’t over when the link is live, or when the “thank you” has been sent. Effective outreach, especially when we’re talking about enterprise campaigns, involves continuous trial and error. In a link building laboratory, outreach is where we get to have the most fun with data.
With outreach, the only way to deliver success is to be open to variability.
We can test:
- Order of information shared in the email
- do we start with an interesting statistic or an informative resource?
- Email length
- Personal details of outreacher
- commonalities, books read recently, mailing address
- Personal details of recipient
- do we refer to their name, location, company, etc?
- Subject Line
- this is probably one of the most important to TEST, TEST, TEST
- The relationship between the subject line and the first sentence
- does one ask a question and the other answers it? Vice Versa?
- Misspellings & grammatical errors
- can dumbing things down work?
- Timing – time of day, day of week, sending in relation to current events
- Ways of demonstrating social proof of resource or outreacher
- “I’ve been published in …” or “Even President Obama thought this was good enough to retweet”
- —-> Note: this not an area for hyperbole
- The descriptions of your pitched resources
- anchor text versus raw URL?
- how do we explain what we are sending?
- Accompanying resources
- do we pitch a single or multiple URLs?
You can measure:
- Open rates
- Response rates
- Link rates
- Traffic to pitched resource(s)
- Resources chosen from pitch
- when pitching multiple resources – if the recipient links to Resource B, C and D, but not A, we know that A needs some work
And this goes for any new vertical or campaign. As you’ll see below, some of our experiments contradicted each other, or popular industry advice, likely because different outreach variables were at play. Remember “there are no rules”?? This is why:
11 Citation Labs Outreach Experiments
Every campaign is a new opportunity to experiment. Here are a few of ours.
For some of the below, we’re able to share the stats from our experimental campaigns. Keep in mind – these are isolated campaigns on particular topics, to certain audiences. Even most “what worked” and “what didn’t work” guidelines have yet to be solidified into Citation Labs “best practices.”
What Worked (for Us)
1. Asking questions in the subject line, or a creating a subject line that compelled the reader to ask a question.
“3 in 5 people eat too much sugar every day” makes one think “Am I eating too much sugar?”
2. Including location-specific details in the subject line or the first few sentences of a high-touch email.
3. Shorter emails.
4. Finding more personal details on the site and mentioning them in the email (“personalization”).
5. Sharing a story in the pitch that explains the mission behind the email.
6. Including in the template “I’d be happy to get on the phone with you.”
Resulted in fewer annoyed or skeptical responses.
7. Mentioning the client you’re emailing on behalf of.
When you’re outreaching, and there’s no clear benefit to you (the person sending the email), people can get suspicious.
In a sharing economy outreach to potential local sponsorship partners, we were 2x as likely to receive a response when we mentioned the name and a blurb about our client in our email.
8. Including an ask-for-help option in outreach.
For some emails, we experimented with mentioning a “ask for help” page on the website we were reaching out from. The thought was that asking our recipients to voice what they wanted, content-wise, would build trust. The results were a bit strange:
No mention of the “Ask for Help” page led to 2x as many results, but only 62% as many links, as mentioning the “Ask for Help” page in our pitch.
What Failed (for Us)
1. Withholding – i.e. offering a statistic or two, but not the entire resource in the first email; instead, asking if the recipient would like to see the resource.
Many of our industry peers have had success with this tactic, so we tried it, but never saw success in terms of link count.
In our experiment, withholding led to 2x as many responses, but only 1/3 as many links.
2. Putting the resources before the bulk of the email.
Another industry tactic we’d heard of, but when we tried it, we received a lower response and link rate.
3. Planning outreach based on the gender of our emailer.
For a long time, we listened to popular advice that a female outreacher is better. But in one Citation Labs experiment, we found that a male outreacher received more responses.
For Autism-based outreach in the US, the outreacher with an obviously male name received 2x as many links as an outreacher with a gender neutral name.
But then! In another experiment, a few months later, we found that a female received more responses.
For eco/environmental outreach in France, the woman outreacher received 3x more positive responses and 13% more links than the man outreacher.
This final example is a great synecdoche for its ten predecessors. Maybe it was the topic, or the country, or the timing, but we’ve found that there’s no perfect “best practice” when it comes to setting up the details of our outreach campaigns.
However, there are some guidelines we’ve developed over the years that have worked in enough campaigns to graduate from a topic or audience-specific success to a Citation Labs agency rule of thumb. But you’ll notice that most of them, unlike our “what worked” and “what didn’t work” learnings, are higher-level approaches instead of one-off tactics.
Citation Labs Outreach Best Practices
For this piece, I’ve interviewed Citation Labs team outreach experts and managers to compile our agency’s best attempt at a resource for link building and PR outreach advice. (Thanks to Valerie French, Andrea Pretorian, Kristina Welch and Garrett French!)
- Over-qualify. When prospects don’t fit into a particular outreach list, we place them into a topic-oriented list, and as soon as we find X number of linkers who cater to [Topic Y] audiences, we’ll do a [Topic Y] outreach campaign based on [Topic Y] content we’ve created for this audience. This allows us to plan for content we know linkable audiences desire and to be very specific with our outreach.
- Don’t leave template writing to a junior member of the team. Experiment ideas come with experience.
- At the same time, it’s easy to burn out in this work. Rotate copywriters’ projects so they get varied experience and can keep outreach fresh.
- Justify the email – this is much easier if you’re writing with a company mission in mind.
- Relate to the person you’re emailing with a “we” moment before breaking into an “I” statement. (For this one, we took some structure tips from this reddit post on speechmaking.)
- Weird characters, such as apostrophes, dollar signs, percent signs, and even certain words that you wouldn’t think of as harmless can set off spam triggers when used in your subject line.
- A no doesn’t have to be a “no.” Sometimes you can turn it around by getting on the phone with a linker or blogger (see #6 under “what worked), or by answering their questions promptly and thoroughly.
- Have people in the inbox every day, sending responses to create that relationship. If we placed something heart health related and then develop another heart health content piece to pitch, we have the rapport to reach out again.
- Don’t consider an outreach campaign “done” until a month after the last email’s gone out. People will respond either in the first 48 hours, either in 1 week or in 2 weeks. And then you get the really weird people who respond 6 months later. Most of the links we’re going to get will go up within 2 weeks, up to a month. Set these expectations with clients and executives upfront.
Outreach is not the last step for a link builder. It’s the last five steps. Or eight or ten steps. Experimenting is many people’s happy place, but the success of most individual campaigns (and therefore the bread & butter behind the experiment wheel) often comes down to the number of links developed. Having enough bandwidth to do what works, but having enough left to experiment is a balance that deserves some trial and error in its own right.
8 thoughts on “11 Email Experiments + The Citation Labs Outreach Best Practices Record”
Great post/thanks for sharing those insights. Had a few questions for you:
1) When doing the two-step approach, what kind of content were you pitching?
2) You mention highlighting the location. Is that when location is the primary piece of the content or not? For example if a city ranked #1 in a study or something like that.
3) Can you share more specific examples of subject lines that create curiosity? I get adding a question but that seems like it would be weird in pitch emails.. although I could be wrong. I like the idea of using stats that make one ask themselves a question though!
1) We were pitching informative written content in the health space. And just to clarify, we weren’t saying “withholding doesn’t work” – just that it doesn’t always work.
2) I think it can be for both. Obviously, it makes more sense if you’re reaching out with a Guide to Cleveland that you mention some things about Cleveland. But even if you’re reaching out in a high-touch way email with location-agnostic content, commenting on city landmarks, etc., especially when emailing people with a physical store or address, can be a way to personalize an email.
3) It could be as simple as using a … or posing the SL as an invitation. You want to be genuine and not too baity, but I think it’s okay take a lesson from Buzzfeed and step out of the box.
Great article Megan, certainly set off a few light bulb moments. I’m actually just testing ‘Withholding ‘ on a rather large campaign at the moment (not seeing a great placement rate!). Keep it up!
Thank you for sharing! I love seeing such specific tips from real experiments. Very useful.
Outreach is this odd world where it’s both hard to do & easy to overthink it. Great to be reminded that experimentation is key.
Megan, this is outstanding. The importance of the “psychology of outreach” is rarely discussed and this record is filled with it, while not specifically saying it. Even something as simple as the “mood” someone is in at the time can affect how they react to your email, and that’s beyond our control. We can “best practice” all day long, but at the end of the day, outreach is about people, and people are human, and humans are a fickle bunch. You could raise Hemingway from the dead to write your outreach and still hear crickets.
Thanks Eric! And yes, the “psychology of outreach” is definitely what we’re getting at here. Just as with traditional psychology – there’s not a one-size-fits-all solution to every problem.
Great list of actionable tips here! I noticed that you list mentioning the client you’re emailing on behalf of as best practice. I’m assuming that you’re not using branded company emails for your outreach? Just curious if there are any studies on the effectiveness of reaching out as a third party on behalf or a client, or directly as the client using a company email?
Good thoughts Brian! There are times when we’re working closely enough with our client’s team on a highly branded effort that it makes sense for us to have our own email. We have not been able to test sending email@example.com versus sending firstname.lastname@example.org, because usually those branded efforts are so sensitive, but I would bet serious $$ that @clientdomain.com has an advantage with certain linkers (bloggers, nonprofits maybe), while actually being a disadvantage for others (anyone who highly distrusts marketers).