Four things I learned about Search Engine Optimization for publishing
A Dispatch from the Content War
Who are you and what are you doing here?
My name is Josh Gee, and I was the Director of Audience Development at Boston Globe Media Partners (BGMP). I drove strategy and execution for search optimization, email (newsletters and marketing), paid traffic, and business development/partnerships for boston.com,bostonglobe.com, and their constellation of awesome smaller sites.
I helped build this, and this was my baby. In short, my grandmother was very happy that I had a job in marketing of some sort at a big company. About 6 months ago, I published a post about email that a lot of people liked. It wasn’t as good as my former colleague Matt’s post about social, but hey, he’s wicked smart.
I didn’t know much about search when I started with BGMP. In fact, I remember saying in my interview, “I don’t know anything about search.” Of course, 4 months later they said, “Josh, you’re in charge of search now so figure it out.” Through a ton of homework and learning from people smarter than myself, I managed to make what I think will be significant SEO gains across the BGMP family of sites. There are two people/organizations who I should thank up front:
Hiring Shawna Wright, who came from an SEO consulting firm, was probably the best thing I did during my time atBGMP.
The second best thing I did was bring on Define, an Audience Development consulting firm (specializing in search) who were some of the best consultants I’ve ever worked with.
Lots of what I’m about to share I’ve learned from them. So let’s go.
What is SEO?
Search Engine Optimization really involves two things:
- Making sure your site is built so that the massive programs that crawl the Internet discovering and indexing pages for sites like Google can easily digest and understand your content.
- Creating content that matches with what people are looking for using sites like Google.
1. Search still matters
You don’t read glowing profiles of search-optimized news startups. Even the Huffington Post, which grew to dominance based largely on search, is focusing more energy on social. Social is huge and only getting bigger, but it wasn’t until summer 2015 that it passed search as the top traffic driver to media sites. Just because social is rising, doesn’t mean search is falling.
It’s not a zero-sum game. You can’t ignore something like search just because everyone is asking you about Snapchat. When Millennials want to dig deeper on a subject, 57% say they use search compared with 7% that say they use Facebook. Search can still be a huge driver of traffic. Check out the chart below, which is based on an anonymized aggregation of over 100 sites across a number of categories. In scale, it represents billions of visits and even more billions of page views.
At BGMP we saw about twice as much traffic from social as from search, but BGMP’s search engine optimization left a lot to be desired. We had a ton of reputation juice but a lot of things to fix, which meant lower-than-average search traffic. But even with the work that was needed, we saw a ton of traffic.
Even if search is losing to social, it’s still huge. If search is driving 35% of your traffic, are you giving it 35% of your attention? Probably not. Luckily, search is much more established than social so it’s easier to know what will work.
2. Ignore *your* search results. Pay attention to your site’s search traffic
Some variation of this should be printed above the desk of every Audience Development / Social staffer in the world. Not so much for them, but for business leaders who amble over and ask “Why isn’t this story at the top of Google News results?” Looking at your results and freaking out is basically the same as looking at your Facebook feed and saying, “Why isn’t our story showing up when I log into Facebook?” It’s missing the forest for the trees.
Not having a specific story at the top of Google News isn’t a crisis for search; a prolonged decreasing in your overall traffic is.
Many factors go into why a specific story is or isn’t at the top of a specific Google search, including an individual users search history and plain dumb luck. Search optimization is like the Arnold Palmer line, “The more I practice, the luckier I get.”
Focus on what you can control and let go of what you can’t.
This guy knows what I’m talking about.
3. Focus on technical and structural improvements before editorial
When I first dove into SEO, I assumed the majority of the task would be editorial. Some collection of “More tags, fewer tags, and using ‘Patriot Tight End Rob Gronkowski’ instead of ‘Gronk’.” Most people I worked with on both the newsroom and the business side had similar assumptions. Search has been around a long time and has changed a lot. Everyone has some training in it and lots of weird, apocryphal knowledge is still kicking around.
It quickly became clear that the biggest benefits would come from focusing on foundational technical improvements, not tags or content. Over and over I was reminded of a scene in The West Wing, where a character talks about racing sailboats. Every minor tweak is in the service of going just a little bit faster.
We spent 3 months on technical improvements before even thinking about editorial, and we should have waited even longer. I’ve read about traffic increases of 1000% before a single editor changed anything.
For publishers, dynamic XML sitemaps should be your top priority. Search engines crawl media sites faster than other kinds of sites(I think), but still not as fast as we publish content. A good XML sitemap pings search engines every time you post a new piece of content, ensuring that places like Google Search and Google News pick up your stories as soon as they are up. You can find out more about building sitemaps here and here.
Google’s admin panel is also getting a lot better about alerting you to crawl errors. More than half of optimization is making sure you’re not being penalized, which can be a major piece of kelp (it’s from the video) and drag down your whole site.
Focus on your technical improvements before you start tackling anything else. The great part is that it’s called optimization. Get to a point where you’ve done what you think is worthwhile and then you can move on.
4. Search Content is helpful content
Search content gets a bad rap. There is a lot of bad search content out there which makes some people loathe to consider it. But it doesn’t have to be that way. Search content’s not always going to be sexy, but it can be helpful to the reader. And with a good strategy in place it can be helpful to your outlet with a base of reliable traffic.
This is my favorite piece of search content:
I probably encounter this article 2–3 times a year and always the same way: I’m running to the airport. I have an hour and a half until my flight. It’s not enough time to stop on the way, but I wouldn’t mind a beer and a quick snack. I take out my phone and Google “Where to eat at Logan” and this pops up.
It wasn’t until the third time I saw it that I realized a few things this article does really well.
It’s headline matches what people are actually searching for — The first piece of advice I got about search back in 2009 is probably still the most useful: “People don’t search aspirin, they search headache.” A more typical newspaper headline for this story might be “A culinary tour of Logan” or something else that no one would ever actually look for. That works when competing for attention in the daily churn, but won’t last longer than a day or two. Meanwhile a story like this one from Boston.com, which did OK when it launched, grew and grew from early October right through election day. It was boston.com’s #2 story by search traffic during my time at BGMP and the biggest story of that month.
“People don’t search aspirin, they search headache.”
This story has a long shelf life — Stuff on search engines lasts forever. The story about where to eat at Logan, because it matches how people actually search, will get discovered again and again and again. It probably didn’t do great when it launched. It might not have even gotten homepage or social play. A catchier headline probably would have done a lot better. It’s not about that.
At boston.com, this was our top story by search traffic during my time there. It did nothing for us during 11 months of the year, but, you guessed it, it dominated the site during October. It had been doing that for years before I even started.
How hard would it be for your team to come up with a hundred stories like this which fit your editorial brand and voice? Or start with just ten? Then how hard would it be to update them once or twice a year to make sure they stay relevant? It’s not glamorous but it would bring in a solid base of traffic. And the more folks are finding your content that way, the higher your breaking news and other stories will feature across search sites.
If you’re not sure how to title stories or what to write about, tools like Google Trends, Google Keyword Planner, or Buzzsumo can give you a sense of what people are searching for and what words they are using to do it. Moz.com also has some good free tools for seeing if you site is optimized.
Once you’re a little better at search, start trying to figure out search terms in real time and revise your breaking news stories to give them more lift based on how people are searching for that news.
SEO content doesn’t have to be scummy, and it doesn’t involve cramming your headlines with as many terms as possible to try and trick the engines. Create pieces of content that help people answer questions they are asking, and then title them using the terms people use to ask those questions.
tl;dr: Search matters and you’re probably not paying enough attention to it. Ignore individual results and look at the overall traffic. Focus on technical optimization — remember the sailboat. Write stories that answer questions and use headlines that reflect how people actually search for things.
Editor’s Note: I drafted this last summer, then left it in drafts for like 6 months while I went backpacking and started a new job working on this. I found it and realized that it’s still pretty relevant. I updated it with new data, but if any seams show, that’s why.