SEO News

Posted by Dr-Pete

Last week, Google launched its latest feature, the “In-depth articles” block. Like News results or local packs, in-depth articles are a rich SERP element that sits in the left-hand column but doesn’t count as a standard, organic result. Here’s an example, from a search for “rainforest“:

We originally spotted in-depth articles in testing as early as July, and as of August 6th the feature officially went live for English queries on Google.com. Over the weekend, I re-tuned our MozCast 10K engine (which tracks a set of 10,000 queries and their features) to take a deeper look at in-depth articles. This post covers what we know so far.

Variations on a theme

All in-depth article blocks we’re currently tracking have three results – I’ve seen no exception to this rule yet, although that could change as Google collects more data and adapts. There are a few minor variations to how in-depth articles appear. Here’s a complete snippet, which includes an image thumbnail, title, description, publisher icon, publisher, and author (from a search for “presidential candidates“):

Some in-depth article listings don’t have authorship (from a search for “wedding pictures“):

Finally, some listings don’t have publisher icons or names (from a search for “jobs“):

So far, every in-depth article result I’ve seen in the wild has had an image, title, description, and either a publisher name or domain name. Image thumbnails seem to be taken directly from the articles and cropped.

In testing, we saw some in-depth article blocks in the middle of search results, but every example I’ve seen since launch has appeared at the end of the results page – after organic results, but before the bottom ad block. That’s only based on anecdotal evidence, as we’re not currently tracking the position, and Google is likely to mix things up as they move forward and test new variations.

One oddity – in-depth article blocks seem to appear on pages with nine organic results, suggesting that the in-depth block itself may be treated as result #10. It’s getting harder and harder to tell the true count of rankings, but it looks like natural result #10 is getting pushed to page 2, and the block is simply inserted.

Some basic statistics

Across the 10,000 queries that MozCast tracks, 352 displayed in-depth articles the morning of August 12th, which equates to roughly 3.5% of queries. By volume (using Google’s “global” volume metric), these queries accounted for 6.9% of total volume for our 10K data set, suggesting that the search terms tended to be higher-than-average volume.

Google has suggested that in-depth articles will typically trigger for “broad” topics, but that’s a bit vague, so let’s take a look at a few examples from different ends of the spectrum. First off, here are ten high-volume searches (as measured by Google’s “global volume” metric) that triggered in-depth articles on 8/12:

• jobs
• ancestry
• 50 shades of grey
• wedding dresses
• forever21
• bruce springsteen
• smartphone
• led
• pregnancy
• medicare

While these cover the range from a popular novel to a trendy mall store, it does seem like searcher intent is fairly vague in these queries. Someone searching for “led” could be shopping for light bulbs or trying to figure out when Robert Plant is playing near them. The in-depth results for “jobs” contained one article about Steve Jobs:

There’s been some speculation that “broad” might refer to “head” queries (often, single-word searches). Here’s the distribution of the 352 queries by number of words (the number in parentheses is the percentage for the entire 10K data set):

• 1-word = 37.5% (21.1%)
• 2-word = 50.3% (45.6%)
• 3-word = 9.1% (24.4%)
• 4-word = 2.6% (7.0%)
• 5+-word = 0.6% (2.0%)

It’s important to note that the keyword set we use does not contain very long-tail queries and is generally skewed toward shorter phrases. The average word count of all 352 queries is 1.80. For reference, the average word count for our entire 10K data set is 2.24 – so, Google does seem to be leaning a bit toward shorter queries. For reference, here are the five longest queries that showed in-depth articles in our data set:

• church of jesus christ of latter day saints
• the girl with the dragon tattoo movie
• department of homeland security
• post traumatic stress disorder
• mitt romney for president

Our 10K engine tracks a wide variety of queries (by volume, competitiveness, length, industry, etc.), but they do tend a bit toward commercial keywords. We don’t have exact data on brand vs. non-brand queries or commercial vs. informational, but it does appear that in-depth queries are appearing across a wide range of intent.

The news connection

Clearly, it’s hard not to see a news and big media connection in these in-depth articles. Are in-depth articles a replacement for news results? No (at least not for now) – many of the results we tracked had both in-depth articles and a news box. For example, a search for the popular novel “50 Shades of Grey” showed standard news results:

…as well as in-depth articles (note, that there’s no overlap between the articles):

Are posts with news results more likely to show in-depth articles? It certainly looks that way. Across our entire 10K data set, 16.8% of queries contained a news result block on August 12th. For that same time period, 55.7% of queries with in-depth articles contained news results. There’s almost definitely some algorithmic connection between these two entities.

The big winners (so far)

So, given the news connection, do the major news sources have an advantage? At least for now, it seems that way. The 352 searches with in-depth articles on August 12th contained 1,056 articles, which were housed on 123 unique root domains. The top 10 root domains accounted for almost 57% of the total allotment of in-depth articles. Here are the top 10, in order:

1. nytimes.com (20.4%)
2. wsj.com (6.1%)
3. newyorker.com (4.5%)
4. guardian.co.uk (4.3%)
5. wired.com (4.1%)
6. vanityfair.com (3.9%)
7. businessweek.com (3.8%)
8. nymag.com (3.3%)
9. theatlantic.com (3.3%)
10. thedailybeast.com (3.2%)

Within our data set, the New York Times alone accounted for one-fifth of the articles listed in in-depth article blocks. Most of the heavy hitters were generally considered news sites – other big brands like Yahoo.com and MSN.com had isolated articles, but Google didn’t seem to show them any particular favoritism.

To be fair, some smaller news sites and niche sites did show up in the list. Here’s an in-depth article listing from the West Virginia Gazette, for example (from a search for “routers“):

Here’s an example of a niche publication, Yoga Journal, getting listed (from a search for “knee pain“):

Clearly, big publications have an early-mover advantage right now, but what’s unclear is whether that advantage is baked into the in-depth article algorithm or is just a consequence of other authority and content factors. So, that leads us to the million-dollar question: what does it take to break into the in-depth box?

Getting in on the action

While big news organizations have an advantage, there’s no compelling evidence that in-depth articles are a private club. In fact, Google has already posted a support document with advice on getting listed in in-depth articles. I’ll give you a quick-and-dirty summary:

• Use Schema.org article markup
• Set up authorship markup
• Set up a Google+ account, including your logo
• Properly handle paginated articles
• Use “first click free” for paywall content

Ana Hoffman wrote a good post that goes into more detail on these in-depth article support factors. Of course, these aren’t sufficient conditions to get listed – domain authority, content quality, and traditional ranking factors undoubtedly are also at play here. The good news is that Google is telling us that you do have a chance at getting in, and there are ways to help the process.

I suspect Google will be experimenting with and expanding in-depth articles over the next few months, so all of this data is preliminary and subject to change. If you’re a news site or have reputable, long-form content, I’d strongly consider at least putting the signals above into place. If anyone manages to break into an in-depth box, we’d love to hear your story.

Sign up for The Moz Top 10, a semimonthly mailer updating you on the top ten hottest pieces of SEO news, tips, and rad links uncovered by the Moz team. Think of it as your exclusive digest of stuff you don’t have time to hunt down but want to read!

Posted by Annie Cushing

Image from the National Archives

If you want to see how your competitors are gaining a strategic advantage, one of the best tactics to overtake them is to take a deep dive into their backlinks. They leave breadcrumbs behind that reveal their best tactics. Then pivot (no pun intended), glean ideas from their brilliance, and do it even better!

Required skill: pivot tables

If you don’t know how to use pivot tables, you need to check out this video walkthrough. I teach you everything you need to know and then some.

Download example pivot table

I redacted my client’s data from the pivot table in the Excel sheet, but you can get an idea of how I pulled together the data in the “Raw Data” tab and then see how I organized my pivot table in that tab. You can, of course, organize yours however you feel is best. But hopefully this will provide a good jumping-off point. I also sorted my pivot table by domain authority in descending order and then filtered out links from [free-subdomain].wordpress.com.

You can download the Excel workbook from Dropbox.

Steps to pull data together

Step 1: Pull your site’s (or client’s) backlinks — using Open Site Explorer, Majestic SEO, ahrefs, or whatevs — as well as a few of your main competitors. Then pull them together into a formatted table.

Step 2: Add another column and label it “Site.” What I typically do is add the domain (without the http:// or www to minimize noise) and double-click the bottom-right corner of the cell to fill down to the bottom of the data set. Rinse and repeat each time you add a new batch of backlinks. When you finish, you’ll have a single table that contains a mashup of backlinks.

Step 3: Extract the domains from the backlink URLs using the LEFT and SEARCH functions. If you haven’t done this before, I demonstrate how in this video tutorial. (Or you can download the demo workbook from the post and just copy the formulas.)

Step 4: Create your pivot table using these settings:

Step 5: As a general rule, I don’t like how Excel merely indents rows in the default, “Compact” pivot table format. In our data set here, where we have three different values pulled down in rows (Domain, Site, and URL). This can cause your rows to get really congested, and it can be hard to differentiate them. For this reason, if I pull multiple values into the Rows field list, I prefer the Outline layout. You can check it out in the Excel file download.

To set your pivot table to “Outline,” click on any cell in the pivot table and go to Pivot Table Tools > Design tab > Layout > Report Layout > Show in Outline Form (Mac: PivotTable tab > Design > Layout > Outline Layout).

Step 6: I pulled individual observations into text boxes to the left of the pivot table. I used text boxes for a couple reasons:

• I didn’t want to mess up the heights of rows inside my pivot table.
• I could attach links to the text boxes that linked to the individual cells in the pivot table.

Step 7: Dive into the data and start reverse-engineering strategies (using the observations I pulled out as a template, if you’d like). I can assure you there are many more to excavate from that data set!

Video tutorial

For you visual learners out there, I pulled together a 10-minute video walkthrough of some of the key steps I took to organize the data for analysis.

Sign up for The Moz Top 10, a semimonthly mailer updating you on the top ten hottest pieces of SEO news, tips, and rad links uncovered by the Moz team. Think of it as your exclusive digest of stuff you don’t have time to hunt down but want to read!