Wednesday, March 1, 2017

Mastering Google Search Operators in 67 Easy Steps

Posted by Dr-Pete

Any SEO worth their sustainably harvested pink Himalayan salt knows that Google offers a variety of advanced search operators – special commands that take you above and beyond regular text searches. Learning search operators is a bit like learning chess, though. It's easy to memorize how each piece moves, but that's about 1% of your path toward mastery. I know that the pointy-hat guy in chess moves diagonally, but that doesn't mean I'm about to take on Kasparov or Deep Blue.

Instead of just listing all of the operators and telling you what they do, I'd like to try something different. This post is a journey in 67 parts, split into five functional stories:

  1. Content Research
  2. Title Research
  3. Plagiarism Check
  4. Competitive Research
  5. Technical SEO/Audits

You can skip around, but I'd suggest following the story from the beginning. When you're done, you'll understand not only what each operator does, but how to use it in real-world situations and mix-and-match it with other useful operators.


I. Content Research

Crafting original content in 2017 requires wading into the sea of content that's already been created, and Google remains the most complete map of that sea. Advanced search operators are invaluable research tools for content marketers. Let's walk through a sample content journey...

1. Find all the content

tesla

Let's say you've got a blog post to write about the inventor Nikola Tesla. You hop over to Google and search "tesla," only to find a lot of results like this:

Google has decided that Tesla Motors is the dominant intent for this phrase, which doesn't help you very much for your current project.

2. Narrow your search

nikola tesla

So, of course you add more keywords and narrow your search. Now you're on the right track:

Anyone who's ever run a Google search understands this, but there's an important point here that we often overlook. Whenever you string together more than one word in a Google search, Google connects them with a logical AND. This is true of both keywords and operators. If you combine operators, Google will assume that you meant AND and will try to meet all conditions.

3. Mind special characters

tesla ac/dc

Let's say you want to specifically find pages with the phrase "ac/dc", so you try the search above:

Notice the highlighted words – Google has returned anything matching "AC" and "DC" separately. In this case, they've treated the forward slash as the same as a space, which probably isn't what you intended.

4. Force exact match with quotes

tesla "ac/dc"

By putting quotation marks around a phrase, you can force an exact-match search. This requires Google to match the specific, full phrase – with all terms and in the order specified:

This is a lot closer to what you probably expected. Notice the highlighting in the second result, where Google seems to have matched "AC-DC". This is a lot closer than the previous attempt, but Google is still taking some liberties with the forward slash. Be sure to do a sanity check of results any time you use non-alphanumeric characters in a search.

5. Force a logical OR

tesla OR edison

If you specifically want a logical OR between keywords or operators, use the "OR" operator. OR must be in all-caps, or, alternatively you can use the pipe symbol (|):

Note that, in most cases, Google is still going to give priority to results that contain both terms. Specifying logical OR is most useful when two terms only co-occur rarely.

6. Group terms with parentheses

(tesla OR edison) alternating current

Some operators, including OR, are more useful in complex searches. Here, we're using parentheses to group "tesla OR edison" and then are adding "alternating current" as an AND condition:

Requiring all three terms might be unnecessarily restrictive. By using both ANDs and ORs in the same search, we're giving Google a bit more flexibility. Since you probably don't want to memorize the precedence of all Google search operators, I highly recommend using parentheses whenever you're in doubt.

7. Exclude specific terms

tesla -motors

Maybe you want to know what other uses of "tesla" are out there, beyond Tesla Motors. You could use the (-) operator to tell Google to exclude any result with "motors" in it:

Browsing these results, you can see quickly that Tesla is also a band and a unit of measurement. In addition, Tesla the company makes products other than cars. Keyword exclusions are also called "negative keywords" (thus the minus sign).

8. Exclude multiple terms

tesla -motors -car -battery

Just like positive keywords, you can chain together negative keywords:

Keep in mind that each minus sign should only be paired with a single keyword or operator.

9. Exclude exact-match phrases

tesla -motors -"rock n roll"

You can exclude full phrases by using the (-) sign followed by the phrase in quotes:

You can combine individual negative keywords with negative exact-match phrases as needed.

10. Match broadly with wildcards

tesla -motors "rock * roll"

What if you specifically wanted to include more about the rock-n-roll band, but you didn't care whether it was spelled "rock-n-roll," "rock and roll," or "rock & roll," etc.? You can use the asterisk (*) operator as a wildcard to replace any single word:

Wildcards behave most predictably within an exact-match phrase, allowing you to find near-matches when you can't pin down your search to a single phrase. The (*) operator only operates on the word level. There is no single-character wildcard operator.

11. Find terms near each other

tesla AROUND(3) edison

Here's a nifty one. Maybe you want to find results where "Tesla" and "Edison" not only appear in the document but are fairly close to each other. The AROUND(X) operator tells Google to only return results where the two words are within X words of each other:

Phrases like "Tesla vs. Thomas Edison" show up as matches, but an article where the two men were mentioned in separate paragraphs wouldn't.

12. Find near exact-match phrases

"nikola tesla" AROUND(2) "thomas alva edison"

What if, for some reason, you really needed references to include full names? You can combine AROUND(X) with exact-match phrases (in quotes):

AROUND(X) only works on the entities immediately preceding and following it, so be careful when combining it with other operators or phrases that aren't exact-match. Note that AROUND(0) returns strange results – if you want to return two words only if they appear together, use an exact-match phrase instead.

13. Find content on specific sites

nikola tesla site:pbs.org

The "site:" operator is an advanced command that lets you specify a specific domain you want to search on. We usually think of it as a technical SEO and audit tool, but it can also help you refine content searches. Let's say you remembered reading an article on PBS about Tesla, but lost the URL:

Typically, you'll use "site:" with a root domain (i.e. leave subdomains, like "www", off) to match as broadly as possible. Advanced operators like "site:" can be combined with each other and with keywords.

14. Find content on specific TLDs

nikola tesla site:edu

You don't have to include a full domain with "site:". For example, let's say you wanted to find any content about Nikola Tesla on a university website. You could search on all ".edu" domains (also known as a Top-Level Domain, or TLD):

The "site:" operator will not work on a partial domain name. It only accepts full domains, root domains, or TLDs. You can use it on country-specific TLDs (ccTLDs), such as "co.uk" or "com.sg".

15. Find content on multiple TLDs

nikola tesla (site:gov OR site:edu)

Just as with keywords, you can combine "site:" operators with logical OR to search multiple domains:

Often, it's easier and a bit less confusing to run individual searches, but this example is just to illustrate that you can combine advanced operators in complex ways.

16. Dealing with broad matches

discount airfare

Google is getting better at matching synonyms, which is usually good thing, but it sometimes means that results are a lot broader than you might have expected:

Here, a search for "discount airfare" is returning keywords like "cheapest flights," "cheap flights," "airfare deals," and a variety of other combinations.

17. Use exact-match to block synonyms

"discount airfare"

This is another situation where exact-match can help. It doesn't just tell Google to use the full phrase, but it blocks Google from returning any kind of broad match, including synonyms:

Obviously, the results may still contain synonyms (naturally written content often does), but using exact-match ensures that there will be at least one instance of "discount airfare" in each of the results you get back.

18. Exact-match on a single word

discount "airfare"

This may seem counter-intuitive, but you can apply exact match to just one word. In this case, putting an exact match on "airfare" blocks Google from using synonyms just for that word:

Here, Google is free to match on synonyms for "discount" (such as "cheapest"), but every result is forced to include "airfare." Exact-match single words when you want to exclude variations of that word.

19. What to do when exact-match fails

"orbi vs eero vs google wifi"

The other day, I was searching for articles that specifically compared Orbi, Eero, and Google Wifi networking hardware. Something odd happened when I searched on the exact-match phrase:

It's not obvious from the search results themselves, but the first result doesn't contain the phrase anywhere in the body of the text. On rare occasion, Google may match a phrase on secondary relevance factors, such as inbound link anchor text.

20. Search only in the body text

intext:"orbi vs eero vs google wifi"

In these rare cases, you can use the "intext:" operator. This forces Google to find the text in the body of the document. Now, all of the top results clearly have an exact match in the content itself:

Interestingly, the second result reveals what happened with our last search. A Reddit post featured an article from The Verge with an alternate title and used that title as the anchor text. Reddit apparently had enough authority to generate a match via the anchor text alone.

21. Find a set of keywords in the text

allintext: orbi eero google wifi

What if you want to find a set of words, but they don't need to be in an exact-match phrase? You could use a separate "intext:" operator for each word, or you could use "allintext:" which tells Google to apply "intext:" to all of the words following the operator:

All of the results have the target keywords in the body text, in some combination or order. Be very careful about mixing "allintext:" (or any "allin...:" operator) with other commands, or you could end up with unexpected results. The "allintext:" operator will automatically try to process anything that follows it.

(Special thanks to Michael Martinez for working through some "intext:" examples with me on Twitter, and to Google's Gary Illyes for clarifying some of the details about how exactly "intext:" works)


II. Title Research

You've done your content research, and now it's time to pin down a title. You want to capture those clicks, but, of course, you don't want to be unoriginal. Here are some search operator combos for title research.

22. Check for a specific phrase

"tesla vs edison"

You've settled on using "Tesla vs. Edison" in your title, so let's do a quick check on content with that exact-match phrase:

You've pinned down Google to an exact-match phrase, but that phrase can occur anywhere in the text. How do we look for it in just the document title?

23. Check for a phrase in the title

intitle:"tesla vs edison"

Use the "intitle:" operator to specify that a keyword or phrase (in quotes) has to occur in the document title:

Be aware that sometimes Google may rewrite a display title in search results, so it's possible to get a result back where the phrase doesn't seem to match the title because Google has rewritten it.

24. Check multiple keywords in title

intitle:tesla intitle:vs intitle:edison

If you want to check for multiple keywords in a title, but don't want to restrict yourself to exact-match, you can string together multiple "intitle:" operators with single keywords:

Of course, this can be a bit clunky. Luckily, there's an easier way...

25. Check multiple keywords easily

allintitle: tesla vs edison

Like "allintext:", there's an "allintitle:" operator. It will match any of the keywords following it:

This returns roughly the same results as #24, which doesn't make for a very interesting screenshot, but is exactly what we want it to do. Again, be careful combining "allintitle:" with other operators, as it will try to consume everything following it.

26. Check for titles with lists

intitle:"top 10 facts" tesla

Maybe you've got your heart set on a listicle, but you want to make sure it hasn't been done to death. You can combine an "intitle:" operator with a general keyword search on a topic:

These results are all pages that talk about Tesla but have "Top 10 Facts" in the title.

27. Find lists and exact-match phrases

intitle:"top 10 facts" "nikola tesla"

Oops, we 're pulling in results about Tesla Motors again. Luckily, you can combine "intitle:" with exact-match phrases and other, more complex operator combos:

This is much closer to what you probably had in mind, but the bad news is that the "Top 10" things does seem like it's been overdone, even in the realm of Nikola Tesla.

28. Check for Top X lists

intitle:"top 7..9 facts" "nikola tesla"

The range (..) operator lets you search for a specific range of numbers. Maybe you're tired of Top 10, but don't want too short of a list. Let's check out what Top 7, 8, and 9 lists are out there:

This returned only four results, and they were all videos. So, at least you're on the right track, originality-wise. Once you master search operators, you'll eventually reach the mythical end of the Internet.

29. Check the title for this post

intitle:"search operators" "in * easy steps"

Let's put all of this to the test – how original is my title for this post? I'm not expecting an exact match to a post with 67 steps, but what about any post mentioning "Search Operators" in the title that also uses some variation of "in * easy steps" anywhere in the result?

It looks like I did alright, from an originality standpoint. Of course, there are many ways to mix-and-match operators to find similar titles. Ultimately, you have to decide how you define "unique."


III. Plagiarism Check

You've finally published that article, but you suspect someone else may have copied it and is taking your traffic. Advanced search operators can be great for hunting down plagiarism.

30. Find articles with your exact title

intitle:"duplicate content in a post-panda world"

Use the "intitle:" operator with your exact-match title to easily spot whether someone has copied your entire article with no modifications. Here's a search based on a post I wrote a couple of years back:

Ok, you probably didn't need to know about the original article, so let's try again...

31. Find title matches, excluding sites

intitle:"duplicate content in a post-panda world" -site:moz.com

Use (-) with the "site:" operator to exclude specific sites. In this case, we already know that the original title was posted on Moz.com:

It turns out that two of these sites are just linking to the post in kind of a low-quality but not outright malicious way. What you really want to know if someone is copying the text wholesale...

32. Find unique, exact-match text

"they were frolicking in our entrails" -site:moz.com

Another alternative is to run exact-match on a long, unique phrase. Luckily, this particular blog post has some pretty unique phrases. I'm going to keep the Moz.com exclusion:

The first result is a harmless (if slightly odd) Facebook post, but the other two are full, copied-and-pasted duplicates of the original post.

33. Find unique text only in the body

intext:"they were frolicking in our entrails" -site:moz.com -site:facebook.com

If you want to be completely sure that the unique text is in the body of the document, you can use the "intext:" operator. Here, I've added both "intext:" and a Facebook exclusion. Within reason, it's ok to mix-and-match a variety of operators:

Practically speaking, "intext:" often returns similar results to the exact-match phrase by itself. I typically use "intext:" only when I'm seeing strange results or want to make absolutely sure that I'm only looking at document body text.

34. Find a quote you're not sure about

i would rather kiss a wookiee

What if you're looking for a long quote, but you can't remember if you're getting that quote quite right? We often equate exact-match with long searches, but sometimes it's better to let Google go broad:

Here, Google is helpfully reminding me that I'm a lousy Star Wars fan. I've even got an article about all the other people who are wrong about this, too.


IV. Competitive Research

In some cases, your research may be very focused on what kind of content the competition is creating. Google search operators can help you easily narrow down what your competitors are up to...

35. Start with a basic search

tesla announcements

Let's say you want to find out who's publishing Tesla Motors announcements, so you start with the simplest query you can think of:

You're probably not looking for Tesla's own announcements, so you do an exclusion...

36. Exclude obvious sites

tesla announcements -site:tesla.com

You grab the handy "site:" operator and run a negative (-) on Tesla's own site, resulting in:

That's a little better. These are all pretty familiar competitors if you're in the news game.

37. Target specific competitors

tesla announcements site:nytimes.com

Maybe you want to focus on just one competitor. You can use the "site:" operator for that, too:

Obviously, this approach is going to work best for large competitors with a high volume of content.

38. Target a specific subdomain

tesla announcements site:wheels.blogs.nytimes.com

Remember that you can use "site:" with a full subdomain. Maybe you just want to find out what CNN's "Wheels" auto industry blog is posting about.

You can, of course, exclude specific subdomains with "-site:" as well.

39. Target a specific author on a site

tesla announcements site:nytimes.com "neal e boudette"

Maybe you're interested in just a single author. There's no reliable author search operator for organic results, but in most cases, just including the author's name as exact-match text will do the trick:

Make sure to pull up an article first to see how the author's name is presented (middle initial, etc.).

40. Target by keywords, site, and title

tesla announcements site:nytimes.com intitle:earnings

If you wanted Tesla announcements in the New York Times that only mention "Earnings" in the title, then you can mix-and-match operators as needed:

Don't be afraid to get creative. The Google index is a big, big place and there's always more to be found, especially on very large sites.

41. Find related competitors

related:nytimes.com

What if you wanted to branch out to other publications? By using the "related:" operator with a root domain, Google will show you other sites/domains like the one you specify:

The "related:" operator is great when it works, but be warned that it only works for certain niches and typically for larger sites. It's also one of the rare Google search operators that can't be combined with other operators.

42. Find content in a specific path

tesla announcements site:fortune.com/2016

If you want to drill down into a site, you can specify URL folders with the "site:" operator. Forbes, for example, is conveniently organized with year-based folders, so you can easily see just articles from 2016:

Keep in mind that this only works for parts of the URL directly following the domain name. So, how do you search on text in other parts of the URL?

43. Search broadly for a "folder"

tesla announcements inurl:2016

Luckily, Google also has an "inurl:" operator. By searching on a year, for example, you can find that year anywhere it happens to appear in the result URL:

Keep in mind that the text you specify "inurl:" can appear anywhere in the URL, not just at the folder level.

44. Search by a specific date range

tesla announcements daterange:2457663-2457754

What if you really want to narrow down your date range? Google also has a "daterange:" operator which lets you pinpoint publication dates to the day, in theory. For example, here's a search for Q4 of 2016:

Unfortunately, in regular organic results, publication dates aren't always accurate, and "daterange:" can, in practice, return some pretty strange results. You may have noticed, too, that that's not your typical date format. The "daterange:" operator uses the Julian date format.

45. Search by broad date range

tesla announcement 2015..2017

If you don't need your date range to be particularly precise, consider using the range (..) operator with a year on either side of it. As numbers go, years are generally unique enough to return reasonable results:

Please note that this is not specifically a date search, but as cheats go, it's not a bad one. Unfortunately, the range operator doesn't always work properly paired with "inurl:" and other advanced operators.

46. Target just one type of file

tesla announcements filetype:pdf

The "filetype:" operator lets you specify an extension, such as PDF files. Let's say you only want Tesla announcements that have been published as PDFs:

Other file extensions to try are "doc" (Word), "xls" (Excel), "ppt" (PowerPoint), and "txt" (text files). You can also use "filetype:" to specify certain varieties of web pages, including "html", "php", "asp", etc. Keep in mind that the file extension typically has to be listed in the URL, so these searches are not exhaustive.

47. Find sites linking to competitors

link:nytimes.com tesla

The "link:" operator lets you do competitive link research. For example, the search above looks for all documents relevant to Tesla that have links from The New York Times:

Ok, so mostly this tells you that The New York Times links a lot to The New York Times. That's probably not quite what you were looking for...

48. Find links excluding the source

link:nytimes.com -site:nytimes.com tesla

Let's combine "link:" with a negative (-) "site:" operator to remove links from The New York Times:

Please note that Google has deprecated a lot of the functionality of the "link:" operator and the results it returns are just a sample (and, potentially, an unreliable sample). For in-depth competitive link research, we strongly recommend third-party tools, including our own Open Site Explorer.

49. Search inside link anchor text

inanchor:"tesla announcements"

You can use the "inanchor:" operator to search inside linked text. So, for example, the search above looks for sites being linked to from sites using "tesla announcements" in the linked text. In other words, the results represent the targets of those links (not the sources):

Please note that, like the "link:" operator, the "inanchor:" operator represents only a small sample of the index and is no longer actively supported by Google. Use it with a grain of salt.

50. Search multiple words in anchor text

allinanchor: tesla announcements "model x"

Like the other "allin..." varieties, "allinanchor:" applies to every word after it, looking for all of those words in the anchor text, but not as an exact-match:

The three link-based operators ("link:", "inanchor:", "allinanchor:") can be useful for your initial research, but do not expect them to return a full, accurate representation of all links to your site or your competitors.


V. Technical SEO/Audits

Advanced Google search operators can also be powerful tools for understanding how sites are indexed and for performing technical audits. Technical SEO is a complex subject, of course, but here are a few examples to get you started:

51. Glimpse into a site's index

site:amazon.com

It all starts with the "site:" operator, which, at its most basic level, can help you get a glimpse of how Google indexes a site. Here are a few results from Google's index of Amazon.com:

Please note that the result count here (and for any large-volume search) is at best an estimate. Given an estimate of 119,000,000 pages, though, we can be assured that the real number is massive. On the scale of any decent-sized site, you're going to want to drill down...

52. Filter out the "www" subdomain

site:amazon.com -inurl:www

To drill deep into a site's index, the combination of "site:" with "inurl:" will quickly become your best friend. For example, maybe you want to see only pages on Amazon that aren't under the "www" subdomain. You could use "site:" along with a negative match (-) on the "inurl:" operator:

Even in the first few results, you can see a sampling of the other subdomains that Google is indexing. This can give you a good starting point for where to drill down next.

53. Filter out multiple subdomains

site:amazon.com -inurl:www -inurl:logistics -inurl:developer -inurl:kdp

You can extend this concept pretty far, building successively on earlier searches to return narrower and narrower lists of pages. Here's an example with four "-inurl:" operators:

I've done this with over a dozen "inurl:" statements and am not aware of any fixed limit on how many operators you can combine in a single search. Most sites aren't big enough to require those kinds of extremes, but it's good to know that it's possible if and when you need it.

54. Focus on a single subdomain

site:developer.amazon.com

Alternatively, you can focus on a single subdomain. For this, I generally prefer to include the subdomain in the "site:" operator instead of using "inurl:". Otherwise, you could find the text anywhere in the URL:

You could extend this concept to dive deeper into any of the sub-folders returned here ("/ios", "/ja", etc.) and even combine a more specific "site:" operator with additional "inurl:" operators.

55. Filter for non-secure pages

site:amazon.com -inurl:https

Interestingly, you can use "inurl:" to include or exclude secure (https:) pages:

If you're moving a site from "http:" to "https:", this trick can help you make sure that new pages are being indexed properly and old pages are gradually disappearing from the index.

56. Search for a URL parameter

site:amazon.com inurl:field-keywords

You can also use "inurl:" to target URL parameters on dynamic pages. For example, let's say you want to see what kind of internal search pages Google is indexing on Amazon:

Please note that there's no way to specify a URL parameter – Google may find the text anywhere in the URL. On the bright side, many URL parameters tend to have unique names.

57. Search multiple URL attributes

allinurl: amazon field-keywords nikon

Much like "allintitle:" and "allintext:", there's an "allinurl:" operator. In this example, you're looking for internal search pages on Amazon that have the word "Nikon" in the URL:

Unfortunately, "allinurl:" suffers from two problems. One, you can't reliably combine it with "site:", which limits your options. Two, it tends to return strange results. For example, notice that the top results for my US search were from Amazon France. In most cases, I recommend using multiple "inurl:" statements instead.

58. Find stray text files

site:amazon.com filetype:txt -inurl:robots.txt

You might be wondering if you left any stray documentation files laying around your site that happened to get picked up by Google. You can do this using a combination of "site:" and "filetype:":

In this case, you want to exclude "robots.txt" (using "-inurl:") because Amazon has dozens of Robots files. This combo is a good way to clean up files that have been accidentally left live on a site.

59. Dig deep into duplicate content

site:amazon.com "hot wheels 20 car gift pack"

A site like Amazon has massive potential for internal duplicate content. By using the "site:" operator with exact match phrases, you can start to pin down near-duplicates:

In this case, Google is still returning almost 1,000 results. Time to dig deeper...

60. Dig through duplicate titles

site:amazon.com intitle:"hot wheels 20 car gift pack"

You can specifically using "site:" plus "intitle:" to find pages on a site that may be exact duplicates.

Believe it or not, Google still returns over 100 matching pages. Let's keep at it...

61. Find title duplicates with exclusions

site:amazon.com intitle:"hot wheels 20 car gift pack" -inurl:review -inurl:reviews

You dig in and notice that many of the results in #60 are review pages, with either "review" or "reviews" in the URL. So, you build on the previous search and add two exclusions:

Voilà... you're down to just a half-dozen results. You just leveled up in technical SEO.

62. Find similar products with different counts

site:amazon.com "hot wheels * car gift pack"

Maybe you're curious about other Hot Wheels gifts packs that represent similar products but not exactly the same one. You could replace "20" with the wildcard (*) operator:

Unfortunately, wildcards don't play well with the "intitle:" operator, so you'll generally be restricted to exact-match phrases outside of advanced operators.

63. Find similar products with exclusions

site:amazon.com "hot wheels * car gift pack" -20

Given all of the previous searches, you probably don't need to know about the 20-packs, so you can add an exclusion on the number 20 (just treat it as a word with negative match):

Looks like there's a healthy number of 5-car gift packs as well. The plot thickens...

64. Follow the rabbit hole to Wonderland

site:amazon.com "hot wheels * car gift pack" -20 -5

It's time to take the red pill and find out just how deep this rabbit hole goes. You can keep adding exclusions and take out the 5-packs as well:

Finally, you're nearing the bottom. This process may seem a bit obsessive, but auditing large sites is a process of identifying potential problems and drilling down until you either you pin down the issues or decide they aren't worth worrying about. Once you master them, advanced search operators shine at drill-downs.

65. Bonus: Show me the money!

site:amazon.com "hot wheels" $19.95

I woke up in a cold sweat at 2am realizing I had forgotten a search operator (sadly, while you may find it funny, this is not a joke). I warned earlier that special characters can produce weird results, but one that Google does recognize is the dollar sign ($):

This isn't really a site audit example, but it fits well with our Amazon story. Keep in mind that, while Google will honor the ($) in the results, they could appear anywhere in those results. Many Amazon pages list multiple prices. Still, it can be a useful tool to add to your arsenal.

66. Find results in a price range

site:amazon.com "hot wheels" $19..$20

You can also combine a ($) search with the range operator (..) and search a range of prices. Let's say you wanted to find any pages mentioning "Hot Wheels" and prices in the $19-20 range:

While this tactic can definitely be useful for general product research, e-commerce sites can also use it in an audit to find pages with incorrect or outdated prices.

67. Find other TLDs for your brand

site:amazon.* -site:amazon.com

This last tip could be either an audit trick or a way to track down the competition, depending on how you use it. Use the wildcard (*) in the top-level domain (TLD) to find any site with the same name, and then exclude the main site:

For a large site, like Amazon, this could help you find other legitimate TLDs, including country-specific TLDs (ccTLDs). Alternatively, you could use this trick to find competitors who have registered your brand name under other TLDs.


Wait, You're Still Here?

Congratulations for making it this far. I hope you've picked up at least a handful of useful tricks and the confidence to experiment. If you have favorites I'm missing, please feel free to share them in the comments. I'm sure there's a good trick or ten I've never seen.

If you need a quick reference, we've launched a new Search Operators reference and cheat sheet in the Learning Center. This resource reflects the current state of Google's search operators, as best we know, including deprecated operators.


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!

Tuesday, February 28, 2017

How to Create Content That Keeps Earning Links (Even After You Stop Promoting It)

Posted by kerryjones

Do your link building results look something like this?

  1. Start doing outreach
  2. Get links
  3. Stop doing outreach
  4. No more links

Everyone talks about the long-term benefits of using content marketing as part of a link building strategy. But without the right type of content, your experience may be that you stop earning links as soon as you stop doing outreach.

In this sense, you have to keep putting gas in the car for it to keep running (marketing “gas” = time, effort, and resources). But what if there was a way to fill up the car once, and that would give it enough momentum to run for months or even years?

An example of this is a salary negotiations survey we published last year on Harvard Business Review. The study was picked up by TechCrunch months after we had finished actively promoting it. We didn’t reach out to TechCrunch. Rather, this writer presumably stumbled upon our content while doing research for his article.

techcrunch-link.png

So what’s the key to long-term links? Content that acts as a source.

The goal is to create something that people will find and link to when they’re in need of sources to cite in content they are creating. Writers constantly seek out sources that will back up their claims, strengthen an argument, or provide further context for readers. If your content can serve as a citation, you can be in a good position to earn a lot of passive links.

Read on for information about which content types are most likely to satisfy people in need of sources and tips on how to execute these content types yourself.

Original research and new data

Content featuring new research can be extremely powerful for building authoritative links via a PR outreach strategy.

A lot of the content we create for our clients falls under this category, but not every single link that our client campaigns earn are directly a result of us doing outreach.

In many cases, a large number of links to our client research campaigns earn come from what we call syndication. This is what typically plays out when we get a client’s campaign featured on a popular, authoritative site (which is Site A in the following scenario):

  • Send content pitch to Site A.
  • Site A publishes article linking to content.
  • Site B sees content featured on Site A. Site B publishes article linking to content.
  • Site C sees content featured on Site A. Site C publishes article linking to content.
  • And so on…

So, what does this have to do with long-term link earning? Once the content is strategically seeded on relevant sites using outreach and syndication, it is well-positioned to be found by other publishers.

Site A’s content functions as the perfect citation for these additional publishers because it’s the original source of the newsworthy information, establishing it as the authority and thus making it more likely to be linked to. (This is what happened in the TechCrunch example I shared above.)

Examples

In a recent Experts on the Wire podcast, guest Andy Crestodina talked about the “missing stat.” According to Andy, most industries have “commonly asserted, but rarely supported” statements. These “stats” are begging for someone to conduct research that will confirm or debunk them. (Side note: this particular podcast episode inspired this post – definitely worth a listen!)

To find examples of content that uncovers a missing stat in the wild, we can look right here on the Moz blog…

Confirming industry assumptions

When we did our native advertising versus content marketing study, we went into it with a hypothesis that many fellow marketers would agree with: Content marketing campaigns perform better than native advertising campaigns.

This was a missing stat; there hadn’t been any studies done proving or debunking this assumption. Furthermore, there wasn’t any publicly available data about the average number of links acquired for content marketing campaigns. This was a concrete data point a lot of marketers (including us!) wanted to know since it would serve as a performance benchmark.

Screen Shot 2017-02-27 at 1.16.47 PM.png

As part of the study, we surveyed 30 content marketing agencies about how many links the average content marketing campaign earned, in addition to other questions related to pricing, client KPIs, and more.

After the research was published here on Moz, we did some promotion to get our data featured on Harvard Business Review, Inc, and Marketing Land. This data is still being linked to and shared today without us actively promoting it, such as this mention on SEMRush’s blog and this mention on the Scoop It blog (pictured below).

scoop-it-citation.png

To date, it’s been featured on more than 80 root domains and earned dozens of co-citations. It’s worth noting that this has been about far more than acquiring high-quality links; this research has been extremely effective for driving new business to our agency, which it continues to do to this day.

Debunking industry assumptions

But research doesn’t always confirm presumptions. For example, Buzzsumo and Moz’s research collaboration examined a million online articles. A key finding of their research: There was no overall correlation between sharing and linking. This debunked a commonly held assumption among marketers that content that gets a lot of shares will earn a lot of links, and vice versa. To date, this post has received an impressive 403 links from 190 root domains (RDs) according to Open Site Explorer.

How to use this strategy

To find original research ideas, look at how many backlinks the top results have gotten for terms like:

  • [Industry topic] report
  • [Industry topic] study
  • [Industry topic] research

Then, using the MozBar, evaluate what you see in the top SERPs:

  • Have the top results gotten a sizable number of backlinks? (This tells you if this type of research has potential to attract links.)
  • Is the top-ranking content outdated? Can you provide new information? (Try Rand’s tips on leveraging keywords + year.)
  • Is there a subtopic you could explore?

Additionally, seeing what has already succeeded will allow you to determine two very important things: what can be updated and what can be improved upon. This is a great place to launch a brainstorm session for new data acquisition ideas.

Industry trend and benchmark reports

Sure, this content type overlaps with “New Research and Studies,” but it merits its own section because of its specificity and high potential.

If your vertical experiences significant change from one year, quarter, or month to the next, there may be an opportunity to create recurring reports that analyze the state of your industry. This is a great opportunity to engage all different kinds of brands within your industry while also showcasing your authority in the subject.

How?

People often like to take trends and add their own commentary as to why trends are occurring or how to make the most of a new, popular strategy. That means they’ll often link to your report to provide the context.

And there’s an added promotional benefit: Once you begin regularly publishing and promoting this type of content, your industry will anticipate future releases.

Examples

HubSpot’s State of Inbound report, which features survey data from thousands of HubSpot customers, has been published annually for the last eight years. To date, the URL that hosts the report has links from 495 RDs.

Content Marketing Institute and MarketingProfs have teamed up for the last seven years to release two annual content marketing benchmark reports. The most recent report on B2B content marketing has earned links from 130 RDs. To gather the data, CMI and MarketingProfs emailed a survey to a sample of marketers from their own email marketing lists as well as a few lists from partner companies.

In addition to static reports, you can take this a step further and create something dynamic that is continually updated, like Indeed’s Job Trends Search (171 RDs) which pulls from their internal job listing data.

How to use this strategy

Where can you find fresh industry data? Here are a few suggestions:

Survey your customers/clients

You have a whole pool of people who have been involved in your industry, so why not ask them some questions to learn more about their thoughts, needs, fears, and experiences?

Talking directly to customers and clients is a great way to cut through speculation and discover exactly what problems they’re facing and the solutions they’re seeking.

Survey your industry

There are most likely companies in your industry that aren’t direct competitors but have a wealth of insight to provide to the overall niche.

For example, we at Fractl surveyed 1,300 publishers because we wanted to learn more about what they were looking for in content pitches. This knowledge is valuable to any content marketers involved in content promotions (including ourselves!).

Ask yourself: What aspect of your industry might need some more clarification, and who can you reach out to for more information?

Use your internal company data

This is often the easiest and most effective option. You probably have a ton of interesting data based on your interactions with customers and clients that would benefit fellow professionals in your industry.

Think about these internal data sets you have and consider how you can break it down to reveal trends in your niche while also providing actionable insights to readers.

Curated resources

Research can be one of the most time-consuming aspects of creating content. If someone has pulled together a substantial amount of information on the topic in one place, it can save anyone else writing about it a lot of time.

If you’re willing to put in the work of digging up data and examples, curated resource content may be your key to evergreen link building. Let’s look at a few common applications of this style of content.

Examples

Collections of statistics and facts

Don’t have the means to conduct your own research? Combining insightful data points from credible sources into one massive resource is also effective for long-term link attraction, especially if you keep updating your list with fresh data.

HubSpot’s marketing statistics list has attracted links from 963 root domains. For someone looking for data points to cite, a list like this can be a gold mine. This comprehensive data collection features their original data plus data from external sources. It’s regularly updated with new data, and there’s even a call-to-action at the end of the list to submit new stats.

Your list doesn’t need to be as broad as the HubSpot example, which covers a wide range of marketing topics. A curated list around a more granular topic can work, too, such as this page filled with mobile email statistics (550 RDs).

Concrete examples

Good writers help readers visualize what they’re writing about. To do this, you need to show concrete evidence of abstract ideas. As my 7th grade English teacher used to tell us: show, don’t tell.

By grouping a bunch of relevant examples in a single resource, you can save someone a lot of time when they’re in need of examples to illustrate the points they make in their writing. I can write thousands of words about the idea of 10x content, but without showing examples of what it looks like in action, you’re probably going to have a hard time understanding it. Similarly, the bulk of time it took me to create this post was spent finding concrete examples of the types of content I refer to.

The resource below showcases 50 examples of responsive design. Simple in its execution, the content features screenshots of each responsive website and a descriptive paragraph or two. It’s earned links from 184 RDs.

Authority Nutrition’s list of 20 high-protein foods has links from 53 RDs. If I’m writing a nutrition article where I mention high-protein foods, linking to this page will save me from researching and listing out a handful of protein-rich foods.

How to use this strategy

The first step is to determine what kind of information would be valuable to have all in one place for other professionals in your industry to access.

Often times, it’s the same information that would be valuable for you.

Here are some ways to brainstorm:

  • Explore your recent blog posts or other on-site content. What needed a lot of explaining? What topics did you wish you had more examples to link to? Take careful note of your own content needs while tackling your own work.
  • Examine comments on other industry articles and resources. What are people asking for? This is a gold mine for the needs of potential customers. You can take a similar approach on Reddit and Quora.
  • What works for other industries that you can apply to your own? Search for terms like the following to see what has been successful for other niches that you can apply to yours:
    • [Industry topic] examples
    • types of [industry topic]
    • list of [Industry topic]
    • [Industry topic] statistics OR stats
    • [Industry topic] facts

No matter which way you choose to proceed, the time investment can help you garner many links down the line.

Beginner content

Every niche has a learning curve, with various words, concepts, and ideas being foreign to a beginner.

Content that teaches noobs the ins and outs of your vertical has long-term linking potential. This type of content is popular for citations because it saves the writer from explaining things in their own words. Instead, they can link to the expert’s explanation.

And the best part is you can tap your internal experts to provide great insights that can serve as the foundation for this type of content.

Examples

101 Content

Moz’s Beginner’s Guide to SEO is a master class in how comprehensive beginner-level content becomes a link magnet. Not only does the guide have backlinks from more than 1,700 RDs, it also edges out the home page as the most-trafficked page on the site, according to SEMrush.

“What is...?”

Beginner content need not be as massive and thorough as the Moz guide to be linkable. It can be as simple as defining an industry term or concept.

Moz’s meta description page, which has backlinks from 244 RDs, is a solid example of an authoritative yet simple answer to a “what is?” query.

Another example is the first result in Google for the query “what is the Paleo diet,” which has 731 links from 228 RDs. It’s not a 10,000-word academic paper about the paleo diet. Rather, it’s a concise answer to the question. This page has served as an excellent source for anyone writing about the Paleo diet within the last several years.

screenshot-robbwolf.com 2017-02-21 14-17-01.png

If a lot of adequate top-level, definition-style content already exists about topics related to your vertical, consider creating content around emerging terms and concepts that aren’t yet widely understood, but may soon be more mainstream.

The perfect example of this? Creating a definitive explanation about content marketing before the entire world knew what content marketing meant. Case in point: Content Marketing Institute’s “What is Content Marketing?” page has amassed an impressive from 12,462 links from 1,100 root domains.

How to use this strategy

Buzzsumo recently released a new tool called Bloomberry which scours forums including Reddit and Quora for questions being asked about a keyword. You can search by time period (ex. questions asked within the last 6 months, all-time results, etc.) and filter by source (ex. only see questions asked in Reddit).

Use Bloomberry to see what beginner questions are being asked about your keyword/topic. Keyword ideas include:

  • [Industry topic] definition
  • How does [industry topic] work
  • [Industry topic] guide
  • What is [industry topic]

After doing the search, ask yourself:

  • What questions keep coming up?
  • How are these common questions being answered?

Bloomberry is also useful for spotting research opportunities. Within the first few results for “SaaS” I found three potential research ideas.

bloomberry.png

Pro tip: Return to these threads and provide an answer plus link to your content once it’s published.

Yes, you still need to promote your content

Don’t mistake this post as a call to stop actively doing outreach and promotion to earn links. Content promotion should serve as the push that gives your content the momentum to continue earning links. After you put in the hard work of getting your content featured on reputable sites with sizable audiences, you have strong potential to organically attract more links. And the more links your content has, the easier it will be for writers and publishers in need of sources to find it.

What types of content do you think are best for earning citation links? I’d love to hear what’s worked for you – please share your experiences in the comments below.

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!

Friday, February 24, 2017

3 Tactics for Hyperlocal Keywords - Whiteboard Friday

Posted by randfish

Trying to target a small, specific region with your keywords can prove frustrating. While reaching a high-intent local audience is incredibly valuable, without volume data to inform your keyword research, you'll find yourself hitting a wall. In this Whiteboard Friday, Rand shares how to uncover powerful, laser-focused keywords that will reach exactly the right people.

Hyper Local Keyword Research

Click on the whiteboard image above to open a high-resolution version in a new tab!

Video Transcription

Howdy, Moz fans, and welcome to another edition of Whiteboard Friday. This week we're going to chat about hyperlocal keyword research. Now, this is a big challenge, not only for hyperlocal-focused businesses, but also for all kinds of websites that are trying to target very small regions, and many of them, with their keyword research and keyword targeting, on-page optimization.

The problem:

So the problem tends to be that most keyword research tools, and this includes things like the Google AdWords Tool, it includes Moz's Keyword Explorer, or KeywordTool.io, or Übersuggest, or anybody you want to use, most of them are relying on volume data.

So what happens is when you see a bunch of keyword suggestions, you type in "Sequim," for example, Sequim is a tiny town on Washington's peninsula, so across the Puget Sound from where we are here in Seattle. Sequim has a population of like 6,500 people or something like that, so very tiny. So most searches related to Sequim have no volume data in any of these tools. As a result, you don't see a lot of information about: How can I target these keywords? What are the right ones to go after? You don't know whether a keyword has zero searches a month, or whether it has four searches a month, and those four searchers are exactly who you want to get in front of, and this is really problematic.

There are three solutions that we've seen professional SEOs use and that some of us here at Moz use and the Moz Local team uses, and these can be real handy for you.

Solution 1: Use keyword data for larger, similar regions

So the first one is to basically replicate the data by using keyword information that comes from similar regions nearby. So let's say, okay, here we are in Sequim, Washington, population 6,669. But Port Angeles is only a few miles away. I think maybe a couple dozen miles away. But its population is more like 20,000. So we've got four or five times the keyword volume for most searches probably. This is going to include some outlying areas. So now we can start to get data. Not everything is going to be zero searches per month, and we can probably backtrack that to figure out what Sequim's data is going to be like.

The same thing goes for Ruidoso versus Santa Fe. Ruidoso, almost 8,000. But Santa Fe's population is almost 10 times larger at 70,000. Or Stowe, Vermont, 4,300, tiny, little town. Burlington is nearby, 10 times bigger at 42,000. Great. So now I can take these numbers and I can intuit what the relative volumes are, because the people of Burlington are probably similar in their search patterns to the people of Stowe. There are going to be a few differences, but for most types of local searches this will work.

Solution 2: Let Google autosuggest help

The second one, Google autosuggest can be really helpful here. So Google Suggest does not care if there's one search a month or one search in the last year, versus zero searched in the last year. They'll still show you something. Well, zero searched in the last year, they won't show you anything.

But for example, when I search for "Sequim day," I can intuit here, because of the ordering that Google Suggest shows me, that "Sequim day spa" is more popular than "day care." Sequim, by the way, sounds like a lovely place to live if you are someone who enjoys few children and lots of spa time, apparently. Then, "day hikes."

So this technique doesn't just work with Google itself. It'll also work with Bing, with Google Maps, and with YouTube. Another suggestion on this one, you will see different results if you use a mobile device versus a desktop device. So you might want to change it up and try your mobile device. That can give you some different results.

Solution 3: Use lexical or related SERP suggestions

All right. Third tactic here, last one, you can use sort of two styles of keyword research. One is called lexical, which is basically the semantic relationships between words and phrases. The other one is related SERP suggestions, which is where a keyword research tool — Moz Keyword Explorer does this, SEMrush is very popular for this, and there are a few others — and they will basically show you search terms the links that came up, the search results that came up for "Sequim day care" also came up in searches for these terms and phrases. So these are like SERPs for which your SERP also ranked.

You can see, when I searched for "Sequim day care," I did this in Keyword Explorer, because I happen to have a Moz Keyword Explorer subscription. It's very nice of Moz to give me that. You can see that I used two kinds of suggestions. One are related to keywords with similar results, so that's the related SERPs. The other one was based on closely related topics, like the semantic, lexical thing. "Sequim day care" has given me great stuff like "Banbury School Nursery," a nearby town, "secondary schools in Banbury," "Horton Day Nursery," which is a nursery that's actually near there, "Port Angeles childcare," "children's nursery."

So now I'm getting a bunch of keyword suggestions that can potentially be relevant and lead me down a path. When I look at closely related topics, I can see things like closely related topics. By the way, what I did is I actually removed the term "Sequim," because that was showing me a lot of things that are particular to that region. But if I search for "day care," I can see lots of closely related topics, like day care center, childcare, school care, special needs children, preschool programs, and afterschool programs. So now I can take all of these and apply the name of the town and get these hyperlocal results.

This is frustrating still. You don't have nearly the data that you have for much more popular search terms. But this is a good way to start building that keyword list, targeting, experimenting, and testing out the on-page work that you're going to need to do to rank for these terms. Then, you'll start to see your traffic grow from these.

Hyperlocal may be small, but it can be powerful, it can be very targeted, and it can bring you exactly the customers you're looking for.

So good luck with your targeting out there, and we'll see you again next week for another edition of Whiteboard Friday. Take care.

Video transcription by Speechpad.com


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!