What I Learned Publishing 200+ Blog Posts on CXL

This is my last week at CXL. It’s bittersweet. I started on a Monday and published my first post on a Thursday. Since then, it’s been rinse and repeat for nearly two-and-a-half years. In sum, I wrote 46 posts and edited another 156. That works out to about a half-million words and a new post […]

The post What I Learned Publishing 200+ Blog Posts on CXL appeared first on CXL.

This is my last week at CXL. It’s bittersweet. I started on a Monday and published my first post on a Thursday. Since then, it’s been rinse and repeat for nearly two-and-a-half years.

In sum, I wrote 46 posts and edited another 156. That works out to about a half-million words and a new post every 4 days for 870 days.

Through it all, here’s what I figured out—and what I failed to solve.

5 things I learned 

1. Your brand isn’t what you just published; it’s what people see most often.

If, like CXL, you have 750+ posts floating around the Internet, your content brand isn’t what your strategy is now—the one that’s affected, say, your last 20 or 30 posts. It’s the other 700 posts that account for 95% of your traffic (since new stuff takes time to rank).

google analytics organic traffic to blog posts.
Organic user growth to posts I worked on. More users landed in November 2020 than my first eight months on the job combined. Organic takes time.

If your top five blog posts—which were published years ago—account for 30% of new users, then those posts are your brand for one in every three people. When someone finds a broken link, missing image, or dated bit of advice, your reaction may be, “Oh, that post? We did that years ago.” But for them, it’s all they know.

The division of traffic, not recency, shapes your content brand. That’s all too easy to forget. We work in the moment. We form forward-looking strategies.

Ask: If none of our content had a publish date, if there were no time-ordered list of posts, which articles would we say are the core of our brand? The most important topics and high-traffic posts may be buried deep in your blog roll. They should be top of mind.

We now update twice as many posts as we publish. It doesn’t take as long (one to four hours is enough to satisfy users and search engines), but it’s the most valuable thing we’ve done for organic traffic and the biggest opportunity for our content brand.

triage document to prioritize blog post updates.
Our triage document prioritizes content updates based on historical traffic, potential traffic, and evergreen-ness.

We started by fixing the big mistakes (broken stuff), and will, in second and third iterations, work on smaller things—off-brand tone, so-so articles, etc. 

2. Set a hard deadline that’s barely feasible—then use that pressure to improve processes. 

An old boss used to say, “Deadlines are your friend.” For her, deadlines staved off procrastination. But they’re also a cap on quality.

Everyone publishes on a budget. Time is part of that budget. Journalists have to make the morning paper. Magazines have to get out each week. Even academics—the most immune to deadlines—have to wrap up articles and books in time to earn tenure. 

It’s a balance. If you told me I had to publish a post every day, the quality would go down. But if you didn’t demand any cadence, there’s no way I’d say, “Twice a week sounds good.” That’s at the fringe of feasibility.

I can’t tell you how often I scrambled to get a post out the door in time for the Thursday morning newsletter. (I failed every six weeks or so.)

slack conversation about newsletter.
This conversation with Kyndall, who sent out the newsletter, happened…often.

But that pressure has been a great catalyst for process development. It’s absolutely possible to get a high-quality post out the door in 15 hours. A fixed deadline inspires efficiency. (It also banishes the low-ROI fiddling that can drag out final revisions for an extra week.)

Over the course of a year or so, running the blog went from a 40-hour-a-week gig to a 25-hour-a-week effort. The surplus time then went to stand-alone projects, post updates, etc.

(Yes, I could’ve used it to build more cushion into the publishing schedule, but that would only lower stress—which was manageable—not add business value.)

If CXL didn’t have a 10-year precedent of what a “good” post was or how often we usually published, it would’ve been difficult, from Day 1, to know that we had the right balance of publishing more versus publishing better.

But if you need a starting point, know that posts like ours take about one hour per 100 words.

3. Don’t try to create a perfect strategy; execute a good strategy more often.

I came to CXL after several years at an agency. Agency life teaches you the value of pretty slide decks. You need them to win client buy-in. If you have to rework (or abandon) a strategy, clients lose faith.

As a result, you spend a lot of time trying to create an airtight strategy—one that you’re confident will work, with milestones spread across several months. But you’re relying on a lot on hypotheticals and “best practices.”

No strategy guarantees success, especially at a micro level (e.g., picking topics for blog posts). We’ve published some great things that few people saw. We’ve published some mediocre posts that drive tons of traffic and leads. (Google’s standards are lower than ours.)

One post went viral because—unbeknownst to us—it gave exquisitely timed reassurance to an anxious influencer, who then shared it.

You can’t plan for this stuff. What you can plan for is the general shape of things—a makes-sense-but-not-foolproof strategy. From there, you’ll have more success if you execute that strategy more often. (Executing on it more often also gives you real-world feedback—the best way to refine your strategy.)

dart board with three darts.
Once you can hit the board with your darts, start throwing more darts—don’t bother looking for a technique that guarantees a bullseye. 

4. If you want a unique voice, write about topics you’ve got no shot to rank for.

Yes, this is counterintuitive, but hear me out.

If you’re just starting in content marketing, these are the typical steps for a by-the-book, keyword-targeted strategy:

  1. Identify the most relevant topics you should write about.
  2. Realize that the SERPs for the “core” aspects of that topic are way too competitive (i.e. dominated by big sites with tons of links).
  3. Find related, long-tail topics that have less search volume and lower competition.
  4. Publish on those until your domain is strong enough to go after the original keywords.

It makes sense. It drives traffic. But it will destroy any chance to stand out. The strategy, in effect, asks, “What keyword volume will remove all incentives to publish something unique?”

As a distribution channel, search rarely rewards tone, design, or angle. So much content looks the same because everyone’s style guidelines try to please the same search engine.

Flip the script:

  1. Identify the most relevant topics you should write about.
  2. Pick topics that big sites own, ones for which you have no prayer to rank for.
  3. Create content on those topics with a novel take or presentation.
  4. Embed those unique elements—of language, visuals, whatever—in all the content you create.

If you work to get eyeballs on your content before search is involved, you’ll establish the standards you need to get attention, now and later. (Retrofitting existing content is clumsy and expensive.)

Put another way: Quantitative data is the proof, but qualitative data is the story. This isn’t a shock to those who do conversion research—quant data is the “what”; qualitative data is the “why.”

But if you want a compelling research study, you’d better ask at least one open-ended question. Code those qualitative responses (ideal H2s) and use individual excerpts to add raw energy to your write-up. That’s what pulls people through an otherwise dry report and gets social shares.

The quantitative data, by contrast, wins the links—it’s the benchmark data that people love to cite.

A final note: Don’t overcomplicate original research. The U.S. News & World Report studies on cities, colleges, hospitals, and schools rely on about a dozen data points. All but a few are publicly available. The rest fall into one of three categories:

  1. A purchased dataset (e.g., from Gallup);
  2. One of their other studies (e.g., using hospital and school data for place rankings);
  3. Some non-scientific email survey.
methodology for u.s. news and world report studies.
Ah, yes, nothing says “authoritative” like a SurveyMonkey poll of 3,000 randos.

The third category is key. It generates “proprietary” results—the one extra data point you need to keep anyone else from replicating your results.

Grab some quantitative stuff that’s already out there. (Kaggle has a ton; Siege Media cataloged many sources.) Run a survey with an open-ended question. Mash mash. Publish.

If only everything were so simple.

3 problems I didn’t solve

Too many of these sorts of posts are triumphant—challenges faced, challenges met. I’m walking away knowing that there are things I didn’t get right or haven’t solved.

It’s engaging to work on hard problems; it’s disappointing not to see them through to a cathartic resolution.

1. Blogs are hamster wheels. I should’ve stepped off more often.

This is the single biggest reason I didn’t solve more problems. The weekly fist-pump moment of this job is right after you publish the second post of the week, usually Wednesday afternoon or Thursday morning.

The rest of the week is a denouement—a much-needed winding down to get ready for next week’s plot twists and frenetic climax. The blog is always the main event, and it requires heads-down work, not introspection.

That slows progress. As Amos Tversky, Daniel Kahneman’s long-time research partner, said, “The secret to doing good research is always to be a little underemployed. You waste years by not being able to waste hours.”

hamster next to wheel.
The author in an undated photo. (Image source)

I could’ve made blog production more efficient—or published better stuff—had I taken the occasional Monday and said, “Okay, this week, the most important thing is improving how we operate, even if it means we don’t publish anything.”

I never did.

The result was incremental improvement. The 15-hour reduction (from 40 hours to 25) that took over a year was probably achievable in six months had I stepped out of the wheel once a quarter.

For anyone running a blog or thinking of starting one (godspeed), build thinking time into your editorial calendar—at least one week per quarter. No one outside your company will care if you don’t publish for a week, and you’ll get better a hell of a lot more quickly.

2. Those “side projects” are usually unfinished—or not very good.

Like any startup, we have tons of ideas. Also like any startup, we don’t have the capacity to execute them. We do, of course, have the boundless enthusiasm to think we can execute them.

That spare 15 hours after the last post goes out isn’t the best time for creative problem solving. It’s a great time to update blog posts, do email outreach, or other fuzzy-brain tasks, but you—or I, anyway—need the first fruits of my brain to execute complex content projects.

With rare exception, the additional content projects I worked on fell into one of two buckets:

1. They didn’t get done. We have several dusty strategy docs and a few half-baked projects.

For example, we grabbed thousands of screenshots of Amazon’s homepage over the last 20 years. We planned to analyze them to create a visual history of one of the web’s most ardent champions of iterative design.

Summer 2019 was the last time I touched it. We adjusted priorities. We had a domain migration that December. And, oh yeah, we started developing a new product (Adeft).

screenshot of amazon homepage from 2011.
You could be forgiven for thinking this was Amazon’s first-ever design, but it’s less than 10 years old.

2. They weren’t very good. Here’s another example: We wanted to increase traffic by targeting head terms that were too broad for a blog post (e.g., “email marketing”).

Hub pages were a cheap option because we could add a small amount of original content (e.g., definitions, FAQs), then automatically pull in relevant posts, webinars, and courses based on WordPress tags.

But we had only bare-bones design and dev resources. “We’ll put out a beta version and see if it gets traction,” we said. But these pages needed to earn links to rank. They needed to be so good that we were proud to promote them. They weren’t.

So, each one drifted around Page 3 or 4 of search results, as invisible as a slim volume in a cavernous library.

photo of huge library.
The Tianjin Binhai Library in China. (Image source)

In retrospect, it would’ve made sense to outsource some of these projects—give them to people for whom they could’ve been the number-one priority.

That, or we should’ve shoved the blog out of the way for a week here or there. But it’s a hard sell (to yourself or others) to sacrifice output of a thing that you know works.   

About 60–70% of a given week can go toward serious, creative work. The rest isn’t wasted (plenty of fuzzy-brain work is super valuable), but I have too often assumed that all hours are equal.

3. Some experiments take a long time to run (maybe too long for a startup). 

The biggest “before-and-after CXL” change in my thinking has been to go from a strategy- to experimentation-centric workflow.

I have Peep and, as a discipline, conversion optimization to thank for that. I move faster. I spend less time pondering. (Why speculate when you can get user feedback?) New ideas justify a test, not a strategy.

But the A/B testing mindset often defaults to two- or four-week cycles. Those are comically short time periods if you work in content or SEO.

So how, at an experimentation-focused startup, do you iterate rapidly if you need six months to get back data? Should you stick to a strategy if a shift in your product or market resets priorities?

I don’t know.

Conclusion

For the life of me, I can’t find who tweeted(?) it, but someone on some platform once wrote that when things, as they do, get all startup-y, it’s best to “roll down the windows, turn up the music, and keep driving.”

That’s good advice. And this has been great fun.

The post What I Learned Publishing 200+ Blog Posts on CXL appeared first on CXL.

How to Analyze SERPs to Win Big in Rankings

Google is always changing how it displays search results. The starting point for any effective SEO strategy is understanding what Google chooses to show and why. Only then can you figure out what you need to create or adjust on your site to show up more often—and in higher positions.  In this article, you’ll learn […]

The post How to Analyze SERPs to Win Big in Rankings appeared first on CXL.

Google is always changing how it displays search results. The starting point for any effective SEO strategy is understanding what Google chooses to show and why. Only then can you figure out what you need to create or adjust on your site to show up more often—and in higher positions. 

In this article, you’ll learn a key part of that process: analyzing SERPs. I’ll also show you how to apply your analysis to win more relevant clicks.

SERPs aren’t static. They’ve changed—a lot.

Over the last decade, SERPs have changed in two crucial ways:

  1. SERP features, not blue links, are the top result (or results) more often.
  2. Searches have become intensely customized for intent.

Beginning in 2013, Google began showing “answer boxes,” an early test of featured snippets. These answer boxes used the Knowledge Graph to show a box that answered a search query. Users no longer needed to click through to get the definition they wanted.

Definition of SERP.

Featured snippets have since expanded the answers that Google can pull directly into search results. By 2018, Google expanded featured snippets show more than one featured snippet for certain queries.

How to setup forward calling search example.

As “a single featured snippet isn’t right for every question,” multiple featured snippets had the potential to answer more user queries—and reduce user clicks. Indeed, by 2019, research by Sparktoro found that more than 50% of searches end without clicks.

Traditional organic results have moved in one direction: down. But not every SERP looks the same, and not every SERP feature sticks around forever—which is why SERP analysis is so important. 

Let’s use the query “what is digital marketing” as an example.

What is digital marketing.

Here we can see:

  1. A “People also ask” section right below the first result. 
  2. A knowledge graph panel for “Internet Marketing,” including a “People also search for.”
  3. Recommended videos from YouTube related to digital marketing. 

In this example, the “People also ask” section and “Videos” take up a significant portion of space, reducing the visibility of organic results shown on Page 1. 

The SERP is even more challenging to break into on mobile, which features two ads above all else:

What is digital marketing search on mobile.


SERP analysis like this can help you gauge the potential organic traffic from a query before you start creating content (or, if necessary, how to tweak existing content).

You can learn:

  • What content will target the right people with the right intent;
  • Whether you can compete—with SERP features or other sites.

Let’s walk through the process. 

How to do a SERP analysis (before you create content) 

The goal of this type of SERP analysis is to: 

  1. Confirm that your content plan matches the current SERP content (i.e. will satisfy intent);
  2. Determine if it’s possible to compete with other sites in the SERP (based on links).

1. Confirm that your content plan matches the current SERP content.

Searching for “Apple” returns a SERP filled with SERP features all about Apple, the technology company. Zero results are about Apple the fruit. Search for “buy Apple,” however, and you’ll get a SERP geared toward purchasing (not learning about the company).

The SERP features that appear are clues to the intent behind the search. If you target the wrong intent with your content, you’ll have a difficult time ranking and, even if you do, any traffic that arrives is more likely to bounce, which has its own (negative) SEO impact.

Search intent.

Search results are Google “showing its hand”—letting you in on what it’s learned about the intent behind a search. If a particular feature dominates results, your content should follow the same form and answer the same questions.

Small query shifts, big SERP impact

Subtle changes to query language can dramatically change the perceived intent. Take the search “my SEO sucks,” for example. The first result is an SEO agency (My SEO Sucks), and no featured snippet shows up. 

But a slightly different version of that search, “why does my SEO suck,” returns a Moz article from 2008, and a “People also ask” box shows up. Search “my website’s SEO sucks” and ads, a Quora answer, a “People also ask” box, and a related search feature listing SEO tools all appear.

Search for something that provides Google even less context, like “SEO,” and you’ll get a mix of results that serve varying intent. Google is essentially saying, “I don’t know what you want, so here are several options.”

If your SERP analysis returns a SERP with mixed intent, you probably haven’t chosen a specific-enough target—those mixed intent results are a mash-up of the best results that focus on a more specific intent. Rather than targeting the broad query, you’ll likely benefit from focusing your content in something more long tail (which will still have the potential to rank for that broader query).

2. Determine if it’s possible to compete with other sites in the SERP.

Once you know your term’s intent, the subsequent analysis determines how competitive it is for sites to rank. You can assess your ability to rank by looking at two primary metrics:

  1. Domain-level link metrics like Moz Domain Authority or Ahrefs Domain Rating (DR).
  2. Page-level link metrics like Moz Page Authority or Ahrefs URL Rating.

It’s not an exact science, but you should expect higher competition for queries that have greater search volume and/or stronger buying intent.

If the DR of the top sites for the keyword you’re targeting are all above 90, you’re going to have an extremely hard time trying to break into that SERP—unless you have a really strong backlink profile yourself. And there’s no sense targeting a term for which you’ll never see the light of Page 1. 

For example, the SERP for “how to do a push up” is dominated by authoritative sites, with The New York Times (not pictured below) claiming the featured snippet, and wikiHow, Men’s Health, and Nerd Fitness all high on the first page: 

How to do a pushup.

So what do you do? Target a related term (i.e. get in front of a segment of the same audience) without going up against the same competition. “Push up progression,” a quick SERP analysis reveals, receives significantly fewer searches a month, but the competition is far less stiff:

Push up progression.

A newer fitness site can get within striking distance of the top by creating content around a “push up progression.”

Tools to power your SERP analysis

Ahrefs Keyword Explorer is one of my favorites. Their SERP overview of the search “SEO tools” looks like this:

Ahrefs SERP overview.

From this dashboard, you can see all the competitive information about the current SERP to estimate what it will take to rank.

Moz and SEMRush also have useful keyword research tools that lay out SERP features and competitors. Ahrefs, Moz, and SEOquake all have browser extensions to show metrics for each result in SERPs, so you can analyze SERP features at the same time you assess competitiveness.

By this point, you should know the type of content that will match user intent and feel good that you can compete with the other sites in the SERP. It’s time to apply your analysis to how you create or improve the page you want to rank.

7 ways to get more traffic from SERP features

Making your search snippets more clickable will get you more clicks from the exact same position in search results. Here’s how to make the most of the SERP features that show up for the keyword you’re targeting.

Featured Snippet.

Featured Snippets answer specific search queries and show up at the top of Google search results. As they’re part of organic results, featured snippets are often referred to as “Position 0.” While there’s no specific strategy that guarantees you’ll land (and keep) one, there are a few things you can do.

For example, let’s take the featured snippet (from the Freshbooks Small Business Resource Hub) that appears for the query “straight line depreciation.”

If you click into the article on Chrome, you can see that Google now highlights the text that answers the question. (This relates to Google’s improved ability to rank passages.)

What is straight line depreciation?

You can see that Google highlighted the relevant content. A clear, short summary—at the top of a page dedicated to the topic—makes it much more likely that Google will deem your content the best fit for a featured snippet. As Moz notes: “The optimal length of a featured snippet paragraph is roughly 40 to 50 words, or around 300 characters.”

In this instance, the entire page focuses on answering the question, “What is straight line depreciation?” This makes it clear to search engines what the page is about and, as a result, more relevant to people searching for “straight line depreciation.”

2. People Also Ask (PAA) boxes

PPA box.

A PPA box lists questions related to a search query and, below, answers from a page with a link to the source. These boxes show up a lot in SERPs, but they don’t provide much visibility for publishers (compared to featured snippets) because they require a couple of extra clicks. 

Straight line depreciation search.

Even so, PAA boxes give you an idea of the related questions that you may want to answer on your page (to, for example, make it more likely that you’ll earn the featured snippet or simply rank higher). Alternatively, they may also give you an idea of new pages you could create as stand-alone articles or as part of a content hub.

A great way to find pages that you rank for but don’t show up in PAA boxes is to use the Ahrefs Organic Keywords section in their Site Explorer. You can filter results by SERP features and select “People Also Ask.”

Ahrefs research.

For example, if you’re Freshbooks, you can see that you own the first position for “straight line depreciation” but are just sixth for “balance sheet.” Investopedia is a tough competitor from a link perspective—but maybe a sharper focus for your content could improve your position.

The PAA boxes below the featured snippet give you some ideas of what you might need to cover on your page. Do you already answer these questions? As well as Investopedia? Are the answers clearly defined so that search engines can find them?

Balance sheet PAA.

 3. Video carousels and clips

The video carousel gives users an option to preview videos relevant to their search. Clips, currently in (Google’s often prolonged) beta, let users navigate directly to specific points in a video

Video carousels example A.

Let’s get meta by looking at HubSpot’s video on how to optimize your YouTube videos for search.

Video carousels example 2A.

You see the common theme across the thumbnails showing up here? They all include a person, a big title, and a colorful background. Also notice that the top two titles are most aligned with the words in our query. That’s not by accident.

For relevant phrases that don’t make it into the title, tags can help. Tags are added by the uploader of a video and are not seen by users—but are easy to find. To look at the keywords a competitor video uses, search the source of the page for “keywords”:

Video tags.
The video on how to optimize your YouTube videos for search uses the keywords “how to optimize youtube video, hubspot, marketing, business, SEO, social, media, blog, growth, youtube, youtube seo, how to rank on youtube.”

For best results, add around a dozen specific (“how to optimize youtube video”) and broad (“youtube SEO”) tags to your video that sum up what it’s about. It’s a good sign that your tags are fully optimized if you can understand what your video is about just by reading them.

Don’t go overboard with your tags. About 5–8 is plenty. Including too many is counterproductive, as you’ll send mixed signals to YouTube’s algorithm about your video topic. 

4. FAQ schema

FAQ result.

Frequently Asked Questions (FAQ) display a drop-down list of questions and answers from an organic search result.

Here’s a brief walkthrough on how to add FAQ schema to your site:

When it comes to developing content, you can find questions to answer on your page a few different ways:

  • Google’s autocomplete & People Also Ask section;
  • Quora;
  • Answer The Public;
  • Internal site search.

Add only relevant FAQs and keep your answers short and straightforward 

5. Map Pack

Map pack search result.

The Google Map Pack (or Local Pack) appears for local-intent keywords and lists two or three local businesses related to the search. It links to profiles with photos, reviews, and other information about the local businesses.

Local SEO is its own beast, but here are some starting points:

  • Verify your business on Google.
  • Fill out all the information on your Google My Business profile.
  • Get online reviews (+ respond to reviews).
  • Build local citations (with a consistent name, phone, and address).
  • Build backlinks (local links + relevant topic links).

If we take a peek at the SERP for “wrongful foreclosure attorney near me,” Lawsuit Legal appears at the top of the “regular” organic results but is nowhere to be found in the map pack:

Google Map Pack.

Even if you rank first organically, you’ll lose out on traffic (especially on mobile) if you don’t claim and optimize your Google My Business profile. Thirty minutes of work on the LawsuitLegal profile could yield thousands in new business.

Google My Business.

6. Images

Images appear in multiple places on the SERP—anywhere Google thinks visual content provides better results. Images can appear in rows, blocks, and even between organic listings. (Side note: Always use images in your blog posts as images can boost your conversion rate.)

Image SERP.

 Want to get more organic traffic from images? Here’s your punch list:

Sitelinks show up for brand queries, so it’s a space you probably already own. However, optimizing your sitelinks can help get users to relevant pages in fewer clicks.

Sitelinks.

Google’s advice to improve sitelinks highlights a few standard SEO best practices:

Provide a clear structure for your website, using relevant internal links and anchor text that’s informative, compact, and avoids repetition.

Allow Google to crawl and index important pages within your site. Use Fetch and Render to check that they can be rendered properly.

If you need to remove a page from search completely, use a “noindex” robots meta tag on that page.

If relevant, high-value pages aren’t showing up as sitelinks, you may want to:

  • Review your website’s structure and ensure the navigation is clear.
  • Check that top pages are in your XML sitemap.
  • Build more internal links to key pages.
  • Cross check that your page titles are accurate and descriptive.

Conclusion

As Google continuously pushes out new features in search results, keyword research alone isn’t enough—you need to pull data from real-world SERPs.

A SERP analysis needs to be part of every piece of content you create. You need to know what people expect when they use search terms and whether the content you create is competitive with what’s already there.

Keeping up to date and monitoring SERPs doesn’t have to be complicated. Here are a few accounts and groups to make it easy:

The post How to Analyze SERPs to Win Big in Rankings appeared first on CXL.

3 Smart, Sustainable Ways to Monitor Your Organic Search Rankings

“Why don’t we rank first for [keyword]”? Every SEO analyst gets this question. And every business investing in SEO uses keyword rankings to judge performance. Despite the flood of “organic rankings are dead” articles in recent years, I have yet to see a single business or agency that has given up on tracking keywords. So […]

The post 3 Smart, Sustainable Ways to Monitor Your Organic Search Rankings appeared first on CXL.

“Why don’t we rank first for [keyword]”? Every SEO analyst gets this question. And every business investing in SEO uses keyword rankings to judge performance.

Despite the flood of “organic rankings are dead” articles in recent years, I have yet to see a single business or agency that has given up on tracking keywords.

So are all those articles wrong? Or is everyone paying attention to the wrong metric? What makes sense when it comes to rank tracking in 2021 and beyond?

Monitoring rankings still makes sense—even though it’s gotten a lot harder.

SEO is one tiny piece of the larger marketing puzzle. It’s about making your site accessible to a search crawler and, in turn, easily discoverable for users.

While improved rankings don’t necessarily translate into more traffic (or conversions), it’s easy for down-funnel metrics like engagement or leads to seem beyond the scope of your work, or at least secondary. Your job as an SEO, you think, is to help a site show up higher in results—if you’re good at your job, the site will rank higher.

Analytics.

Being able to prove value, of course, requires being able to measure progress. That’s gotten more difficult—and not just because of “(not provided).” Google has been advancing its search algorithm for years, introducing personalization, localization, multi-format search elements, and 0-click SERPs.

At some point, knowing your actual organic positions became not merely ineffective (i.e. higher rankings didn’t lead to more leads and sales) but also almost impossible:

  • How can you even determine your page’s current search position if you—or your preferred rank-tracking tool—displays a different set of search engine results than what your target audience may see? 
  • How, then, can you evaluate the effectiveness of your SEO campaign? How do you know if SEO efforts are paying off if you can’t tell whether your positions are improving? 

The solution is data blending—combining several datasets to create a new dataset that can deliver confidence in how your rankings are trending, even if absolute, pinpoint accuracy for your rank for a given keyword remains forever elusive. (After all, if different people get different results, there is no single rank.)

When it comes to rankings, we have a few sources that provide reliable data:

  • Google’s own tools (mainly Google Search Console);
  • Third-party rank trackers that have successfully bypassed barriers like personalization and localization;
  • Third-party web analytics platforms that still show search-query data.

Using several data sources and blending the data will get you closer to assessing the value of your SEO work. Here are a few ways to do it.

1. Use a third-party tool to compare and trend GSC data.

Ever since Google Analytics locked out its keyword data as “(not provided),” Google Search Console (GSC) has been the only reliable source of ranking data.

GSC does provide free, in-depth insights, which you can play with and even integrate into your WordPress backend by using a plugin. (You can also use GSC data to identify which sitelinks Google shows for brand queries.) But its functionality is sadly limited, especially when it comes to trending ranking data.

Which keywords are gaining or losing traction? Which pages have been gaining or losing traffic? For how long has any particular traffic trend gone on?

You can make simple comparisons of positions and organic clicks between two periods, but building a more complex report to answer the questions above is likely to take you hours. 

Third-party tools that integrate with GSC can help. SE Ranking is a multi-feature SEO suite that includes many neat features, like monitoring extra search elements (e.g., featured snippets, image carousels, and organic sitelinks).

The nice thing about this platform is that it blends its own ranking data with GSC data, allowing you to:

  • Compare the tool’s findings with GSC’s average position. They will always differ because Google shows an average position, and you’re monitoring rankings within a particular location (e.g., country or city). But comparing the two datasets allows you to estimate your visibility more accurately and validate the rankings you see.
  • Clearly monitor your trend for each query. Expect the two data sets to mostly agree: You’re unlikely to decline in your average position according to GSC while gaining in SE Ranking. (And, if you are, you know something is up that you otherwise would’ve missed.)
SE Ranking position.
See SE Ranking position + recent trend vs. GSC average position + recent trend to create a more reliable organic visibility report.

Once you identify a trend, you can create your plan of action:

  • Do nothing and watch, which makes sense when you see +1 / -1 types of “natural” variations.
  • Diagnose a drastic change (when both datasets agree). I consider a loss of more than 10 positions a huge change but only if the initial position was within the top 10. Otherwise, I wouldn’t spend my time on it. Here’s a pretty easy, yet solid, guide on diagnosing position loss.
  • Check if pages that are trending positively have strong CTAs, forms that work, and no broken links to interrupt a (hopefully) increasing swath of visitors.

2. Reconnect keyword data to on-site behavior.

While Google Analytics is pretty much the default analytics solution for most websites, some independent platforms can add some missing data points—like the queries your organic users ran to end up on your site.

Finteza is one such platform. The tool starts collecting and showing data immediately after you install the script.

Editor’s note: When asked about how Finteza collects the referring keyword data from search engines, their team said: “Finteza gathers data from many Search Engines (Google, Bing, Yahoo, Yandex, Baidu and so on) to provide a complete picture [. . .] there’s no any estimation inside Finteza. System is showing all gathered details.”

You can access your traffic-driving keywords by:

  • Going to the “Sources” section and clicking “Search.”
  • Clicking “Search keyword” tab.
Search Keyword.

The default search keyword report consolidates data from all search engines you appear in. This is helpful because you can get a better picture of all the possible ways your site can be found.

To get a more detailed look into any keyword’s performance, click it in the report and proceed to other sections of the analytics platform. All further reports are limited to the traffic driven by that particular keyword.

This is a great way to identify your best-performing keywords. For example, you can identify how any particular search query performs at various stages within your conversion funnel:

Finteza traffic stats.
Finteza allows you to monitor dozens of funnels that show how your traffic performs at each step of the buying journey.

While digging into your highest-traffic keywords is a nice exercise, I like to get a higher-level look at my keyword performance. For that:

  • Click “Events” to access the list of all on-site actions you’re tracking.
  • Select one of the events by clicking it.
  • Head back to Sources > Search and click the “Search keyword” report.

This report will now include the number of conversions for the selected event:

Events for traffic.

You can export this report as a CSV, then import it into SE Ranking (or your preferred rank tracker) to monitor it closely. I usually group these keywords under a separate cluster (e.g., “best converting” to monitor them closely).

SE Ranking tool.

There might be a better way to combine Finteza’s data with your rank tracking solution, but I’m not aware of one, so I manually combine these two sets.

3. Identify (and improve the performance of) high-ranking keywords that drive few or no clicks.

Any reputable SEO will tell you that rankings are useless unless they drive (quality) traffic. If your pages rank without generating meaningful clicks, you’ve got a problem to fix.

Here’s how.

The easiest way to find your site URLs that drive few or no clicks from search results is to use GSC:

  • Log in to your GSC dashboard and click through to the “Performance” section.
  • Click to the “Pages” tab.
  • Sort results by “Clicks” to surface pages that drive no clicks. (You can filter pages for a baseline number of Impressions to focus on pages that show up often but don’t win clicks.)
Google SearchConsole click stats.

Unfortunately, as it often happens, GSC data is nice to have, but it’s not really actionable. It’s difficult to visualize this data across your site to help prioritize your actions.

Luckily, there are tools to help. 

Jet Octopus is an online SEO crawler that can process thousands of URLs without killing your computer. The tool also integrates with GSC, allowing you to blend its data with your site structure and trends. (Screaming Frog integrates with GSC, too, but it may eat up all your RAM.)

The first step is letting the tool crawl your entire site which, surprisingly, will not take much time, even for huge sites. 

When setting up your crawl, you can connect your GSC account in a couple of clicks. Based on my experience, you should be able to access your blended data the same day. It may take about 24 hours for extremely large sites (more than 100K pages).

Next, click the tab called “GSC Keywords” and browse around. You’re likely to find many useful reports here. My favorite section is called “Data tables,” where you can locate some useful reports like the “Cannibalization” section (i.e. pages fighting to rank for the same search queries) and “New pages” (URLs recently found in GSC reports).

GSC keywords.

But my personal favorite report here is called “Zero click pages.” This one is a solid roadmap for your optimization strategy.

As the name suggests, “Zero click pages” are pages that appear in search results but don’t drive any clicks. Compared to the GSC report, this one includes a variety of filtering and sorting options to help you dig as deep as your site requires.

For example:

Count total queries.

Unlike GSC, the tool shows the total number of search queries for which any page ranks. This is helpful in a few ways, including:

  • Estimating how many queries you can do better for (the higher the number, the higher the odds the page will be able to drive more traffic if it ranks higher).
  • Evaluating your true average position at a glance. If your average position is 30, and there are only 5 search queries, you know that it likely ranks on Page 2 or 3 for all of those. If your average position is 30 but the page ranks for 90 queries, some of those may be on Page 1.
Query count stats.
Pages with an average position of 20-something are already doing well. It may take just a little effort to get them to drive organic clicks.

See the trend.

Which one of those zero-click pages is losing or gaining in rankings? Depending on the answer, your actions may be completely different.

If a page isn’t bringing in any search traffic and keeps losing positions, I wouldn’t even touch it. I’d rather wait until it stabilizes (or even consider getting rid of it, perhaps by merging it with a different page).

On the other hand, if the page is gaining in rankings yet delivering zero clicks, there’s hope. I’d definitely look closely at that page, its current search queries, and see if I need to expand and improve the copy to accelerate its growth and finally generate clicks.

Clicking these numbers under “Queries count” will show exactly which queries for which these pages are gaining search impressions.

Sending internal links to a page is one of the easiest ways to improve its organic search visibility. As such, if a page generates no search traffic, one of the first things we look at is whether it has enough internal links.

Jet Octopus lets you quickly identify pages with no internal links. All you need to do is to blend GSC data with your recent crawl data by clicking the “In Links” option, then changing the filter to “URL is NOT present in In Links.”

Click Apply, and you have a list of pages that have no internal links and send zero clicks from search:

Join dataset.
  • In many cases, those pages can be removed. (Since they have no clicks and send no traffic, chances are, they’re useless.)
  • In some cases, those pages are an easy fix. Send them some internal links (e.g., from your menu), and you’ll see their organic traffic grow.

Likewise, you can export your zero-click queries to monitor them in your rank tracking solution. (Again, I recommend creating a separate group to narrow down your rank tracking report to those keywords.)

Conclusion

SEO data is often scattered around various tools and dashboards, limiting your reporting and evaluation of SEO’s ROI. While an obsessive focus with a single keyword is outdated, ranking trends are still an important way to figure out if your SEO work is having an impact.

Fortunately, there are pretty doable ways to blend two or more data sets into one, which can help validate ranking data and make it easier to translate your analysis into impactful work.

The post 3 Smart, Sustainable Ways to Monitor Your Organic Search Rankings appeared first on CXL.