How to Grow Traffic with SEO Title Testing [In-Depth]

I’ll do you a favor, if you’re just here for the title tag optimization ideas, click here to skip to the ideas section.

For those of us who really want to understand title testing at a deep level, let’s dive into the good stuff.

First, a word of caution: this is a long and in-depth guide, so if you’re looking to get the gist of SEO title testing without having to read through all 8,000 words then head over to my shorter guide below.

The Simplified Guide to Title Testing for SEO

One of the most fascinating things about SEO is that there is a whole universe out there of infinite hypotheses which no one is testing right now. You can stand out from the noise by finding a few and testing them.

If you already understand title tags, you know these elements are right smack-dab at the very top of the SEO food-chain. Along with content and links, page titles are the apex predators. The King Kongs. The great white sharks of SEO optimizations.

Executed correctly, testing your page titles will do more damage than a spicy deep-fried chimichanga in an all-you-can-eat buffet.

More affectionately though, I like to call title tags SEO’s lowest-hanging fruit.

Why is Title Tag Testing So Pivotal for SEO?

Shark references and chimichangas aside, I do believe that title tag optimization remains SEO’s lowest-hanging fruit opportunity, even now as Google has taken more control of title tag representation in SERPs. (more on that topic here).

Of course, the actual lowest-hanging fruit opportunities can change from website to website, so title tag testing is not a universal top priority. Generally speaking, testing title tags is one place where I can get very quick traffic gains with very low time investment.

Here are a few more reasons to conduct title tag tests:

1. Page titles are still considered one of (if not “the”) top on-page ranking factors, next to content.
2. Titles are one of the fastest & easiest optimizations to execute.
3. They’re the very first thing people see when discovering your page in search.
4. Optimizing meta titles has direct engagement with many of the most exciting elements of SEO: user-intent, SERP analysis, keywords, analytics, and much, much more.
5. Page titles have extremely high impacts on click-through-rates.
6. They provide your website an opportunity to stand out from the pack, and increase your brand exposure.
7. Meta titles aren’t only useful for search engines and clicks. They appear on browser tabs, bookmarks, and social media shares.

When you put all of these things together, what you get is low-effort / high-reward SEO opportunities; the low-hanging fruit.

Doesn’t low-hanging fruit just taste delicious? 😋 

High-Impact Title Tag Tests in the Wild

If you’re looking for more proof, here’s a look at some very high-impact title tag test results out there in the wild, wild web.

• Coderwall implemented one title tag test that led to a 57% traffic increase.
• TrustRadius grew sitewide traffic by 22% with a single title tag test.
• Moz saw a 20% uplift by removing “Whiteboard Friday” as a boilerplate title insertion.
• SearchPilot reported one brand experiencing a 15% lift in organic traffic by placing the company brand at the beginning of their meta titles.
• Elise Dopson increased her search traffic by 54% through meta tag optimization alone.
• Primary Arms increased Organic REVENUE by 47% through title testing.

How to Test Your Titles: The Basic Process

Alright, now for the nitty-gritty. Here’s what a basic testing process looks like for a single-URL test. (Note: It’s possible to achieve even higher impacts by scaling this process across hundreds, or thousands of URLs.)

1. Identify your top title optimization opportunities and prioritize them by URL.
2. Conduct SERP analysis on the top keyword(s) to understand search intent and analyze the top-ranking pages.
3. Build a crystal-clear hypothesis for your test.
4. Write a new, optimized page title to support your hypothesis.
5. Launch your test by publishing the new title and submit the page URL to Google Search Console for re-indexing.
6. Benchmark your current performance metrics before executing the test.
7. Wait at least 2-3 weeks.
8. Analyze your results & end the test.

Key callout here: I can’t stress enough how important it is to benchmark your tests before optimizing. Good record-keeping goes a long, long way.

1. Identifying & Prioritizing Title Tag Test Opportunities

In reality, there are dozens of ways to discover title testing opportunities on a website but these two methods are most common.

Method 1: Best for under-optimized websites.

1. Build a spreadsheet with your website’s URLs and corresponding page titles.
   1. Usually by crawling your website with a tool like ScreamingFrog, Sitebulb, or other site crawler.
   2. Alternatively, you can use tools like Ahrefs, Moz, or your website’s CMS in some cases.
2. If you have relevant keyword data to easily add in, definitely do.
3. Scan for the obvious opportunities, missing titles, bad titles, short titles, missing keywords, etc.

Method 2: Best for mature, previously-optimized websites.

1. Go to your performance report in Google Search Console.
2. Make sure to click on the “Average CTR” & “Average position” buttons.
3. Export your report to Excel or Google Sheets.
4. On the “queries” tab, add an “Available Clicks” column that subtracts the “Clicks” column from the “Impressions” column.
5. This new column will not give you the URLs to optimize, but it will give you the best queries (primary keywords) to prioritize your optimizations around.
6. From here, you could do more work to map the primary keywords to your URLs, but I recommend most people stop at this step, prioritize by keyword instead of URL, and use that extra time-savings to go straight to the title optimizations.

2. SERP Analysis for Title Testing & Optimization

Far too many people skip this step in the title testing process. Sometimes it needs to be skipped for the sake of saving time, but it’s one thing to skip SERP analysis intentionally and it’s a whole other thing to skip it out of a lack of experience.

Trust me, it is important. Especially for high-competition keywords and search results.

Just doing a bit of SERP analysis can take your testing hypothesis from good to GREAT!

Questions to ask yourself while analyzing the SERPs

• What major trends are you seeing in titles for the top 10 results? Top 3?
• Can your website compete for the keyword?
• What are the page types ranking for this keyword? Are they blog pages? Product pages? Evergreen content? UGC?
• Is the content of your page relevant to the search intent(s) surrounding the keyword? (I.e. Does your page deliver on the users’ questions or needs?)
• Are the competing pages ranking well due to relevancy, or authority, or both?
• What SERP features is Google displaying outside of the ten blue links?
• Is the search intent branded or unbranded?
• Is the search audience mainly top of funnel, middle of funnel, or bottom of funnel?

Getting good at SERP analysis really boils down to observation skills. You’re looking for commonalities in the results, and you’re also looking for edge cases. You can spend a lot of time here, or a little time. Even a quick glance to see what searchers are looking for is better than no SERP analysis at all.

More questions to think about when analyzing:

• What value can your page or website offer that others cannot?
• Can you craft a title that speaks more directly to users’ search intent than the competing page titles?
• What would make your title tag stand apart from the crowd?
• What ideas can you borrow from your competitors’ title tags?

3. Establishing a Clear & Measurable Hypothesis for Your Test

“There’s two possible outcomes: if the result confirms the hypothesis, then you’ve made a measurement. If the result is contrary to the hypothesis, then you’ve made a discovery.”

– Enrico Fermi

Changing your page titles without creating a hypothesis is a lot like throwing spaghetti at the wall and hoping it sticks. Sure, sometimes this method works, but even in those cases when the random test does work, you’ll learn far more with a clear hypothesis than you will without one.

The best and simplest formula for a great title testing hypothesis looks like this:

Changing my page meta title from X to Y will _____.

The two questions you should be asking yourself right now are: What exactly are you aiming to achieve with your test? And how are you going to measure it?

Here are the most common testing hypotheses I’ve used:

• Changing my title from X to Y will help my page rank higher for the primary keyword.
• Changing my title from X to Y will increase CTR for my primary keyword.
• Changing my title from X to Y will increase the average CTR across all keywords for the page.
• Changing my title from X to Y will increase the total number of clicks to my landing page.

But, you might think, “why would I want my hypothesis to be anything other than ranking higher?”

Well, in some cases you’ll find that ranking higher for your target keyword results in fewer clicks across the long-tail of your page’s keywords. In other cases, you might set up a test to learn which specific messages and value props resonate most with your audience, which means CTR could be used as your primary success measurement.

Measuring Your Test

One more benefit your hypothesis brings is the ability to clarify exactly which metrics you will use for benchmarking and measuring the test results.

Are you aiming for higher rankings on a single keyword? Benchmark & measure against the target keyword ranking.

Is overall URL traffic more important to you than the primary keyword? Benchmark and measure clicks to your URL.

Are you testing to improve CTR? Same deal for CTR.

If you’ve really got your solid conversion attribution & analytics in place, you can even test for conversions in scenarios where your page receives a high enough volume of data to support this level of measurement.

For the vast majority of tests, the Google Search Console Performance Report is the absolute best data source for benchmarking and measuring clicks, impressions, CTR, and keyword positions because this is the closest available source of data to the live test environment (SERPs).

4. How to Write KILLER Title Tag Experiments

Writing a KILLER title tag is both art and science. Experimentation is the science part, creative thinking grounded in best-practices is the art.

Like any form of art, there is no “right way” to optimize a title tag, but there are a zillion ways to screw one up. So before we dive into best practices and new optimization ideas, here are a few of the most common mistakes people make with page titles.

• Your title tag is missing/empty. (pretty obvious)
• Your title tag is devoid of your pages primary keyword.
• Your title tag is stuffing keywords.
• Your title tag is auto-generated by your CMS to only include the page name. (ex. “Home – Domain.com”)
• Your title tag is duplicated across multiple pages. (When everyone is a hero, no one is.)
• Your title doesn’t match search intent.
• Your title doesn’t match the actual page content.

Character Count

“The best headlines never fi” – Bernard Levin

A very key point to note here: your meta title does not actually have to fit within a specific character count. Firstly, Google’s cutoff for title length is by pixels, not characters. Second, if your page title goes beyond Google’s pixel limitations, it will get truncated with an ellipses (…) but it will not count against your page’s SEO performance in any way other than potential click-through-rates. In fact, you may even want to test longer, truncated titles just to see if the added relevance in a longer title may improve the page’s rankings and performance.

With that in mind, Google will typically truncate page titles at the following pixel counts for each of the following devices:

• Desktop – 545 pixels
• iPhone X/6/7/8 – 375 pixels
• iPhone 6/7/8 Plus/Pixel 2 – 411 pixels
• iPhone 5/SE – 320 pixels
• Galaxy Note & S3+ – 360 pixels
• LG Optimus 370 – 384 pixels
• Nexus 5X – 412 pixels

Even though most professionals still prefer to keep titles below the 545 pixel range for desktop (or about 55-60 characters), you should definitely test this for yourself!

Pro tip: Try this bulk title tag checker to check the pixel width for a list of titles.

Okay, onto the good stuff…

Must-Do Best Practices For Any Decent Page Title

Don’t try to reinvent the wheel. The winning strategy for page titles is to do what has already been proven to work extremely well (the best practices) and then layer your own creative ideas in with good judgment.

At a bare, bare minimum, your titles should:

• Include your primary keyword
• Establish a powerful tie between the search intent and your on-page content
• Set the tone for what users are going to see when they get to your page
• Place your brand or website name at the end of the title tag with a separator

17 Ideas for Writing Title Tag Variants that Win in the SERPs

Beyond the bare minimum, these are some of our most killer best-practices and trends among powerful page titles:

1. Test keyword placement locations.

People navigate quickly on search, so if your keyword is at the end, it’s likely to get less attention than at the beginning, where people are more likely to make the connection between their search and your page relevance. There is also some debate about this as a ranking factor. Even if that isn’t the case today, re-prioritizing your keyword to the front can be a very positive optimization. Test this for yourself to see if your title performs better with your target keyword placed at the front, middle, or end of your title tag.

Example:

2. Use parentheses () or brackets [].

Parentheses and brackets are very strategic for adding context mid-title and as an added bonus, they often help your result get a little more attention by drawing the users’ eyes to your search result.

Example:

3. Deliver on a second promise to touch secondary search intents.

People sometimes search with a secondary search intent that the keyword itself won’t tell you. If you search for animal charities, you might also want to know which charities are the most effective. Or if you do a search for “what is keyword research” with intentions of starting your first keyword research project.

Searching this way means you’ve got an explicit search intent that is obvious from the keyword alone, but there might also be a less-obvious implicit search intent that compliments the explicit search intent. If you can go above and beyond by discovering and addressing these secondary implicit intents in corresponding page titles, this test has a tendency to earn and dominate top positions.

Example:

4. Add keyword modifiers to grow the long-tail.

Beyond every primary keyword is a whole world of long-tail variations. These variations often get missed when you optimize for one keyword, and one keyword only. To find them, try digging into more research tools or check your Google Search Console performance report to decipher which modifiers are most-commonly-searched. Then add one to your title to see if you can capture more long-tail traffic.

Example:

5. Add synonyms and keyword variants (but don’t overdo it).

This one can get dicey if you’re overdoing it, so let’s be very clear: DO NOT KEYWORD STUFF. At most, you should only work in one, maaaaybe two keyword variants.

Reasons for running a test like this one could be that your keyword has a synonym worth mentioning, or you might want to see if you can get more long-tail keywords by adding in a related term, or you might be aiming to see if Google will rank your page higher on the primary keyword by adding a variant to the mix. Just be sure to test this approach with readability for the user in mind, and don’t over-optimize.

Example:

6. Use numbers.

Adding a number to your page title has been proven by SEOs and content marketers alike to draw people’s attention and induce higher click-through-rates.

Example:

7. Include a CTA.

Just like you would add a CTA on your web page to entice users to take action, CTAs often work well in page titles!

Example:

8. Add a phone number (if your business relies on calls).

For businesses that rely on phone calls, adding a phone number to title tags has been a highly-effective way to achieve more direct calls from your website. The downside to this is that some users might not click through to your website by calling your business directly. The upside though, is of course more calls to your business.

Example:

*Pro tip: leverage a call-tracking service, like CallRail to create and track unique phone numbers across a wide array of your website’s landing pages and test title tag phone numbers in bulk.

9. Spark emotion (and where appropriate, negative emotion).

Marketers and the media have been using emotional language in headlines, titles, and news stories for centuries because it’s a writing style that works. Find an emotion that suits the response that you want to induce with your audience. Do you want them to feel joy? Excitement? Humor? Lust? Pick an emotion, any emotion.

You may be tempted to stick to positive emotions for the sake of brand protection, but in some cases, you can make really powerful use of negative emotion in a title tag without damaging your brand reputation at all. And since humans are hardwired to respond more to negative emotion than we are to positive emotion, you just might strike gold with a carefully-crafted negativity line.

Example:

10. Tap into FOMO.

No one likes to be the last kid picked for the soccer team. Why is that? Because FOMO. We’re evolutionarily wired for FOMO as one of many social bonding survival mechanisms. Knowing this, you can use FOMO in several ways to spice up your title tags.

Example:

This idea works incredibly well. So well, in-fact, that it may just be my favorite test due to how successful it’s been. That second promise (& How to Avoid Them) is the FOMO element. It says, “if you don’t learn this, you’ll miss out on something that could really save your SEO strategy.” It also utilizes our 3rd tip, “deliver on a second promise.”

You can find more ideas for FOMO testing from Crazy Egg’s FOMO testing ideas article.

11. Give them something extra (for free).

You may need to do more work here than just a title change, but this idea works insanely-well. Pair your article with something that has related value, like a free downloadable template, a PDF, a video, or anything that can compliment the primary search intent.

Example:

12. Incite curiosity.

Curiosity is a very similar concept to the FOMO element. Marketers call this, “the curiosity gap.” As the name implies, it ‘s the gap between what you know and what you don’t know. Here’s an example:

Example:

If you’re a frequent traveler, you probably already want to figure out whatever it is that the airlines are keeping secret from you.

13. Add The Current Year (and/or Month).

Another tried and true title optimization tactic. Not only does this give you an opportunity to add numbers to your page titles, it demonstrates freshness and timeliness. People are hungry for content that is up-to-date. The one downside here is that if you’re going to employ this tactic, you should be prepared to actually update your content from year to year.

Example:

14. Use superlatives. (best, top, easiest, worst, most, highest, etc.)

This is your opportunity to invoke your inner Donald Trump and write the best, most unbelievable title tag anyone has ever seen! Why do superlatives work so well? Because best-in-class sells. People don’t filter Amazon for products with mediocre reviews.They want the best. You can address people’s innate desire to seek out the extremes by adding superlatives.

Example:

15. Test separators between phrases or segments. (|, :, -,  •, ::, ›, », .)

They may seem insignificant, but these separators can really make a difference in helping each separate thought stand out within a single title tag. Additionally, these can vary in pixel space, and different separators may stand out over others.

Example:

16. Use all caps/uppercase (1-2 words max).

Disclaimer: Do not use all caps for the whole title. Full uppercase titles are really not found anywhere on page one that we’ve seen, so it will probably hurt your rankings or at the very best, Google will replace it with an auto-generated title.

That said, you can definitely craft a very strong title using all caps in one, maybe two words max to get that little bit of extra attention.

Example:

17. Try emoticons, emoji, & unicode characters.

These tests can come with some downside. The biggest challenge is that there are a lot of queries that Google will not show emojis for, so you won’t always get them to display as intended. Another downside is that they aren’t always extremely effective in terms of generating higher performance. Still, I would encourage you to test them out for yourself.

Here are some examples of emoticons, emojis, and unicode characters for page titles.

An emoticon is a set of characters that combine together to form a picture. (¯\(ツ)/¯, :‑D, \o/ )

Example:

Credit to Northward for this very clever emoticon title.

An emoji is one of these things that I’m sure you’ve abused in your text & slack conversations. (🤣, 🥇, 💯)

Example:

Note: Emojis have a VERY high tendency to get stripped from titles more often than emoticons and unicode characters when the title is displayed in Google’s search results. If you find that your emojis are being stripped, then a great alternative location is to test emojis in your meta description instead. 

Unicode covers all the basic characters in the writing system as a standard across all languages. Some unicode characters function very similar to emojis, only a bit more analog in design. (✆, ✈, ★, ➝, ✓)

Example:

Pro Tip: Use a Pixel Counting Tool to Visualize Your New Title

I can’t remember how many times I’ve written a new title, only to realize it looks terrible in a live SERP environment. This SERP Preview tool from Portent is my go-to for visualizing my title optimizations before I set them live. I’ve caught DOZENS of mistakes and improvement opportunities just by dropping my optimized title into this tool.

It’s designed to help you check the length for truncation limitations, but in reality the benefit of doing this goes well-beyond the basic length checks. This habit is also where I’ve gotten some of my very best testing ideas.

5. Launching Your Test 🚀

Finally, it’s execution time!

Launching the test is extremely simple (for most websites), but this is definitely where the whole process gets even more fun.

Here’s what you’ll do:

1. Publish the new, optimized title on your website’s CMS (or request this from your developer if you can’t set it live, yourself).
2. Head over to your Google Search Console account and copy/paste the test page URL into the URL Inspection Tool.
3. Click on “Request Indexing.”

4. Return to your benchmarking sheet to indicate the launch date and GSC URL submission.

Houston, we have liftoff! 🚀

6. Benchmarking & Tracking Performance

Thinking back to Step 3 where we established a clear and measurable hypothesis, you should have already established the criteria for which data point(s) you’re going to use to measure your test. To recap, the most common measurement KPIs are:

1. Ranking (or position) of your primary keyword.
2. Overall clicks to your target URL.
3. Average CTR for your target keyword or URL.
4. Conversion actions or custom events.

In most cases, all of these KPIs will grow together, but it’s best to select one metric as your primary KPI for benchmarking and measurement purposes. For the sake of this article, we’re going to go with the most popular test, “ranking higher for your primary keyword.”

Ideally, it’s best to keep an eye on all angles of the test’s performance. 

How to Benchmark

Digital marketing professionals have varying ways in which we like to do benchmarking. Some don’t benchmark their data at all (not recommended) and some benchmark with a quick GA annotation (quick, but harder to track).

The best way I’ve found is to have a running backlog of past, current, and future experiments so that you can view your tests in a centralized location and maximize your learnings for future tests.

Annotating in GA

This is how I did benchmarking when I first started with title testing, and honestly, it worked extremely well in the early stages. Google Analytics annotations are fast, reliable, and a good practice even if you do decide to benchmark with more refined processes. There were just a few downsides.

• First, annotating isn’t really benchmarking. It’s annotating. You can still analyze the data on your own, but that information is easily lost after you’ve taken a look and decided on a new title. What’s missing in that equation is the ability to go back in time (well into the future) and remember why you chose that particular title, or what the data was saying.
• Another big limitation is that GSC restricts all historical data to the last 16 months. If you ever want to go back and see the data more than 16 months out, it’s tough noogies (at least for GSC data).
• Last but not least, GA is technically a centralized location, but it isn’t really all that easy to see multiple test results in a single view. Anytime you want to understand your tests, you’ve got to go back to discover them, one by one, and re-analyze them from scratch.

Recommendation: Build a Title Testing Log in Google Sheets

One of the best time-saving hacks for title testing is to create a backlog of testing ideas. The backlog saves a ton of time by allowing you to frontload the work that it takes to identify opportunities, run SERP analysis, establish hypotheses, and write new test variants. It also serves as your centralized location for measuring, tracking, and learning from all of your current and past title tests.

You could also build this in Airtable, Excel, Trello, or another customizable tracking tool of choice. But because Google Sheets is so customizable and so widely-used, it’s still one of the best go-to benchmarking platforms.

 

Your testing backlog should include these fields (at a minimum):

• Page URL – Use this to delineate tests and not your titles or H1s (which are more susceptible to change).
• Primary Keyword
• Original Title
• New (Optimized) Title
• Your Test Hypothesis – Changing my title from X to Y will ____.
• A place to track & remind you to submit the page to GSC
• GSC Submission Date – To be used as your test start date.
• Test end date – So that you know how long the test ran.
• Your KPI Benchmark – At the start of the test.
• Your KPI Result – After the test.
• A place to mark your conclusion (Positive, Negative, Neutral, or Insufficient)

You can also include more fields, but in my experience these are the most important fields to keep.

A Key Note on Benchmarking

Anytime you’re comparing before and after results, it’s important to make sure you use a comparable collection of “before” data. I’ve found this to be especially important whenever I track query positions in GSC. Most people’s natural tendency will be to jump into GSC and jot down whatever number is showing up as soon as they locate their primary keyword, but this is a critical error because GSC position data is being averaged across multiple device types, locations, and browsers. That means that the position data you see in the default date range is unlikely to match your test duration, and your comparison will be skewed.

A better plan is to establish your testing duration beforehand, set GSC to that same time duration for your “before” data, and record that number in your benchmarking sheet.

7. Waiting for Your Results ☕ / Experiment Durations

Now that you’ve launched and benchmarked your test, take a break! You’ve earned it.

How Long Should You Let Your Test Run?

The 2-Week Rule

There’s a rule that I use, called “the 2-week rule.”

Like the name? I made it, myself.

When you launch your title test, it’s not uncommon to start seeing some movements within days or hours of launching your test, but early results are not always the rule. Plus, Google’s algorithms require several days to learn about the change, before stabilizing the conditions to a more stable environment.

To account for these scenarios, the two week rule says that you never want to conclude any title test that has been indexed for less than 14 days.

That’s the minimum duration, not the standard.

A lot of tests will need 3 weeks, 4 weeks, or even longer before they reach a place of being able to demonstrate meaningful insights. It helps to set reminders so that you can check back in two to three weeks, but let your tests run longer if you don’t see any results within a 2-4 week time period.

8. Analyzing Your Results & Ending the Test

Two or more weeks have flown by and you’re ready to check your results. What’s next? How do you analyze the results?

First, reflect back to your original hypothesis and benchmarked KPI. In our case, it’s the primary keyword position.

Ideally, your test duration will line up closely with the same time frame as your benchmarked data for a (mostly) clean comparison.

Analyzing Your Results

1. Head back over to your GSC performance report.
2. Line up your before & after date ranges in the comparison tab and click “apply.”
3. Your before and after lines should line up like this:

If they don’t line up at the start and end points, your comparison dates are off, so go back and readjust the dates.

4. Make sure to click on the “Average position” square to apply the position data.

5. Locate your primary keyword and analyze the data!

Note: It may help to export the data to excel google sheets.

6. Record your after results and end date in the benchmarking spreadsheet.

Ending The Experiment

You’ve got a winner (hopefully it’s the optimized variant 🤞) and it’s time to decide where to go next.

If your newly-optimized title was a success

Most of the time, people like to let this one ride for a bit longer after the test has ended, but you can test again if you wish. My recommendation is to let it ride for at least another 4-8 weeks before testing a new variant.

You may even be satisfied enough with the results that you keep the page as-is and move onto other marketing initiatives.

If your test returns negative results

Negative results are extremely common, and they still provide a ton of value to learn new insights for future testing.

A negative result usually indicates that something about your optimized title just didn’t resonate with users, or with Google, or both. When this happens, most people will revert the page title back to the original and re-index the page so that you can bounce right back to the results that your page was generating before the test.

The gain from this experience is knowledge. What was it about your original title that worked so well? Or, what is it about your new title that didn’t work? There’s always something to learn from a negative test result, and you can apply these findings to future testing initiatives.

Conclusion

If you’ve read through to this point, give yourself a warm congratulations. You’re now fully up-to-speed with everything you need to know about title testing for SEO.

SEO testing takes time and patience, but when best practices are followed, the positive growth effects are a proven way to grow over the long run.

Most importantly, remember to follow the best-practices in this guide for benchmarking and monitoring your tests, so that you don’t accidentally create a scenario that leaves a negative result in play for too long.

Do those things well, and you’ll be title testing with the best of ‘em!

When you’re done, don’t forget to record your test results, and run new tests with fresh hypotheses.

—Bonus Content—

Advanced Title Testing Strategies

What was covered in the guide is an overview of how to run basic, one-off title tests. If you’ve got that down, and you want to branch out into some advanced strategies, these are some advanced strategies that SEOs also find effective.

Group Testing

Group testing is one of the most powerful ways to scale your title testing across hundreds, or thousands of URLs in one go. If your website has a large group of pages with dynamic (or templatized) page titles, you can run highly significant tests in a very short period of time by grouping these pages into a group test.

Let’s say you’re running a travel website that lists 1,000’s of vacation listings pages, each with the same templated format. Your page titles might look something like this, where the brackets dynamically interchange values for each page location.

Vacation Rooms Available in {CITY_NAME}, {ST} | YourSite.com

Running a group test on these pages is just a matter of re-optimizing the dynamic title format, setting the changes live with your developer or CMS, and then making a new sitemap submission to Google Search Console. The benchmarking and measurement of the test would occur on a group scale, rather than an individual URL level.

If you were to run a test on these pages, you might decide to see what happens if you add the starting price per night to the title, so the new title looks like this.

Vacation Rooms Available in {CITY_NAME}, {ST} from ${PRICE_LOW} Per Night | YourSite.com

The amount of information and statistical significance that you can get from running tests like this on a group level is powerful, almost to the point of addiction once you start seeing how quickly these tests run, how much data you can get back from them, and how big the results can be (for positive experiments).

Key point of consideration: whenever you are running a group test, you must have at least one dynamic variable that distinguishes each individual page from another. Without a dynamic element you’ll have a very big duplication problem on your hands.

Split Testing

Split testing works much the same way as group testing does, except instead of testing the entire group with one variant, you break the group into two or more sets so that you can test two or more variants. Most split tests will run a control group with the original and a variant group with the optimized title.

The reasons you might want to run a split testing experiment might be because:

• Testing all your titles together poses too much risk to the business if the test is negative.
• You need to create a more controlled testing environment where the time duration for your control group and your testing group is concurrent, as opposed to a before and after test. This could be especially important for businesses that experience varying degrees of seasonality, events, or other changes that would skew the results of a before and after test.
• You want to run multiple variants in your test instead of just one.

The execution of split tests is definitely trickier than a simple before and after group test, but it can be done with the help of a development team, or by using an SEO testing tool.

The #1 thing to keep in mind when you’re executing a split test is to split your groups out in such a way that each group is as equivalent as possible in terms of  content & performance. The page content may not always be identical in a group setting, but as an example, if your travel site is targeting two separate countries, you might not want to group the pages by region because each region is going to have a different # of pages, different traffic volumes, and different page content. So your split test could end up running 70% of the test on region A’s pages and only 30% of the test on region B’s pages. Not good.

A more effective split test could be to blend the pages based on clicks data. Here, you would split each region’s pages into two groups with similar levels of click volume. So you would have region A broken out into groups 1 and 2, and region B also gets broken out into groups 1 and 2. In both cases, the volume of clicks should be very close to 50% each between groups 1 and 2.

So the splits would look like this:

Region A, Group 1 = ~50% of total clicks in region A

Region A, Group 2 = ~50% of total clicks in region A

Region B, Group 1 = ~50% of total clicks in region B

Region B, Group 2 = ~50% of total clicks in region B

Finally, when that’s been accomplished, you would merge the 4 groups back into two groups for split testing.

Control Group = Region A, Group 1 + Region B, Group 1

Variant Group = Region A, Group 2 + Region B, Group 2

This method will set your split test up for a far more accurate experimentation environment than it would have if region was your only criteria for the split.

Optimizing for CTR

When you’re optimizing a title for CTR, the testing methodology can be a bit different than it would be when rankings.

If you’re trying to measure absolute CTR the process is much more straightforward than it would be for a test that aims to measure relative ctr.

What are absolute and relative CTRs?

Absolute CTR is position-agnostic. No matter where your keyword is ranking, the data counts for absolute CTR. This metric is a little less indicative of user-behavior, and much more indicative of the overall keyword (or URL) performance.

Relative CTR is position-dependent, meaning that if the ranking position changes, then the relative CTR needs to account for that new position.

An easier way to think about the two is like this. If you’ve got a page X that ranks consistently in the #1 position for keyword Y. Let’s assume that for keyword “Y” position #1 gets 25% of all clicks on a regular basis. Now, if we perform an optimization that generates 2% higher CTR than the original title and the ranking stays the same, then we’ve just improved the relative CTR (as well as the absolute CTR).

On the other hand, if the same page were to have ranked in position 2 before the test, and was receiving 16% of the clicks, but after the test we ranked in position 1 and the CTR went from 16% to 25%, then the absolute CTR would have improved due to the ranking increase, BUT the relative CTR would not have changed because the #1 position is still receiving 25% of the overall clicks.

Now, it’s very easy to test for absolute CTR because these tests don’t have to remove ranking changes from the equation, but unless your keyword stays extremely stable throughout the duration of the test, it’s much harder to test for relative CTR.

On the other hand, if you have a primary keyword that consistently ranks #1 across the board (such as a branded keyword) then it becomes easy to test for relative CTR. One great test to try out here might be brand messaging. Try testing two different messages on a keyword like this one. Whichever message increases your relative CTR is likely the one users are resonating with the most.

The Future of Title Testing: AI, Algorithm Changes, & More

Like everything else in the SEO world, title optimization is bound to change and evolve over time because of the constant changes and advancements to Google’s algorithms. That works today, won’t necessarily work tomorrow, and those page titles you tested last year won’t necessarily stay competitive this year.

What will the future of title testing look like? As Google tests and evolves their search engine results pages, how will page titles be displayed in the SERPs of the future?

The actual future is speculative of course, but there are some clues we can follow to piece together a few ideas.

Google’s History of Tweaks and Tests With Respect to Title Tags

The past isn’t always an indication of future tests, but on the other hand history can, and does repeat itself. Some of the changes we’ve seen to title tags have been:

• Adjusting the title length limitations for desktop and mobile.
• Testing the colors of titles as they appear in SERPS.
• More recently, Google has ramped up dynamically rewriting titles (though, it should be noted that the original titles are still the default for algorithmic ranking purposes).

Another source of information that some find valuable in understanding where Google could be headed is through following the patents that Google files. A patent like this one that Bill Slawski describes, has a pretty direct impact on SERP layout which may indicate future changes to sitelinks and snippets.

AI’s Impact on Title Tags

First of all, as we’re seeing Google dynamically rewrite page titles it’s clear that AI is already in the business of title tag optimization. Mostly on Google’s side of the equation.

But we could just as easily start to see more AI making its way into the webmaster side of title tag creation.

A major announcement in June of 2020 came about when OpenAI announced that users could request access to GPT-3, a highly-sophisticated API that uses natural language processing (NLP) to generate AI-based copy at levels that are nearly-indistinguishable from human written content.

The adoption of GPT-3 and AI-generated content is still in its infancy, but it isn’t hard to imagine a future in which AI-powered content begins to take hold across the web, and that includes page titles.

What to Do When Google Rewrites Your Page Title

As mentioned, Google has been taking more liberties lately with rewriting page titles that display on their results pages. It should first be noted Google is using your hard-coded page title to rank your page, which means that any rewrites should not be impacting your rankings. While Google says that this is a good thing for websites, Lilly Ray & other SEOs have seen some really strange title rewrites in the wild.

Danny Sullivan proposed one clever solution that Google could do to add an “and I really mean it” tag, but no one is really sure if that idea will grow wings.

Meanwhile, the short answer to this dilemma is that there are no guarantees while Google maintains control, but you should keep testing and iterating. If you can create a better title than Google’s AI, then your optimization stands a good chance of winning out.

Another potentially-impactful test in these scenarios could be shortening your title tag. One trend that Brodie Clark noticed among Google’s page title rewrites is a lot of shortened page title versions.

Title Tag Mythbusting: Common Misconceptions

There are few misconceptions lingering on the web about title tags worth clearing up because the advice that you might find online is either wrong, or ineffective, or both.

Myth #1: Isn’t the title tag the same as the H1?

No, definitely not. Some web pages will display identical text in both of these tags, but the two are not the same and Google does not treat them the same. Titles are much more influential to SEO, although H1s are indeed used to help Google understand web pages.

Myth #2: You should always keep your titles under 55-60 characters.

Although most page titles will get truncated on Google search results right around this # of characters, there is no rule that you must keep your titles within this limit. In-fact, Google will use the entire length of the title to evaluate the page for rankings, so the truncated portions are still factored as ranking signals.

Myth #2: Google will ignore stop words in your page titles.

It is widely-believed that Google ignores stop words altogether in page titles. However, as Bill Slawski has pointed out, it’s possible that Google still uses stop words because some are meaningful, but stop words which are not meaningful will get ignored.

Myth #3: CTR isn’t a ranking factor.

This myth currently seems to be a matter of speculative opinion between SEOs with more evidence appearing to support the idea than there is evidence to debunk it, which means that the myth here is likely true. But, even if it is correct, that doesn’t mean that CTR won’t help you rank higher. It’s possible that CTR may still impact some queries and not others, or that it’s impact is extremely minimal, or that CTR could lead to other indirect behavioral signals that could impact rankings. We recommend optimizing for CTR with the goal of growing traffic and engagement, not to influence rankings.

24 More Title Testing Ideas

1. Create a stronger alignment between the page title and search intent.
2. Create a stronger alignment between the page title and the content on the page.
3. Touch on a pain-point, then offer a solution.

Credit Card Late Fees and How to Avoid Them | YourBrand.com

4. Test a shorter title.

Credit Card Late Fees | YourBrand.com

5. Analyse your top KW modifiers and add one or two of them to your title for more long-tail clicks.

Recruiting Software: Solutions for B2B Recruiting Teams | YourBrand.com

6. Add adjectives to make your titles more descriptive.

5 Powerful Email Marketing Tools | YourBrand.com

7. Test varying date formats: Year, month + year, “this decade.”

Sales Leadership Strategies You Can Implement Right Now | YourBrand.com

8. Make it rhyme.

Content Marketing Plays

9. Use alliteration/repetition.

21 Powerful Content Marketing Solutions for 2021

10. Invoke a popular term, or slang term.

This Content Campaign is a Brilliant Lowkey Flex | YourBrand.com

11. Add product specs like color, size, model, price, or features.
12. Add a location, or change location formats like city & state, state only, city only, etc.
13. Test first-person versus third-person.
14. Borrow ideas from the competition’s page titles.
15. Test a less-creative, but more direct title. Sometimes dry candor works best.
16. Test question words (who, what, when, where, why, and how).
17. Inject humor into the title (only if it’s appropriate).
18. Add a time anchor. 5 Bodybuilding Techniques You Can Do This Week/Year/Month/Right Now
19. Use value-words like free, cheap, easy, etc.
20. Include the word, “lessons.” Everyone wants to skip hard mistakes by learning from someone else’s experience.
21. Name-drop. The Rock’s Top Tips for Shredded Abs.
22. Add,”[VIDEO]” if the content contains a video. How to Audit Your Google Tag Manager Account [VIDEO]
23. Switch it around. Move the second portion of your title to the front and vica versa. The 10 Best Financial Planning Software Solutions versus Financial Planning Software: The 10 Best Solutions
24. Add a word count.
25. The word, “proven.” 45 Proven Ways to Increase Your Conversion Rates
26. Add some controversy.
27. Give a nod to science and data. 8 Scientific Ways to Improve Your Content Strategy. Or, 8 Ways to Improve Your Content Strategy, According to the Data.