Today’s Ask An SEO question comes from Michal in Bratislava, who asks:

“I have a client who has a website with filters based on a map locations. When the visitor makes a move on the map, a new URL with filters is created. They are not in the sitemap. However, there are over 700,000 URLs in the Search Console (not indexed) and eating crawl budget.

What would be the best way to get rid of these URLs? My idea is keep the base location ‘index, follow’ and newly created URLs of surrounded area with filters switch to ‘noindex, no follow’. Also mark surrounded areas with canonicals to the base location + disavow the unwanted links.”

Great question, Michal, and good news! The answer is an easy one to implement.

First, let’s look at what you’re trying and apply it to other situations like ecommerce and publishers. This way, more people can benefit. Then, go into your strategies above and end with the solution.

What Crawl Budget Is And How Parameters Are Created That Waste It

If you’re not sure what Michal is referring to with crawl budget, this is a term some SEO pros use to explain that Google and other search engines will only crawl so many pages on your website before it stops.

If your crawl budget is used on low-value, thin, or non-indexable pages, your good pages and new pages may not be found in a crawl.

If they’re not found, they may not get indexed or refreshed. If they’re not indexed, they cannot bring you SEO traffic.

This is why optimizing a crawl budget for efficiency is important.

Michal shared an example of how “thin” URLs from an SEO point of view are created as customers use filters.

The experience for the user is value-adding, but from an SEO standpoint, a location-based page would be better. This applies to ecommerce and publishers, too.

Ecommerce stores will have searches for colors like red or green and products like t-shirts and potato chips.

These create URLs with parameters just like a filter search for locations. They could also be created by using filters for size, gender, color, price, variation, compatibility, etc. in the shopping process.

The filtered results help the end user but compete directly with the collection page, and the collection would be the “non-thin” version.

Publishers have the same. Someone might be on SEJ looking for SEO or PPC in the search box and get a filtered result. The filtered result will have articles, but the category of the publication is likely the best result for a search engine.

These filtered results can be indexed because they get shared on social media or someone adds them as a comment on a blog or forum, creating a crawlable backlink. It might also be an employee in customer service responded to a question on the company blog or any other number of ways.

The goal now is to make sure search engines don’t spend time crawling the “thin” versions so you can get the most from your crawl budget.

The Difference Between Indexing And Crawling

There’s one more thing to learn before we go into the proposed ideas and solutions – the difference between indexing and crawling.

• Crawling is the discovery of new pages within a website.
• Indexing is adding the pages that are worthy of showing to a person using the search engine to the database of pages.

Pages can get crawled but not indexed. Indexed pages have likely been crawled and will likely get crawled again to look for updates and server responses.

But not all indexed pages will bring in traffic or hit the first page because they may not be the best possible answer for queries being searched.

Now, let’s go into making efficient use of crawl budgets for these types of solutions.

Using Meta Robots Or X Robots

The first solution Michal pointed out was an “index,follow” directive. This tells a search engine to index the page and follow the links on it. This is a good idea, but only if the filtered result is the ideal experience.

From what I can see, this would not be the case, so I would recommend making it “noindex,follow.”

Noindex would say, “This is not an official page, but hey, keep crawling my site, you’ll find good pages in here.”

And if you have your main menu and navigational internal links done correctly, the spider will hopefully keep crawling them.

Canonicals To Solve Wasted Crawl Budget

Canonical links are used to help search engines know what the official page to index is.

If a product exists in three categories on three separate URLs, only one should be “the official” version, so the two duplicates should have a canonical pointing to the official version. The official one should have a canonical link that points to itself. This applies to the filtered locations.

If the location search would result in multiple city or neighborhood pages, the result would likely be a duplicate of the official one you have in your sitemap.

Have the filtered results point a canonical back to the main page of filtering instead of being self-referencing if the content on the page stays the same as the original category.

If the content pulls in your localized page with the same locations, point the canonical to that page instead.

In most cases, the filtered version inherits the page you searched or filtered from, so that is where the canonical should point to.

If you do both noindex and have a self-referencing canonical, which is overkill, it becomes a conflicting signal.

The same applies to when someone searches for a product by name on your website. The search result may compete with the actual product or service page.

With this solution, you’re telling the spider not to index this page because it isn’t worth indexing, but it is also the official version. It doesn’t make sense to do this.

Instead, use a canonical link, as I mentioned above, or noindex the result and point the canonical to the official version.

Disavow To Increase Crawl Efficiency

Disavowing doesn’t have anything to do with crawl efficiency unless the search engine spiders are finding your “thin” pages through spammy backlinks.

The disavow tool from Google is a way to say, “Hey, these backlinks are spammy, and we don’t want them to hurt us. Please don’t count them towards our site’s authority.”

In most cases, it doesn’t matter, as Google is good at detecting spammy links and ignoring them.

You do not want to add your own site and your own URLs to the disavow tool. You’re telling Google your own site is spammy and not worth anything.

Plus, submitting backlinks to disavow won’t prevent a spider from seeing what you want and do not want to be crawled, as it is only for saying a link from another site is spammy.

Disavowing won’t help with crawl efficiency or saving crawl budget.

How To Make Crawl Budgets More Efficient

The answer is robots.txt. This is how you tell specific search engines and spiders what to crawl.

You can include the folders you want them to crawl by marketing them as “allow,” and you can say “disallow” on filtered results by disallowing the “?” or “&” symbol or whichever you use.

If some of those parameters should be crawled, add the main word like “?filter=location” or a specific parameter.

Robots.txt is how you define crawl paths and work on crawl efficiency. Once you’ve optimized that, look at your internal links. A link from one page on your site to another.

These help spiders find your most important pages while learning what each is about.

Internal links include:

• Breadcrumbs.
• Menu navigation.
• Links within content to other pages.
• Sub-category menus.
• Footer links.

You can also use a sitemap if you have a large site, and the spiders are not finding the pages you want with priority.

I hope this helps answer your question. It is one I get a lot – you’re not the only one stuck in that situation.

More resources: 

Featured Image: Paulo Bobita/Search Engine Journal

Mastering effective ad copy is crucial for achieving success with Google Ads.

Yet, the PPC landscape can make it challenging to discern which optimization techniques truly yield results.

Although various perspectives exist on optimizing ads, few are substantiated by comprehensive data. A recent study from Optmyzr attempted to address this.

The goal isn’t to promote or dissuade any specific method but to provide a clearer understanding of how different creative decisions impact your campaigns.

Use the data to help you identify higher profit probability opportunities.

Methodology And Data Scope

The Optmyzr study analyzed data from over 22,000 Google Ads accounts that have been active for at least 90 days with a minimum monthly spend of $1,500.

Across more than a million ads, we assessed Responsive Search Ads (RSAs), Expanded Text Ads (ETAs), and Demand Gen campaigns. Due to API limitations, we could not retrieve asset-level data for Performance Max campaigns.

Additionally, all monetary figures were converted to USD to standardize comparisons.

Key Questions Explored

To provide actionable insights, we focused on addressing the following questions:

• Is there a correlation between Ad Strength and performance?
• How do pinning assets impact ad performance?
• Do ads written in title case or sentence case perform better?
• How does creative length affect ad performance?
• Can ETA strategies effectively translate to RSAs and Demand Gen ads?

As we evaluated the results, it’s important to note that our data set represents advanced marketers.

This means there may be selection bias, and these insights might differ in a broader advertiser pool with varying levels of experience.

The Relationship Between Ad Strength And Performance

Google explicitly states that Ad Strength is a tool designed to guide ad optimization rather than act as a ranking factor.

Despite this, marketers often hold mixed opinions about its usefulness, as its role in ad performance appears inconsistent.

Our data corroborates this skepticism. Ads labeled with an “average” Ad Strength score outperformed those with “good” or “excellent” scores in key metrics like CPA, conversion rate, and ROAS.

This disparity is particularly evident in RSAs, where the ROAS tends to decrease sharply when moving from “average” to “good,” with only a marginal increase when advancing to “excellent.”

Screenshot from author, September 2024

Interestingly, Demand Gen ads also showed a stronger performance with an “average” Ad Strength, except for ROAS.

The metrics for conversion rates in Demand Gen and RSAs were notably similar, which is surprising since Demand Gen ads are typically designed for awareness, while RSAs focus on driving transactions.

Key Takeaways:

• Ad Strength doesn’t reliably correlate with performance, so it shouldn’t be a primary metric for assessing your ads.
• Most ads with “poor” or “average” Ad Strength labels perform well by standard advertising KPIs.
• “Good” or “excellent” Ad Strength labels do not guarantee better performance.

How Does Pinning Affect Ad Performance?

Pinning refers to locking specific assets like headlines or descriptions in fixed positions within the ad. This technique became common with RSAs, but there’s ongoing debate about its efficacy.

Some advertisers advocate for pinning all assets to replicate the control offered by ETAs, while others prefer to let Google optimize placements automatically.

Image from author, September 2024

Our data suggests that pinning some, but not all, assets offers the most balanced results in terms of CPA, ROAS, and CPC. However, ads where all assets are pinned achieve the highest relevance in terms of CTR.

Still, this marginally higher CTR doesn’t necessarily translate into better conversion metrics. Ads with unpinned or partially pinned assets generally perform better in terms of conversion rates and cost-based metrics.

Key Takeaways:

• Selective pinning is optimal, offering a good balance between creative control and automation.
• Fully pinned ads may increase CTR but tend to underperform in metrics like CPA and ROAS.
• Advertisers should embrace RSAs, as they consistently outperform ETAs – even with fully pinned assets.

Title Case Vs. Sentence Case: Which Performs Better?

The choice between title case (“This Is a Title Case Sentence”) and sentence case (“This is a sentence case sentence”) is often a point of contention among advertisers.

Our analysis revealed a clear trend: Ads using sentence case generally outperformed those in title case, particularly in RSAs and Demand Gen campaigns.

Image from author, September 2024

(RSA Data)

Image from author, September 2024

(ETA Data)

Image from author, September 2024

(Demand Gen)

ROAS, in particular, showed a marked preference for sentence case across these ad types, suggesting that a more natural, conversational tone may resonate better with users.

Interestingly, many advertisers still use a mix of title and sentence case within the same account, which counters the traditional approach of maintaining consistency throughout the ad copy.

Key Takeaways:

• Sentence case outperforms title case in RSAs and Demand Gen ads on most KPIs.
• Including sentence case ads in your testing can improve performance, as it aligns more closely with organic results, which users perceive as higher quality.
• Although ETAs perform slightly better with title case, sentence case is increasingly the preferred choice in modern ad formats.

The Impact Of Ad Length On Performance

Ad copy, particularly for Google Ads, requires brevity without sacrificing impact.

We analyzed the effects of character count on ad performance, grouping ads by the length of headlines and descriptions.

Image from author, September 2024

Image from author, September 2024

(RSA Data)

Image from author, September 2024

Image from author, September 2024

(ETA Data)

Image from author, September 2024

Image from author, September 2024

(Demand Gen Data)

Interestingly, shorter headlines tend to outperform longer ones in CTR and conversion rates, while descriptions benefit from moderate length.

Ads that tried to maximize character counts by using dynamic keyword insertion (DKI) or customizers often saw no significant performance improvement.

Moreover, applying ETA strategies to RSAs proved largely ineffective.

In almost all cases, advertisers who carried over ETA tactics to RSAs saw a decline in performance, likely because of how Google dynamically assembles ad components for display.

Key Takeaways:

• Shorter headlines lead to better performance, especially in RSAs.
• Focus on concise, impactful messaging instead of trying to fill every available character.
• ETA tactics do not translate well to RSAs, and attempting to replicate them can hurt performance.

Final Thoughts On Ad Optimizations

In summary, several key insights emerge from this analysis.

First, Ad Strength should not be your primary focus when assessing performance. Instead, concentrate on creating relevant, engaging ad copy tailored to your target audience.

Additionally, pinning assets should be a strategic, creative decision rather than a hard rule, and advertisers should incorporate sentence case into their testing for RSAs and Demand Gen ads.

Finally, focus on quality over quantity in ad copy length, as longer ads do not always equate to better results.

By refining these elements of your ads, you can drive better ROI and adapt to the evolving landscape of Google Ads.

Read the full Ad Strength & Creative Study from Optmyzr.

More resources: 

Featured Image: Sammby/Shutterstock

Microsoft has announced an expansion of Bing’s generative search capabilities.

The update focuses on handling complex, informational queries.

Bing provides examples such as “how to effectively run a one-on-one” and “how can I remove background noise from my podcast recordings.”

Searchers in the United States can access the new features by typing “Bing generative search” into the search bar. This will present a carousel of sample queries.

A “Deep search” button on the results page activates the generative search function for other searches.

Screenshot from: blogs.bing.com, October 2024.

Beta Release and Potential Challenges

It’s important to note that this feature is in beta.

Bing acknowledges that you may experience longer loading times as the system works to ensure accuracy and relevance.

The announcement reads:

“While we’re excited to give you this opportunity to explore generative search firsthand, this experience is still being rolled out in beta. You may notice a bit of loading time as we work to ensure generative search results are shown when we’re confident in their accuracy and relevancy, and when it makes sense for the given query. You will generally see generative search results for informational and complex queries, and it will be indicated under the search box with the sentence “Results enhanced with Bing generative search” …”

This is the waiting screen you get after clicking on “Deep search.”

Screenshot from: blogs.bing.com, October 2024.

In practice, I found the wait was long and sometimes the searches would fail before completing.

The ideal way to utilize this search experience is to click on the suggestions provided after entering “Bing generative search” into the search bar.

Potential Impact

Bing’s generative search results include citations and links to original sources.

Screenshot from: blogs.bing.com, October 2024.

This approach is intended to drive traffic to publishers, but it remains to be seen how effective this will be in practice.

Bing encourages users to provide feedback on the new feature using thumbs up/down icons or the dedicated feedback button.

See also: Google AIO Is Ranking More Niche Specific Sites

Looking Ahead

This development comes as search engines increasingly use AI to enhance their capabilities.

As Bing rolls out this expanded generative search feature, remember the technology is still in beta, so performance and accuracy may vary.

Featured Image: JarTee/Shutterstock