Why it Depends

In SEO you so often hear the words “it depends” ….. but why?

I feel compelled to write this for so many reasons but here are my main two;

1. For SEO practitioners. It’s an unacceptable professional bar to leave it at “it depends” we must explain why, give the scenarios, considerations and context and provide relevant for instances.

2. For SEO clients. You deserve better, hold your SEO company to account. Ask why does it depend – what does it depend on, how does that pertain to my situation?

So, my intent with the rest of this article is to give you some common SEO questions or considerations, and then explain why the answer might be “it depends” and then explain some of the considerations as to why it may depend.

1. Direct vs Indirect Ranking Factors
2. Influence Function, Decay & Aging
3. Causal vs Symptomatic
4. Dynamic Environment – the Web Doesn’t Stand Still
5. How long does it take to rank on page 1?
6. Why don’t we rank for keyword X?
7. How long does an article need to be?
8. Why does a title need to be 80 characters or less. General Questions about title length
9. Should we delete poorly performing content
10. Should our blog be on a sub domain or sub directory
11. How Many Backlinks do I need?
12. Should I Disavow Certain Backlinks?
13. How Fast Does my Site Need to be?
14. Can I use AI in my Content?
15. How Often Should I Publish New Content?
16. Do I Need to Fix Every Error in my Site Audit Tool?
17. Should I Change my URLs to Have Exact Keywords in them?
18. Does Social Media Activity Boost Rankings?
19. Do Meta Descriptions Make a Difference?
20. References & Resources

Direct vs Indirect Ranking Factors

One of the biggest misconceptions and misunderstandings in SEO is the difference between a direct and indirect ranking factor.

Generally, you have on-page and to an extent, technical changes and modifications on your site that you usually have full control over. These are usually direct factors that have a relatively near term effect. So why can’t we just give an explicit time line? Make X change and see Y result in Z time frame? Well ….. “It depends”, the reasons why are things like crawl frequency, how long it takes for the page to be indexed, whether bulk link graph recalculations have happened or if there has or will be an algorithm update that occurs in proximity to the change.

Indirect ranking factors are things that have an influence on your sites rankings, such as getting an external link, a domain change in your backlink profile, people listing and mentioning you (which is a knock-on activity of good marketing). There are also time constants and aging factors for certain parameters. So when you get a good backlink, you don’t immediately see the effect of it.

Influence Function, Decay & Aging

One of the bigger areas that is generally less understood, but is very commonly understood in certain areas of engineering, is the influence function of a certain event.

Just because something happens, i.e an event like getting a backlink, you do not see an immediate 100% benefit of this backlink. 

There is a time for it to become indexed and recognized (this is usually the start of a time lag), then there is an aging effect which is believe to be applied with an exponential decay function.

Causal vs Symptomatic

One very interesting thing in SEO, similar to the direct vs indirect conundrum, is whether an SEO change is causal of an output or whether a measured quantity is just symptomatic.

People like to sound smart when ever anyone mentions correlation and say “correlation doesn’t equal causation”, which is true, and there are definitely cases where this is very valid in SEO (which we will explore more further on), however equally, there are lots of things that correlate that are causal.

It is important to understand what correlation really means and in terms of SEO what they really mean. There are not many very strong correlations in SEO. We are looking more for influence functions and things that move the needle rather than very strong 0.85+ direct correlations.

Dynamic Environment – the Web Does not Stand Still

One of the other big reasons that things may depend is, the web is not static. Your competition is not sitting around waiting for you to produce content, fix technical issues or acquire backlinks. They are doing the same and trying to do it better.

Things change with websites, things break, domains change, and mergers and acquisitions happen not to mention that there are multiple subtle algorithm changes regularly occurring along side the major core algorithm updates.

All of this is to say, that it is hard to make a fully quantifiable prediction of what effect will come from a change.

This is why the industry aggressively tests, experiments, measures, records and scrutinizes.

This is also why generally accepted best practices exist.

This is further a reason why AI will struggle to do SEO very well, as it is trained on a majority data set that is either out of date or just plain wrong.

OK, so with some of the fundamentals to why “it depends” out of the way, lets dive in to some actual examples and explore why “it depends”.

Q. How long does it take to rank on page 1?

This is a fair question which to a client may seem like a very reasonable one, which it is, however, it’s not that straightforward.

Why does it depend?

There are multiple factors that can affect the answer to this question.

The web is not static and nor is your site. Another question or counter to this one is how quickly can you execute?

Some organizations are small and nimble whereas others are not. If it takes a month to change a title tag the answer is probably a lot longer than you would like. 

What does it depend on?

It depends on speed of implementation, removing other technical hurdles and roadblocks.

It depends on what your competitors are doing – they are not standing still.

It depends on which keyword, keywords, phrases or topics you are targeting.

Important factors to consider

• Keyword competition and difficulty
• Backlink profile to competitors for the same terms
• Age and strength of your domain
• Quality of your content
• Technical health of your site

Examples and scenarios 

Example A

A medium age, medium domain srength SMB with an established wen presence and good credentials.

Targeting a new industry segment where the competition is not really talking about it and there are not many results.

If this gets a high score on our proprietary KEI (Keyword Effectiveness Index) and the competitive assessment process shows the strength (or lack thereof) of the competition is within a certain range the answer would likely be, expected to be within 1 month. Produce the content, get the page indexed, build some relevant internal links to it and get some coverage on the topic.

Example B

A brand new site with no established presence for a start up with no history and no credentials (yet) targets ac competitive market segment dominated by long standing large players.

The KEI score is low and the competition gap is huge.

STOLBER Digital Proprietary Keyword Ranking System and KEI metric.

The answer here may be, practically never and the advice would be don’t target it!

This is where the power of detailed technical analytical keyword research comes in.

We determine the right keywords to target where you stand a reasonable chance of ranking.

We bucket things into immediate, short term, mid term, long term and main content keyword categories.

Example C

An enterprise company with an established presence in the marketing and good history and credentials targets an existing space where there are many mid market competitors.

The KEI value is in the mid range and the competition is fair to medium strength.

This would be in the mid-range 6 month time frame, dependant on exact execution of identified criteria.

So – same question, 3 different scenarios, 3 completely different answers and outcomes.

This is why an SEO may say “it depends” but they should be looking at the specific scenario and give appropriate guidance in context of the specific situation.

Q. Why don’t we rank for keyword X?

Why does it depend?

Surprisingly this one is often more simple than many think.

We get asked this question by prospective new clients regularly and when we look, the answer is often “well, you don’t really actually mention it”.

What does it depend on?

This depends on the target keyword being on the page, in the title, in the headings and appearing relatively frequently and early on. It should clearly be what the page is about.

It also depends on the internal linking anchor text cloud i.e do you have links on your site from relevant content pointing to it with relevant terms?

Further, is this within the main content (MC) and theme of your site. Your content needs to live under one umbrella topic, if you fall outside of that, you may struggle to rank if your site tries to eb about too many disparate topics. Again, this is where thorough keyword research comes in.

Important factors to consider

• Competitiveness of the term
• Backlink profile (internal & external)
• Backlink anchor text cloud (internal & external)
• Structure of content and use of term within key elements
• Quality of content
• Technical health of page and site 

Examples and scenarios 

Example A

A site with a major hidden fundamental technical issue.

The site looks good and loads well, it gets a fantastic 99 CWV (Core Web Vitals) score but there is a catch. The tricks used to get that lightening fast CWV score actually prevent content loading until there is a specific user interaciton. For most web users thats common, but for Googlebot (Google’s search engine crawler) it does not, and the rendered page to Google is blank.

This is actually a lot more common than most people realize.

In this scenario, the search engine never sees the content that is on the page, so considered it not relevant.

Typical Core Web Vitals (CWV) Score

Example B

A B2B SaaS company has invested heavily in messaging, the know what they are targeting and their title says it, the keyword is literally in the title – they did their research, they know what they want to target. However thats why the good news stops. The headings on the page and the introductory sections talk about the messaging and how the solutions helps – which is great, but it does not tell a search engine that the page is relevant in any way to what the title claimed the page is about.

This is a common challenge, and it is a balance that the SEO team should be involved with, in getting messaging and keyword relevance into headings, titles and content.

This is the “do you really ever mention it problem” which is actually quite common.

Example C

A tech start-up is using the latest web technology platform to be super nimble and integrate new dynamic functionality and connect to all tools, custom CRM and backend systems, however the content is almost entirely client-side rendered (CSR). This can mean that important elements don’t get loaded and therefore, don’t get seen by a search engine. I have seen many examples where the Google rendered content page title is: [PRODUCT Placeholder Title].

You get the point – your site needs to tell the search engine what its about – not wait for it to try and figure it out.

Example D

Accidental robots.txt blocking of key resources. Sometimes backend sections and URL slugs for libraries and includes are blocked in robots.txt to prevent them from being indexed. Thats generally OK, but if it means that an important library used in display and rendering of key content doesn’t load it can wreak havoc with your rankings.

The worst part about this is that you may never see it happen and may not see it as an issue. To a regular user the content loads fine because that user (and thus user agent) is not having the key resource being blocked.

This one is also fairly common and is only discovered when an SEO takes a proper look under the hood. In fact, I have seen many SEOs not even spot these types of issues.

Q. How long does an article need to be?

Why does it depend?

This is one of those very interesting causal vs symptomatic questions comes in.

It is true that the top 10 organic results generally contain longer-form content, however the correlation is very noisy. While there is a correlation its not the driving factor. It is generally accepted that longer content is a symptom of a high quality, well researched piece of content. The length of the content is symptomatic of the quality of the content, not the cause of the high ranking.

This is quite misunderstood and is why you often see content length needs to eb 2000 words minimum.

What does it depend on?

This depends on the target topic, benchmarking of competitive content and what needs to be covered.

There are two interesting topics to consider SERP Intent vs Search Intent. They sound like they are the same thing and they are similar and related but they are not the same and this plays into content length.

Search Intent is what the search is looking for.

SERP Intent is what the search engine thinks the user is looking for.

Hopefully they are well aligned.

But consider this, you want to rank your e-commerce product for a keyword, but all of the content in the SERP (Search Engine Results Page) is “How to Blog Posts”.

You can write chapter and verse, but if the majority of the rankings are short form video explainer content your very long article will probably never compete.

Important factors to consider

• SERP Intent
• Competitive Benchmark
• Term-specific correlation
• Quality of competitive content
• Your authority to speak on the topic

Examples and scenarios 

Example A

A deep technical exploration of a category with sub sections on each specific sub category. This is an example of a piece of content that needs deep and comprehensive coverage.

Its also something that lends its self very well to passage indexing rather than page indexing, the nature of which and applicability of each should be considered when writing content.

Such an article naturally requires the content length to be high.

Example B

A news item or quick update is typically going to be short. It covers the pertinent facts and appropriate context but that is all. The importance and timeliness of the article is more important than the length.

Barry Schwartz does a fantastic job of these quick news items such as this one New Google Ads Report On New Customers Acquired

Example C

A blog post talking about a specific product or service you offer. This one is fairly subjective, since benchmarking may identify a higher number than is actually needed. The guidance we give is just to write it as you feel appropriate and not specifically target a certain length.

If you can write something with 500 words but the content benchmarking says 800 words minimum, don’t expand and add fluff just to meet a number.

Q. Why does a title need to be 80 characters or less? / General questions about title length. 

Why does it depend?

There is a lot of general guidance and best practice, including many tools that will cite this number, however the reality is much more nuanced than this hard and fast number.

A title over 80 characters will be truncated. Titles longer than 80 characters will typically be rewritten 75%+ of the time, where as in the 55-60 character range “sweet spot” often cited, the rewrite percentage is much more like 50%.

It is important to note that the full title will be used to rank against the algorithm, regardless of whether it is shown, rewritten or truncated in the search results.

What does it depend on?

Loosely, this depends on the type of page. For your home page and key service nd category pages you do want to keep the titles fairly short. You may also want to associate your brand or site name with it. 

You may also have a very detailed sub topic or blog post on a very specific thing, in which case you may need a lengthy title to fully detail and convery what the page is about.

Important factors to consider

• 80 Characters or more typically get truncated
• 80 Characters or more will be rewritten roughly 75% of the time
• The 55-60 character “sweet spot” gets rewritten around 50% of the time
• The full title is used in ranking regardless of whether its rewritten

Examples and scenarios 

Example A

A Home Page typically wants to have a punchy and catchy title that covers your main area or focus – your core term and service area as well as your brand name if it fits. This is your core SERP control and should be fully topically representative of what your site is about and what you do.

This is SERP control and you don’t want it to be rewritten or truncate.

Example B

Similarly to the home page, a top level category or service page should have a title that fully conveys what it is with the core term or phrase present. This can have the brand name at the end too if it fits, but if not be prepared ot drop it. Having the brand name in the title does help to associate your brand with the topic and leverages brand recognition on any SERP.

Example C

A white paper or deep technical coverage of a very specific topic typically requires a longer title to fully convey the important itesm covered. By the nature, you are starting with a long tail term, so having this and an explanation of the specific coverage usually puts you over the 80 character limit. With these types of pages it is typically better to have teh fully descript long title that allows the algorithm to compare gainst the full title than it is to have a title that doesn’t get rewritten as much.

Often time with these types of pages, the title rewriting, which happens dynamically on the fly- relevant to the users query, is likely more beneficial than having a static title.

Q. Should we delete poorly performing content?

Why does it depend?

This one is a little less of an it depends and more often a yes, however many SEOs may answer this with an it depends, and thats OK, as long as they do the appropriate analysis, consider the context, and give you the appropriate advice specific to your site.

What does it depend on?

This one depends on how much content you have, how much content your competition has and the performance of your content. If 95% of your pages have less than 10 visits a year, no ling and poor UX metrics those are very likely all candidates for deletion.

A proper content audit will take into account all of the appropriate factors and give a bracketed score to determine which content ot keep as is, which to improve, what to consolidate, what to redirect and what to delete.

A mass page cull in isolation and as a singular event is often ill advised, but a managed and staged content audit and pruning process can often be beneifical.

Remember that poor performing content can drag your whole site down and contribute negatively to your site wide quality score. This can prohibit you from being visible from certain SERP features and generally reduce your overall organic search performance.

Important factors to consider

• The traffic of rankings for the pages being considered
• Links to the pages of interest
• UX behavior of the pages
• Canibolization of considered pages

Examples and scenarios 

Example A

You have thousands of products on an e-commerce platform, many of the pages are dynamically generated, with slightly different variants (this is also where correct canonicalization and well-managed faceted navigation come in). These additional pages represent crawl bloat and don’t offer anything of unique value. They are still potentially useful for the purchase process and need to be accessible to users. They get very little traffic and have no links.

It does not make sense to purely delete these pages, but they should be excluded from search and have a correct canonical back to their parent or higher level main version.

Example B

You have an information-heavy website that has covered many topics over the years but is now pivoting into focusing on one particular technology or market vertical. Much of the content produced 2+ years ago gets virtually no traffic, doesn’t have any links and has poor UX performance.

These are good candidates for deletion, consolidation or redirection. A thorough content audit will appropriately consider and bracket these types of pages.

This type of site is prime for content pruning.

Example C

You have a medium-sized site and some of the pages don’t get much traffic, but there are some links, not much cannibalization and OK UX metrics.

This type of scenario typically does not warrant any content pruning.

Q. Should our blog be on a sub domain or sub directory?

Why does it depend?

This one is often answered with an it depends and I also see some SEOs recommend using a sub domain, however it is nearly always best to put your blog on a sub directory rather than a sub domain.

What does it depend on?

This one really doesn’t depend on much. Its generally best practice to put your blog on a sub directory rather tha sub domain.

Important factors to consider

• Intent and use of blog
• Ownership and governance of blog
• Backlink profile to main domain

Examples and scenarios 

Example A

A small business wants to start publishing blog posts, they have an existing established website. In this scenario, it makes sense to have the sub-directory blog on the main site.

Example B

An ecommerce site with established market presence and string backlink profile with good traffic. In this scenario, the sub directory is still the best choice over a sub domain. Sometimes people in this scenario question whether the blog has the potential to reduce the performance of the existing site. Aas long as good quality supportive content is produced the effect is net positive and is great value than the same content being produced on a sub domain.

Example C

A large news organization that has several local affiliates and sub groups all producing content under their own governance and organizational structure. In this scenario it does make sense to use sub domains rather than sub directories.

The cases where it makes sense to have a separate sub domain instead of hosting blog content on a sub directory on the main site are few and far between.

Sub domains are considered completely separate entities. It is essentially starting form square one.

There are some migration strategies with certain merger, divestiture or acquisition considerations with different future states where it does make sense to have sub domains too.

Q. How many backlinks do I need?

Why does it depend?

As most SEOs this question and they will say more, however many also respond with an “it depends”. You can’t build strategy on it depends though so we need more than that.

What does it depend on?

It depends on what your competition is doing. If you have more high quality backlinks than your competition, the answer might be none. That is rarely the case.

Important factors to consider

• How many referring domains your competitors have
• Link gap and metric analysis
• Keyword and Anchor text use in backlink profile
• Content to link ratio
• Content gap vs competition

Examples and scenarios 

Example A

A small business in a fairly competitive industry recognizes the needs for SEO and engages an SEO company. Their competitors have been established longer than theirs. They have similar content but they have 100 referring domains pointing to their site, whereas the average competitor has 900.

This does not mean they need 800 more referring domains, but they do have that size gap. After analyzing the types and categories of the referring domains, this can be down selected to a target that is more attainable and relevant, often times 50-100 referring domains based on analysis.

Example B

A medium-sized business that has been established for many years and is generally considered as a category leader has engaged in SEO to broaden its presence and rank higher for non branded keywords as it is relying mainly on its brand reputation and not getting in front of new customers unfamiliar with the brand.

They have more referring domains than any of their competition, but they are lacking in content.

They don’t actually need links. They might want to add specific topical focus on new backlink acquisition but the most likely item to move the needle here is to focus on content that targets long tial searches where they have gaps compared ot their competitors.

So the answer here might be 0 links needed.

Example C

An enterprise company is pushing AI heavily into their business, as are all their competitors. They have a good reputation and an industry recognized leading service, but the competition is pushing heavily into digital PR and their link and content velocity is higher – they will close the gap in 6 months if nothing is done.

Here, making sure that quality links are obtained at a rate equal to or higher than the competition will ensure their market dominance.

The important thing to note here is the link acquisition velicty as this is a leading indicator to future traffic.

They need lots of links! The exact number and gaps can be quantified and b uilt into a strategy.

Q. Should I disavow certain backlinks?

Why does it depend?

This one may be a little controversial. Some SEO tools actively recommend disavowing and give you a toxic link score. Google its self provides the function but warns about doing it and says you should only do it if you know what you are doing.

There are some theories that Google introduced this to use machine learning to identify spammy backlinking domains. That could well be true and likely worked.

What does it depend on?

Google has gotten very good at determining which links are genuine and which are not. They will almost certainly be assigning 0 value to any spammy backlinks, tahts even if the pages they are hosted on are indexed, which they often are not.

If you have really bad obvious bad link neighborhoods linking to your site and or have engaged with a bad SEO agency thate build spammy back links, or have been the victim of a negative SEO attack, it may be prudent to do some degree of disavowing, only after a thorough and detailed analysis.

Important factors to consider

• Are the links even indexed
• Did you pay a company to get the links
• Are similar sites with similar backlink profiles tanking in traffic?
• Have you been hit with a penalty or significant drop
• The disavow file is an absolute, you need to keep and maintain a record of it, submitting a new list will overwrite a previous list if the same domains are not in the new list, so it needs to be maintained

Semrush link profile graph

Semrush link profile characteristics

Examples and scenarios 

Example A

A small B2B website that has never done SEO before and has never paid for links runs an audit through a popular SEO tool and gets a list of toxic links to disavow.

In this scenario, you almost certainly should not disavow links

Example B

An AI tech start up paid a link building agency to spin a bunch of content and drive back links – it worked for a while until the last core algorithm update, now their traffic is tanked.

This is an example of where disavow probably should be used, unfortunately it may be too late and their domain may be tainted. It is possible to recover, but takes a lot of work.

Example C

An enterprise company that has used many marketing agencies and SEO companies over the years has had this recommendation to disavow lots of links including full domains from their SEO agency as a tool putput suggested that they should.

This likely is not a good case for disavow, but individual backlink analysis and audits will determine if this course of action is warranted.

Q. How fast does my site need to be?

Why does it depend?

Google introduced Core Web Vitals (CWV) as part of their Page Experience algorithm into the main ranking algorithm. Many SEOs including my self thought that this may become a significant factor, as they gave us around 6 months advanced notice of this upcoming change – something Google never does!

So the CWV metrics which measure site speed and performance are now something that directly affects the ranking algorithm.

What does it depend on?

A few years on from this introduction and there does not seem to be a significant influence of CWV on rankings. With all things being equal, better CWV will likely make you rank a big higher, but they do not seem to be largely influential.

Now, there are other good non SEO reasons, such as good user experience and general site performance and user retention to have good CWV metrics.

Some tools and techniques to fake out or attain really good CWV scores can actually kill your rankings.

NOTE: There has been a long standing perception that there was a 5 second render limit, however recent research shows this may not be the case

Important factors to consider

• How your site actually renders in Google Search Console matters a lot
• The time to render is important (although the classic 5 second rule is not as hard and fast as we thought)
• Don’t break your site experience to get good CWV
• A good CDN absolves many CWV sins
• CSR (Client Side Rendering) has additional considerations

Examples and scenarios 

Example A

A WordPress site uses a specific plugin designed to give 100 CVW scores. It does, it passess all the tests and gives a rocket ship CWV score. The site owner is very happy. Meanwhile, Google comes along and indexes the site with the updated plugin, which unbeknownst to the site owner 

Example B

A B2C e-commerce site that sells thousands of products at low cost has an LCP if 6 seconds. It’s mainly driven by a large hero image and the space is not defined causing a render blocking resource.

This is the classic issue. In this instance fixing the render blocking resource issue and providing a possibly small optimized webp graphic will make a significant difference that is likely to translate directly to increased revenue from higher conversion rate and lower drop off rates. 

Example C

A technical information site in a niche industry has a 70% desktop and 50% mobile score on page speed insights.

That’s not great and can definitely be improved, however it’s not likely to significantly affect traffic behavior or rankings.

Improvements here can still be made but they are not likely to be impactful. 

Q. Can I use AI in my content?

Why does it depend?

No! Well that’s just my preference.

Google has said that it is fine to use AI generated content as long as it is helpful and meets its guidelines.

What does it depend on?

The reality appears different. There are pitfalls to using AI generated content and we strongly advise against it. 

Important factors to consider

• Scaled content abuse penalties exist
• AI generated content leaves a finger print – it’s statistically determinable and changing a few words doesn’t hide that 
• It’s hard to provide information gain with AI
• It is essentially just regurgitating existing content that doesn’t provide unique value 

NOTE: This was written before the announcement of Claude’s Finger Print in AI Content, however it should be noted, there are other finger print and watermark giveaways in AI content.

Examples and scenarios 

Example A

A technical interview with an engineering B2B SME has the interview transcribed and uses AI to generate an article from it and extract quotes and key points.

This is probably fine and this is a good use of the technology and an efficient way to generate content with the source being an expert.

Example B

A company decides to produce 100 articles per day using automated topic selection and automated AI content generation. 

This is very likely to end badly and can cause your entire domain to be completely deindexed. 

Example C

A busy marketing manager needs to get a press release out about their new product. They use AI to take a few key points and generate the article based on their tone and brand guidelines.

This is possibly OK. We advise against it and prefer fully human written content. 

Q. How often should I publish new content?

Why does it depend?

This is similar to the content length question. Many strategies just revolve around blogging, however there is an analytical way to answer this question.

What does it depend on?

This question depends on the gap between your site and your competitors and the breadth of keywords and topics you are targeting.

Having a managed strategy rather than a spray and pray blogging approach will yield better results. 

Full disclosure and caveat here – this article is not part of any strategy (we do have a regular content strategy) – I am writing this purely because I want to and it seemed useful to me. Sometimes that’s OK. I have always said if you have something to say just say it. It’s always better to get it out there than have it sit silently in a document unpublished somewhere.

Google has said that it’s hard to consider a site an authority on a topic if they have produced less than 30 articles on that topic.

Important factors to consider

• There are practical limitations not most content production processes
• Setting a manageable regular cadence is important
• Auditing existing content performance can be more important than producing new content
• You may need to conduct content pruning 

Examples and scenarios 

Example A

A new site in the healthcare industry needs to establish credibility in an EEAT and YMYL dominated space. Producing regular expert content reviewed and approved by a senior expert listing the persons credentials and linking to their bio is very important.

Example B

A large ecommerce site selling cheap technical products, dominates the space and has a very strong backlink profile. 

Example C

A mid sized B2B engineering firm selling a custom niche solution not well known in the industry and with limited brand awareness. Producing content that satisfies users problems and talks about the real world issues the customers have and how the solution solves their issue is very valuable.

Q. Do I need to fix every error on my site audit tool?

Why does it depend?

The answer is usually no but prioritizing the important ones and understand how they pertain to your specific situation is very important.

What does it depend on?

As a site grows in size technical considerations become more significant. This is especially true of large database driven websites. 

Chart by Aleyda Solis and derived from Mark Williams-Cook chart.

The image above is an excellent depiction of how the importance of certain factors change with the siz eof your site.

Important factors to consider

• How broken is it (serious question to be asked)
• Size of site
• Age of site
• Strength of competition (are they all dumpster fires or well oiled machines?)

Examples and scenarios 

Example A

A small local business runs a Semrush audit and it comes back with a bunch of unminimized Javascript and CSS files and a bunch of missing meta descriptions. Fixing these, whilst best practice, likely won’t move the needle for the business and therefore should not be prioritized. This business likely needs content and links.

Example B

A medium sized B2B company has several broken links, some broken schema, some pages with important elements missing and a broken footer navigation.

This site should likely pay attention to the errors, work with a good SEO company and fix those issues. They are likely causing some issues and holding organic search performance back.

Example C

A large multi national, multi lingual B2C Ecommerce brand with a major presence in a very competitive market has a litany of minor SEO issues and has made some fundamental mistakes and misunderstandings based on prior baad advice. This one is a tech SEOs goldmine. There is likelty a ton of opportunity and they should focus heavily on tech SEO.

Q. Should I change my URLs to have exact keywords in them?

Why does it depend?

No! Almost always no! There are some SEO gurus out there who discuss URL best practices, and that’s fine when starting from scratch.

What does it depend on?

It is almost always better to leave an already indexed URL that search engines know about than it is to change it to be more “keyword friendly” and no provide a URL that search engines don’t know about. Even with a 301 redirect this is usually not the best idea.

So this one depends a lot on the age of the URL.

Important factors to consider

• URL traffic
• URL links
• URL age
• URL indexation state
• Intent of page 

Examples and scenarios 

Example A

A brand new site still in design or concept stage.

Yes absolutely, you are early enough where changing URLs that are not yet discovered could be improved upon.

A good navigation structure and URL pattern helps with analytics later down the line and follows some best practices.

Example B

An established small business that has roughly 30-50% of all of its leads coming from organic search.

NO! DO NOT change your URLs to make them fancy – there is no point and you will likely suffer a traffic loss, even if you aply correct 301 redirects.

Example C

A major B2B company is merging with 2 other companies, there is going to be a big migration project. The URLs will be redirected anyway.

In this instance it may make sense to do some URL consolidation and renaming. It is likely that the new site/s won’t be a 1:1 mirror and some new URLs and structures may be appropriate. In this case, working in best practices to URLs may be acceptable.

MOST people fall into the category of example B – don’t change URLs just because.

Q. Does social media activity boost my rankings?

Why does it depend?

This one is debated, as are many things in SEO, which is why testing and analysis is so important. There are many who believe so. It does stand to reason but as of yet I haven’t found solid direct evidence to support this.

This is one of those indirect and time lagging factors. This is definitely in the indirect category.

What does it depend on?

How social media content is shared, copied, linked to and referenced. Where the material is referenced.

Important factors to consider

• Passive link acquisition through sharing
• Where it gets shared
• If it gets picked up as news and linked to

Q. Do meta descriptions make a difference?

Why does it depend?

This is another one of those soft direct vs indirect factor discussions.

In short, meta descriptions are not used in the ranking algorithm and have zero direct effect on where you rank.

So, end of, case closed, full stop right? ………. Well not quite

What does it depend on?

The meta description lets you control some of the SERP real estate and the language used can influence a click. If your site gets clicked more and fulfils the user’s query, and they don’t go back to Google and keep searching, that is a positive signal for your site and over time will help you get a boost in rankings.

Now this is not a direct effect and it is not easily measurable, but it is still considered an SEO best practice to write compelling meta descriptions. They are also often mirrored into excerpts and in social media sharing like OG and Twitter cards, so how your content is seen when it is shared is also affected by this description. NOTE: Meta descriptions and OG & Twitter cards are all separate elements, but they often contain the same or very similar content.

Also note, while Google does not use meta descriptions in their ranking algorithm, some other search engines do so it is still important.

References & Resources

• August 7th, 2026 – New Google Ads Report On New Customers Acquired – Search Engine Roundtable – Barry Schwartz
• July 23rd, 2026 – Google doesn’t cut off rendering after 5 seconds, but I understand why you might think it does – Tame the bots – Dave Smart
• How Often Does Google Update Its Search Algorithms 
• November 12th, 2021 – Google: “Hard To Call A Site Authoritative After 30 Articles” – Search Engine Journal – Matt G. Southern 
• Google – Using AI Content
• Google Page Speed Insights
• Deciphering SEO Audits
• Claude AI Content Fingerprint
• Aleyda Solis
• Mark Williams-Cook 
• Semrush Site Audit 
• Semrush