SEO Masterclass

Timestamps

• 00:00 - Intro
• 0:58 - SEO baseline
• 1:15 - Search Engine Result Page
• 3:05 - How to track SEO performance
• 3:48 - How things work
• 4:24 - Proxies
• 5:03 - AI SEO
• 8:24 - How Search Engines Work
• 9:11 - Crawling
• 10:33 - Sitemap
• 10:55 - Orphaned Pages
• 13:31 - robots.txt
• 15:43 - Canonical pages
• 17:11 - Domains
• 19:11 - What is Google Search Console?
• 19:33 - What you put on your site matters
• 20:01 - Markup
• 21:30 - Content and Intent
• 23:32 - Performance and experience
• 24:34 - Building authority and trust
• 25:23 - Generative AI performance
• 26:15 - Multitenant Architecture
• 28:19 - Hands On
• 28:35 - Getting started with Data Impulse
• 31:56 - How to improve your SEO
• 32:06 - Intent
• 35:36 - Google Trends
• 36:54 - SERP Suggestions
• 37:09 - Google Adwords Keyword Planner
• 38:05 - Google Search Console
• 39:24 - SERP Preview for Your Site
• 39:51 - Analytics
• 42:09 - sitemap.xml, robots.txt, llms.txt
• 42:41 - Site Performance
• 43:21 - Generative AI Features
• 43:51 - Improving SERP Rankings

It does not matter how much time you spend writing articles or doing keyword research if you skip the fundamental SEO stuff.

For the last several months I've been building my own software to help people build their own websites, and I'm now at the point where I'm really focusing in on SEO, giving any of the sites I've created the best possible chance of ranking in Google or other search engines, including LLMs.

An LLM here is ChatGPT, Claude, Gemini, any of the AI tools you send requests to that can go search the internet on your behalf. Very similar idea to SEO, just specific to AI.

SEO Baseline

Before you make any improvements you need a baseline. How is the site currently performing?

You'll also see the term SERP, which stands for search engine result page. If you look up best tech YouTubers, there's a chance I show up on that page. The page that comes back from the search is the SERP.

People typically think Google search, and doing things right for Google gets you pretty far. But YouTube is the second biggest search engine right now, and a lot of people don't even consider it. If you have a business or a site and you put up some videos that are good for SEO, you can get better results than just creating web pages.

So think in multiple media types. A web page can have a video in it. It can have photos, and those photos can be found through image search, which is another way for people to find your site.

Most of my traffic is actually routed through YouTube. People find my videos, that helps them discover me, and all my links are in the description for them to click through.

How to track performance

Three things.

• Install some form of analytics on your site
• Verify your site in Google Search Console
• Use proxies to see the SERP from somewhere other than your own machine

That third one needs a bit of explaining. Normally you access the internet through your own device, so whatever you search from here, someone else could search the same thing from a different location or with different settings and get a completely different result.

Which means if you search your site up and see it ranking, that doesn't mean it's ranking for everybody everywhere. If you're only looking at your own results, you're pretty much driving blind.

A proxy sits in the middle and lets you check the search engine result page as a client in a different location. For this video I used DataImpulse, the sponsor, which gives proxies in over 195 countries and has 90 million+ residential IPs. Traffic never expires and no subscription is required. I'll go hands on with it later.

AI SEO

This is a big topic and a lot of people are using AI as a new trick to get people to invest in SEO services. Basically "hey, you thought SEO was important? Well it's not anymore, what's important is AI SEO."

You'll hear a couple acronyms thrown around.

• GEO, generative engine optimization
• AEO, agentic engine optimization

There are probably a few other variations. These are all basically variations of the same thing. You might also see LLM SEO, LLM being large language model, the models behind all the major AI tools.

There are some specific things to be aware of for LLM SEO, but at its core the same SEO principles apply. If we optimize the site correctly for SEO there's a really good chance we're already performing well with LLMs, and then there are a couple additional things we can add on top.

So don't think SEO is dead. I want you to embrace it farther, because done correctly you get good results from search engine results and LLM results.

Here's a quote from the Google developer docs:

Is SEO still relevant for generative AI search? In short, yes! The best practices for SEO continue to be relevant because our generative AI features on Google Search are rooted in our core Search ranking and quality systems.

Google's Guide to Optimizing for Generative AI Features on Google Search

Same doc also says you don't need to create new machine readable files, AI text files, markup, or Markdown to appear in Google Search including its generative AI capabilities, as Google Search itself doesn't use them.

That doesn't mean those extra things aren't helpful, because Google is not the only search engine out there. Google specifically says it won't hurt if you do them, just that Google search doesn't use them.

One simple example is adding an llms.txt, which can help AI parse and use your site. There are various others. The big thing is to start with strong SEO fundamentals first, then add in the AI capabilities.

How Search Engines Work

To understand SEO you should first understand a bit about how search engines work. They might seem simple, you type in some words and a couple results come up, but behind the scenes it's a very complex system design problem.

Three main pieces:

• Crawling
• Indexing
• Serving

Crawling is the bots. You might hear crawlers, bots, or spiders. In the world of Google SEO it's Googlebot doing the crawling, but most search engines work pretty similarly. The bot navigates to your pages and media by traveling through the links on your pages.

As it does that, all the information is indexed. You can think of indexing as just a massive database that the important information goes into after the crawler reaches it.

And serving is the SERP. You search something, that search goes to the index, grabs the important information out of the database, and gives back a search engine result page.

Every single one of those pieces has something you need to know about as the admin of your site.

In-Depth Guide to How Google Search Works

Sitemap

The sitemap is a file that shows the crawler what's available within your site. It's important because it shows how to navigate your site.

Orphaned pages

Say we have page A, but it's not linked to from any other page. That's an orphaned page. You can think of it almost as an unlisted web page, there's no direct way to find it unless you know that specific URL.

There are two pieces to this, and they're not always the same:

• the page is not linked to
• the page is not in the sitemap

We could put it in the sitemap but not link to it, and the crawler still finds it. Or we could link to it and not put it in the sitemap, and it can still be navigated to from that original page.

For a very important page, make sure it's both linked to from another page in your site and in the sitemap.

But some people choose to only put the very important stuff in the sitemap, so you might still link to certain things and not list them. That's fine if you basically want to prioritize certain pages within the sitemap.

An example of in-sitemap-but-not-linked might be a custom landing page. If I have software I'm selling to business owners and then I want a specialized page for software developers or content creators, I might put that page in the sitemap but not link to it from the home page. Google can still find it from the sitemap, but an organic user isn't going to land there clicking around the menus.

robots.txt

This is a permission layer, whether the crawler can navigate your site at all. Think of it as a gate they have to go through.

It blocks crawlers, but it doesn't necessarily prevent indexing, especially if your page is linked to from some other site. There's a chance it could still be indexed.

So if you don't want your page in the search engine result page at all, there's a noindex option you can look into that goes in the page's metadata. That's different from what robots.txt does.

Within the robots.txt we can disallow navigation, or even certain paths. Say you have a bunch of variations of the same page, or minor variations in the URL structure, you could disallow those paths to prevent potentially infinite possibilities for the crawler to walk through. Or if the crawler is hammering a page it doesn't need to be crawling, disallow it.

💡 Tip[!tip]
robots.txt is a public page, so don't use it to disallow a bunch of private URLs you don't want people to know about.

Some people go in and start typing keywords in their pages and fail to do the basic fundamental stuff like structuring their sitemap and their robots.txt.

Canonical pages

Duplicate data should be seen as bad. What we want to do is identify what page is the real thing, and that one is said to be canonical. It's canon.

The indexing process will go through similar pages and potentially group them together. And the search engine has information not only on your site but on everybody else's site too. So if another site has what you have, how does the search engine know which one is correct and which is the duplicate?

As your reputation within the search engine increases, if somebody duplicated your site copying your exact page, it would be able to identify that your page is the legit one.

This is a pretty common issue. Sites get duplicated whether it's intentional (copying someone for phishing or scamming) or just being indexed and displayed on some other site. If you've searched around forums you've probably seen different sites with the same data on them.

Domains

You could have coursecatalyst.com/pricing or www.coursecatalyst.com/pricing. You could also have variations with the protocol, HTTP versus HTTPS. And potentially variations with the URL ending in a slash versus not.

The indexing process will figure out which page is the correct one when we have very similar or duplicate pages. There's a good chance it does it right, but sometimes we want to explicitly say hey, this is the main page.

The one without the www is the apex URL, by the way.

We can specify this clearly with canonical tags. When you're linking you can say what the full correct URL is:

<link rel="canonical" href="https://coursecatalyst.com/pricing" />

You can also do a redirect within DNS from one domain to the other, and it's worth reading up on the different types of redirects. Having that redirect makes it very clear which domain is the canonical one and which one is something somebody accidentally typed or linked to from social media.

What is Google Search Console?

It's a way to verify ownership of your domain and get insights into how your page is showing up in Google search engine results, plus different things you can configure to tell Google to change those results. Definitely something to look into if you're managing your own site.

What You Put on the Page

We've talked about crawling and indexing, but what's actually in the page matters too. What you do on the page helps determine what the search engine does when it retrieves all that information, so you have the ability to influence how the page is indexed.

Markup

Some of these are just basic checks to make sure your page is structured properly. You can run various SEO audits on your page to check them.

• Titles, and specifically the title hierarchy: H1, nested H2, H3, all of those
• Proper HTML throughout the page, so you're not using or nesting elements incorrectly
• Meta description, which influences how search engines interpret what your page is about
• Proper anchor text on your links
• Alt text on images, which is good for SEO and also for screen readers
• Good page size, so the page loads fast and doesn't retrieve too much data

This list is massive. Basically anything with basic SEO, just making sure the page is structurally good and correct. All of this applies to every page and has nothing to do with what you're actually putting in the page.

Content and intent

The second big thing is the actual content, and the intent.

Content gets into keyword search and Google Trends, and then high-quality content. Basically creating stuff that's not commodity content, that's harder to reproduce. If you do something very basic and general, that's been done a million times. Infuse as much as you can into it to make it human, unique, interesting, and not easy to reproduce.

Intent is the general idea of what's the purpose of the page, but it's also kind of a formal structure. There are different page intent types:

• Informational
• Commercial
• Transactional
• Navigational

There's also freshness. You might benefit by putting the date or last updated date in the page.

This is where much of the strategy portion lives. Something I do with YouTube videos is make a bunch of videos that point people back to one very important video. That's a pillar approach, everything linking back to some high value content, and you put extra effort into keeping that one up to date and making sure it ranks. The others act as a way to boost it.

That's just one example strategy, there's like a million SEO strategies. Which comes back to being able to track, so set up analytics and you'll know what's working and what isn't.

Performance and experience

We'll cover this lightly. We want the page to be fast. There are different tools to audit that, or you can go to DevTools in Chrome or something similar in other browsers.

Time to first byte is one to care about. When you make a request, how long until you get the very first byte back from the server? That time is kind of wasted, we're not downloading anything yet, it's just getting to the server and back.

Then how long does the whole page take to download, and what's the actual file size of the full page?

And then user experience. Is the page responsive on mobile, are we following accessibility best practices, and so forth.

Building authority and trust

This one is huge and probably the hardest to fake. The biggest way to track the idea is backlinks, where someone else's site links to yours.

That's a huge win, but it's not necessarily going to help you if it's a low quality site. Quality over quantity. You don't want to pay some backlinking service that links to your site from a million low quality sites. You want to get linked to from reputable sites, and that reputation follows to your site. Some people call it link juice.

Generative AI performance

There are steps you can take to increase the chance of your page showing up in LLMs, and you can also track what's working.

In your analytics you'll have a source showing where people came from. If you're seeing results from ChatGPT or Perplexity or Claude, you can check what pages those people are going to, and do more of that.

Multitenant Architecture

Bonus section, only relevant if you're working within a multi-tenant platform.

The site I built lets people create their own site. So this is relevant if you're the app developer creating a multi-tenant platform, or if you're building a site powered by one. Shopify, Wix, Squarespace, Kajabi, all of those.

The main thing to be aware of is that certain things might be out of your control. You might need to check the app's settings to see if you can set up analytics, if you can check your sitemap to confirm it correctly created all the pages you need, if you can connect your site to Google Search Console.

You can still do a lot of it from one of these instances, but it's most likely a collaborative effort with the app your site is built from. They might give you built-in features, or a lot of customizations, and you can work with them directly.

If you're doing something like self-hosted WordPress where you actually install WordPress, that's not a multi-tenant system. That's your own site that you fully control.

Hands On

Now the hands-on portion, so you're not just driving blind, making changes and writing articles hoping something sticks.

You can get started for as little as a dollar per gigabyte for residential proxies. I chose a residential proxy, which gives all the capabilities we need here, and it shows up under products.

I did the proxy setup and configuration in Chrome, but for actually using the proxy I went with Firefox and the FoxyProxy Standard extension. From the extensions go to options, then proxies, then add.

Back in DataImpulse under products you get the login, password, and everything else you need. Start with the proxy host and paste that in, then fill out the rest.

The country of choice gets appended to the username, like __cr.ca for Canada. You can find these values in the proxy configuration in DataImpulse, appended to the username right there. Then hit save.

Now check your IP with any IP checking site such as ip-api, and you can see it's coming from Ontario with that proxy selected.

Doing a search for restaurants near me, we get the "allow access to your location" prompt. I blocked it, since allowing it kind of defeats the purpose of the proxy. And the results are showing Canadian websites. Same for Bing, all .ca domains.

I searched software engineering fundamentals through the Canadian proxy and scrolled through. Most results are .org or .com, there's a .ca for Amazon, and there's a YouTube video, which is one of mine that ranked there.

💡 Tip[!tip]
I wasn't on YouTube doing the search, that video got pulled into the Google search results. YouTube is super valuable for SEO.

Then I disabled the proxy and refreshed. That video still ranked, but so did my website. Switch back to Canada, refresh, and my site is totally gone.

So I'm ranking for my own search in the US and not through the proxy. I might be super excited because I'm ranking, but it might not be a universal experience. I can't fix that unless I know it's a problem.

Over in the dashboard I can see the usage. It looks like I spent 7 cents on all the tinkering in this video.

Understanding Intent

None of the SERP checking matters if we can't rank for anything, so let's get into improving.

Intent is a way to describe what type of content you'd expect for the search term you're using. Sometimes the intent behind a search term is not 100% clear.

Software engineering fundamentals could be a blog or instructional information, like the YouTube video that shows up. Or in the non-Canadian search we see a landing page for software engineering fundamentals, clearly designed to sell something. So I'd say that's a mixed intent search term.

These categories are informally defined, there's not one source of truth for what the options are, but here's the categorization to be familiar with.

• Informational, purely designed to teach or instruct
• Commercial, nurturing content or sales copy, like "best software engineering online courses for 2026" comparing options and listing their own courses, ultimately leading people to a sale
• Transactional, the actual sale page
• Navigational, a page like a login where somebody knows exactly what page they're trying to get to and isn't trying to learn, compare, or purchase

Why does it matter? When somebody searches something there's an implied intent behind it, and that's how the search engine knows what to put on that page.

So when you're trying to rank for certain keywords, first ask what the intent behind that search term would be. The easiest way to figure that out is to look at the results for it. Whatever intent is behind those pages is what's expected to compete for that term.

If you're using primarily informational intent for a search term that has higher intent for commercial or transactional, you're going to have a harder time competing.

Usually there's an overwhelming obvious answer for intent. But if there's not a very clear winner, it's probably a bit broad for you to be trying to compete for. If I wanted to take over the search term "software" that's very general, it's not clear if I'm trying to buy software, download software, or learn about it, so there's a ton of possible results that could fill that page.

Finding Search Terms

Google Trends

Probably the easiest option. Come in, type Agentic AI or whatever it might be, and see the interest over time. I'd zoom out a little, maybe the past 3 months, and get an idea of how things are trending.

Also compare it against something else, as that helps you understand it relative to something you already know, whether it's popular, growing, or shrinking.

From there you can also suggest search terms and use their little AI tool, and it gives you some options. Machine learning, agentic AI, large language models.

These are all pretty general topics though. If you're going for ranking, target something more specific. "What is machine learning used for? Five top use cases" might be an example article. That's going to give you a much higher chance of actually ranking and getting views.

SERP suggestions

Go to the search engine results page and look at the options there, plus the "people also asked" section. Scroll through, get some ideas, check out the AI overview, and at the bottom there's usually some more suggestions.

Google Ads Keyword Planner

For more advanced options. Discover new keywords, search a term, get results. You can put in some keywords or start with a website and it fills in a bunch of keyword ideas.

Go through and figure out what's doing well, find some open opportunities.

💡 Tip[!tip]
The competition value here is for ads. It doesn't necessarily translate to competition for organic searches. The volume is the genuinely helpful column.

Looking at the most valuable terms, software engineering technology has the highest bid, along with software engineering program. I'm guessing a bunch of schools and course programs are advertising, and the competition is pretty fierce there.

Tracking Performance

Google Search Console

Start here, verifying ownership of your site. You'll see the performance for your site in search engines.

Quick example, on Monday I got 8 clicks from 259 impressions, and you can see the actual search terms my site showed up for.

The big things I care about are the impressions and the click-through rate. Click-through rate is the ratio between the two, so for Code Breakthrough we have one impression and one click, a super high click-through rate.

CTR does not necessarily translate to better SEO or higher ranking, but it does translate to more total visits. Your site might be getting ranked in searches and you wouldn't even know until you come here, and then you might find people are not actually clicking. That's an indication you need to change some stuff.

It's nice to know that even though I migrated my entire site to my own platform, I'm still getting rankings for some of these terms.

SERP preview for your site

If you're wondering how your site actually shows up, use the site: operator with your URL and you'll see the different ways your pages are showing inside Google.

site:codebreakthrough.com

This lets you see if anything weird stands out. My first thought is the favicon could definitely be improved. I also don't like that I have the same name twice, so I should go fix that.

Most of the data here is parsed by Google, so it changes based on what's being searched.

Analytics

Google Search Console is one piece of the puzzle. You also want analytics.

Here's my app's analytics page, with visitors for the last 24 hours and the different paths that were visited.

A big thing here is the referrer, how people are coming to the site. There are 20 visits from google.com, and opening that shows what pages were actually hit.

But that's not what I'd primarily use analytics for. I'd figure out what OTHER ways people are finding me. In my case YouTube is a huge driver of traffic, so instead of spending a ton of time on the other options I might just go heavy on YouTube. How can I make that even better since it seems to be working?

Sometimes people focus on improving the worst of the list, trying to boost everything up, when maybe you just want to go heavy on the things that are already working.

You can also find location, environment, and times of day. Another big one is visit duration, we want people to actually stay on the site. And page views per visitor, so I have 833 views split across 366 unique visits from 314 unique individuals, a little less than three page views per visit.

There's a lot of other stuff you can do, like UTM tracking, which lets you be more specific about how somebody reached a page. If I wanted an affiliate program and wanted to track who sent people to my site, I'd use UTM tracking for that.

My platform is multi-tenant, so I can grab a single site, apply, and see the visits for just that site. Then compare against other sites under my platform, or go for the platform page itself.

Most of my traffic is end users, people watching course content, not potential customers of my platform who would be course creators.

The files

Most sites are going to have a sitemap.xml, showing the most important pages for the search engine to go through, plus a robots.txt. On the one I looked at, they added the API and LP directories to the disallow.

And then llms.txt is a little less common, but it'll probably become more common. Here's an example I found, just text describing the product and the different pages and how an AI could go through the material.

Site performance

pagespeed.web.dev is very helpful. You want these scores as high as possible for both performance and SEO, and it goes through and measures all the different values and areas that could be improved.

There's some stuff on AI in there too. For my site, agentic browsing flagged that the accessibility tree is not well-formed, so that's something I should go improve. And there are some not-applicable ones you might want to consider, like WebMCP or llms.txt.

Generative AI features

Within Google Search Console you can also get generative AI features. It's in beta right now, but it shows the impressions you've received through the AI summary within Google.

I was able to get mentioned 12 times in one day, which is pretty awesome. And this gets a lot better if your core SEO is good. I didn't do anything to optimize for LLMs and I'm still getting views.

Improving Your Rankings

It's a little scary because if you make a change it's not guaranteed to improve the results. It could make things worse.

So instead of thinking "how could I get this to rank higher for a search term", I think "what are the least destructive things I can do to this page to likely increase its ranking, without the risk of it being lowered or removed from that search term altogether."

Some of these directly affect your rankings, others just help with overall site reputation and branding. And these things change over time, so what's super important today might be less valuable tomorrow.

Always do these

Custom favicon, that's the little image that shows up with your website. If you're on a multi-tenant platform they probably give you a default one, and you'll want to update it.

Alt text and accessibility, which helps your site's reputation and can potentially avoid some legal issues too.

Outdated information. You might have conflicting info on the page accidentally saying you have a 14-day free trial in one spot and a 7-day free trial in another. Fix that, you want the page correct and up to date.

Meta descriptions, which help with click-through rate. That might not always show in the SERP because Google still might choose what to put there, but it can show up and it can help. Google does say click-through rate doesn't affect ranking, but it's one of those things that's still very important and has an indirect benefit.

Links to your own pages. When you post something new, go update pages and get internal links pointing to it. This is how search engines know certain pages are more important within your site.

Anchor text, meaning you anchor the text of where it's supposed to go, not the "click here". Say "get my 30-day challenge" and highlight the 30-day challenge, instead of "click here to get my 30-day challenge". Crawlers use that to know the intent of the click.

And redirect 404s to correct pages. If people are going to old pages, get those pointed at the latest stuff.

Riskier

Removing links is more dangerous than adding links, because when you remove a link you're reducing the authority of whatever page it points to. So you can have side effects, changing one page and hurting another.

Bigger content changes are the other one. A complete rewrite of my landing page is more dangerous, especially if I've already ranked for things like software engineering fundamentals. I know something in that page is working, so I'm not going to go in and redo everything.

If I am going to redo things I'll do it slowly, working mostly from the bottom up, since the things at the top should be considered more important than the things at the bottom. I'd first work through my FAQ and lesson structure, maybe then my about me, and then the more specific product stuff.

Not that you can't do them, just understand it's riskier. You don't want to change so much that Google no longer finds the page relevant for a search term.

Riskiest

Incorrectly modifying your robots.txt. If you accidentally say no search engines allowed, that's going to be bad.

A change of intent for a page. If I have a URL for software engineering fundamentals and change it from a landing page to an instructional page, that's a change of intent and can harm rankings.

Title changes, which is a big one. You set the page title in the metadata and that's really how you tell the search engine what this page is about. If you're ranking for what your title says, be very cautious about changing it, there's a good chance you get removed from that ranking.

URL or slug changes. Changing from software engineering fundamentals to beginner coding is a completely different URL structure. Not only will the previous page 404, but now there are different terms in the URL. This is basically like restarting. Redirecting the 404 to the new page can help, but it's not going to be as good because the previous search engine result pages I was showing up in will no longer feature that URL.

And domain migration. A big example is going from a subdomain to a full custom domain, where you're basically restarting all your SEO progress.

That said, if you're still on a subdomain you're probably not that ranked and still pretty early on. I'd still switch to a custom domain, because a custom domain is higher value and you have a higher chance of being successful with it. Even if you lose out on a few rankings you'd landed, you end up with a higher value domain that gives you more results over time. It's kind of like investing.

Then 301 the previous domain so it points to the new one. Do this as early as you can so you're not building a bunch of reputation on a subdomain, which is basically its own thing. If you have calebcurry.wordpress.com and switch that to calebcurry.com (first off, how did you get my domain name?) you're basically going to lose all that original progress.

On reversibility

If you do any of these and mess it up and rankings go worse, mostly it's reversible, unless you change the actual URL structure and can no longer visit that page. It's more going to be ups and downs.

So track it over time and make small improvements. What you can also do is add new content with different intent, different URL structures, or different keywords to amplify what you already have.

If I wanted to talk more about software engineering fundamentals, I could create content pieces like the top five software courses, nested under blog/software-engineering-fundamentals, talking about it from a different angle, more informational or more commercial, and then link to the landing page for the actual transaction. That's how you get different keywords and more rankings overall.

And give it time, because it takes a lot of time to build reputation. I transferred my domain to a new platform, removed some pages, added some pages, made big changes, and I'm still getting traffic from earlier SEO ranking without even trying, just because I've been building on the same domain for a number of years now.

Volume vs Percentage

Last tip, percentage improvements are often outweighed by volume improvements.

If I create YouTube content and send a bunch of traffic to my site, that makes my site overall better. To me that gives more results than having 10 views from the search engine, optimizing a bunch of things, and getting it to 20 views.

When you're really early you just want to send as much traffic to your site manually as possible, using social media, YouTube (which is kind of a social media), newsletter, all the things that send people to the URL and not through a search engine. Eventually you start getting backlinks, which improves your site's reputation. As you go you add pages and articles and compound your organic search views over time.

That's going to be way more valuable than over-optimizing nothing. Because if you 2x nothing, you still have zero.

So to get that initial burst I really recommend direct traffic. Then once you have actual traffic, doing those optimizations makes a much bigger change. If you're getting thousands of views and you do a 2x improvement, you now have twice the thousands of views.

Long term, focus on both. Don't worry so much about optimization that you ignore direct traffic through social media and backlinking, but don't ignore SEO either and only focus on sending people there directly.

This is a vast topic, so I encourage you to follow SEO content. Check out the SEO subreddits or different threads and sites that give up-to-date information so you can absorb it over time.

The full Google docs are here if you want to go deeper: Documentation to Improve SEO