SEO Basics That Are Most Often Overlooked When Building a Website

In most web projects, SEO is treated as the final step. The briefing often states: “Build the site first, then do SEO at some point.” The problem with this sequence is that much of what makes a website visible on Google at all is determined during the actual development process, not afterward.

What is overlooked here can only be made up for later with significantly more effort. Sometimes, it can’t even be fully rectified. We see this in nearly every other project that comes to us after it has already been built elsewhere: The fundamentals are missing, and retrofitting them costs many times more than it would have cost to incorporate them from the start.

Here are the fundamentals that are most often overlooked, and what should be taken into account from the very beginning instead.

What SEO Actually Needs to Achieve When Building a Website

SEO isn't just a list of tricks you apply at the very end. It's the combination of technical accuracy, a clear structure, and content that actually matches what people are searching for. All three of these elements are largely established during the design and development phases, not afterward.

If you treat SEO as the final step, you’re essentially not treating it at all. While a basic framework can be fixed, any fixes made to a live website are more time-consuming, riskier, and more expensive than making the same decision before you even click “Publish” in the CMS. That’s exactly why this list belongs in the briefing, not in the post-launch review.

The following seven points are not insider tips. They are fundamentals. Nevertheless, they are missing from most of the projects we take on because they were simply never explicitly requested. This is rarely due to a lack of knowledge on the agency’s part. It’s because SEO doesn’t have its own place in the project plan; instead, it gets lost somewhere between design approval and the launch date.

If you’re familiar with these points, you can specifically request them for your next project—whether you’re the client or the team building the site. That’s exactly what this list is for: not as a way to troubleshoot problems after the fact, but as part of the briefing that’s established from the very beginning.

1. The URL structure isn't well thought out

Many websites end up online with URLs like marschfahrt.de/page?id=384 or with wildly nested subfolders. Neither Google nor visitors can tell what the page is about from these URLs. Ideally, a URL should be readable without even having to open the page.

The problem usually doesn't stem from carelessness, but rather from the workflow. The CMS automatically generates a URL based on the page title or an internal ID; no one double-checks it, and the result is a URL that works technically but provides no indication of the page's content.

What you should do: Plan a descriptive, logically structured URL structure with readable words and hyphens before the first page is built. marschfahrt.de/webdesign instead of cryptic parameters. This may sound trivial, but it’s often overlooked because no one explicitly requires it. It’s best to map out the entire structure on a piece of paper before you start building, so no one has to improvise later on.

2. Redirects Are Often Overlooked During a Relaunch

If an existing website is redesigned and URLs change in the process without the old addresses being properly redirected to the new ones, the site can quickly lose a significant portion of its visibility on Google. Years of hard-earned rankings are then lost, often without anyone understanding why.

This happens more often than you might think, and usually not because of a lack of knowledge. The relaunch is complete, everyone is happy with the new design, and the redirect list slips to the bottom of the project plan because it doesn’t seem like a big deal, either technically or visually. That’s exactly what makes it so dangerous.

What you should do: Before every relaunch, create a complete list of all existing URLs and set up a permanent, server-side redirect (301 or 308) for each one directly to the corresponding new page, without any chained redirects through multiple steps. Don’t do this on a random basis—make sure it’s comprehensive. After the migration, verify that the redirects are actually working, and keep them in place for at least a year—preferably indefinitely. Expect it to take a few weeks for Google to start showing the new URLs instead of the old ones. This is one of the most costly mistakes that can happen during a website migration, and at the same time, one of the easiest to avoid.

3. Meta titles and descriptions remain blank or generic

The title and description in search results often determine whether someone clicks or keeps scrolling. Google doesn’t always use your page title and meta description verbatim; instead, it often compiles them from the page content itself. This makes it all the more important that both are specific and match the content. Nevertheless, on many websites, these fields remain empty or are automatically filled with the page name, without any actual content behind them.

This effect is often underestimated: Two pages can rank exactly the same position in search results yet receive vastly different click-through rates, depending on how well the title and description reflect the actual content. Ranking one position higher is of little use if no one clicks because the text below doesn't deliver on its promise.

What you should do: Write a unique, specific title and a description for each important page—one that actually describes what the page offers, rather than just repeating the company name. Imagine you see ten search results listed one below the other: What would your entry need to include for you to click on it specifically?

4. Images are not optimized for search engines

Large, uncompressed images are one of the most common causes of slow loading times. Google considers load time to be part of the overall user experience (Core Web Vitals)—one factor among many, but not the most important one. Missing alt text also deprives Google of the most important clue as to what is shown in an image. While Google also evaluates the image itself and the surrounding text, the alt text remains the most important piece of information.

With many projects, the problem starts right at the delivery stage: A photo comes from a camera or smartphone in its original size, is uploaded without any adjustments, and the page loads noticeably slower than it should. You rarely notice this right away, but it adds up across the entire page.

What you should do: Compress images before uploading them, use modern formats like WebP or AVIF, give the files descriptive names (for example, dachdeckerei-koeln-team.webp instead of IMG_4711.jpg), and describe each important image briefly and specifically in the alt text. This not only helps Google but also people who rely on image descriptions—for example, because they use a screen reader. Good alt text isn’t an SEO trick here; it’s simply a matter of accessibility, which also happens to help with rankings.

5. The heading structure gets all mixed up

A clear main heading and a logical sequence of subheadings below it make a page easier to read. However, Google does not specify a hard-and-fast rule such as “exactly one H1”; a clean structure primarily helps people and screen readers. In practice, we often see multiple H1 elements on a single page or headings that were chosen based purely on visual size rather than content structure.

This often happens when design and content are developed separately. Someone decides that a text element should appear larger, chooses the H1 formatting for that purpose, and suddenly the page has two or three main headings, even though the content was intended to have only one. Visually, this is hardly noticeable. But anyone using a screen reader will notice it immediately, and for Google, the structure becomes unnecessarily unclear.

What you should do: Treat headings as a table of contents for your page, not as a stylistic device. Use an H1, followed by H2s for the main sections, and H3s only where there’s a genuine subheading. This helps your visitors find their way around quickly, especially on mobile devices, where people tend to skim a page rather than read it. Google may also use a clear main heading for the title in search results.

6. Internal links are completely missing

Many websites consist of isolated subpages that rarely link to one another. However, Google uses internal links to understand which pages are particularly important and how content is related. A page without incoming internal links comes across as a footnote, even if the content is actually central to the site.

This doesn't just apply to blog posts. Service pages also rarely link to one another, even though they often should. Anyone offering web design and SEO should link from the web design page to the SEO page—and vice versa—because visitors often aren't yet aware of this connection themselves.

What you should do: Deliberately link thematically related content to one another, using actual links and anchor text that clearly states where the link leads, rather than “click here.” A blog post that links to the relevant service page not only helps visitors navigate further but also boosts the linked page’s ranking on Google. The easiest way to do this is to create a brief overview of which pages are thematically related; then, as you write, you’ll almost automatically find the right places to insert links.

7. Structured data is skipped

Structured data is additional, invisible information embedded in the code that helps Google correctly categorize content, such as business hours, reviews, or product information. It helps Google categorize content more clearly and is a prerequisite for many rich results in search results. However, there is no guarantee of this, and it does not directly affect the ranking itself.

This point is often overlooked because it’s invisible. No one sees structured data when looking at the finished page, so no one asks about it when the project is accepted. That’s exactly what makes it one of the items most likely to be missing once a project is considered “finished.”

What you should do: When building your website, determine which structured data makes sense for your business model and have it set up right away. For a local business, this primarily includes the name, address, and hours of operation; for an online store, it includes product details, prices, and availability. Be careful with reviews: Customer reviews of your own business that you mark up yourself on your website won’t appear as stars in search results. And FAQ or How-To markup is no longer worth the effort—Google no longer displays either of them. While this can be added later, it’s often forgotten once the project is considered complete.

Why it will be nearly impossible to make up for this later

The reason this list should be addressed before development begins—and not afterward—is simple: Each of these seven decisions affects the website’s basic structure. Changing the URL structure after the fact means having to set up redirects all over again. Correcting the heading structure after the fact means going through every subpage individually. What is a decision during development becomes a project in and of itself afterward.

Added to this is an effect that’s hard to quantify but is very real: Google needs time to re-evaluate a website. Even if all seven issues are corrected retroactively, it often takes weeks or months for that to show up in the rankings. Even a website that’s set up correctly from the start takes time for Google to crawl and index. But it doesn’t have to be fixed first and then re-evaluated a second time.

We're not saying this to scare anyone. We're saying it because we regularly take on projects where this very rework eats up the lion's share of the budget—a budget that would have been better spent during the initial construction.

That doesn't mean an existing website with these gaps is worthless. It just means that the order of the work is reversed: Instead of planning everything out properly from the start, fixes are made retroactively while the site is live and still running. Both approaches are possible. But only one of them is more cost-effective, faster, and carries less risk of losing good rankings during the transition.

A real-world example

A while back, a client came to us with a five-year-old website that had carried over exactly these issues from a previous relaunch. The URLs were automatically generated, there were no redirects from the previous version, the alt text was completely empty, and a handful of pages had duplicate H1 tags. The client felt the site was working well: modern design, fast load times on paper, and positive feedback from colleagues.

The situation was different on Google. The site barely appeared for the search terms that were actually relevant, even though the competition wasn’t any stronger in terms of content. The reason wasn’t the design, but rather the seven points listed above, which no one had explicitly checked during the relaunch at the time.

The correction took several weeks and cost significantly more than it would have to address these issues right from the start of construction. Some of the redirects could only be reconstructed using old Analytics data because no one remembered which URLs had even existed. This is exactly what we mean when we say that it’s nearly impossible to fully make up for this later: It’s not just about time, but also about information that becomes harder to obtain with each passing month.

Checklist for Your Next Web Project

Before your website goes live, you should have these points clarified:

• A clean, descriptive URL structure has been finalized

• During a relaunch, all old URLs are permanently redirected (301 or 308) to the corresponding new ones without any subpaths.

• Every important page has its own meta title and description

• Images are compressed and have alt text

• The heading structure follows a clear logic rather than being based purely on aesthetics

• Related content is linked internally as appropriate

• Structured data is included where appropriate

If you can check off these seven items before launch, you’ve already covered most of the technical SEO basics without having to set up a separate SEO project.

Conclusion: SEO isn't an afterthought

Most of these points take hardly any extra time to address during website development if they are considered from the outset. If they are fixed later on, they cost significantly more—and sometimes even result in a loss of search rankings that is very difficult to recover.

Good visibility on Google isn't achieved through some quick fix after launch. It comes from laying a solid foundation from the very beginning and from having the discipline to actually work through this checklist before the site goes live—not only after the first results fail to materialize.

Related: Google AI Mode and AI Insights: Why Your Website Is Getting Fewer Clicks and What Still Drives Traffic, and Local SEO—The Key to Regional Visibility · Services: SEO at a Fixed Price

Do you want to know how well your website is technically optimized for Google?

We'll take a look at your website and give you honest feedback on where your site stands in terms of the seven points listed above and what's most worth addressing first. Send us a quick message using the contact form. We'll respond within 24 hours—personally and without any sales pitch.

Sources

We have compared all technical recommendations with the documentation from Google Search Central. As of November 4, 2026.

• Best Practices for Site Moves That Involve URL Changes – Google Search Central

• Best Practices for URL Structure in Google Search – Google Search Central

• Influencing Your Title Links in Search Results – Google Search Central

• Control Your Snippets in Search Results – Google Search Central

• Image SEO Best Practices – Google Search Central

• SEO Starter Guide – Google Search Central

• Understanding Page Experience in Google Search Results – Google Search Central

• Link Best Practices for Google – Google Search Central

• Structured data markup supported by Google Search – Google Search Central

• Review snippet structured data – Google Search Central

• Latest Updates to Google Search Documentation – Google Search Central