The honest story of how search data, failed pages, changing keywords, and small technical improvements shaped my website.
When I started building Letras Diferentes, I was not thinking like an SEO specialist.
I was thinking like a creator.
I wanted to make a useful place where people could transform ordinary text into stylish letters for social media bios, gaming nicknames, messages, and profiles.
The idea seemed straightforward:
Build a text generator.
Publish it.
Wait for people to discover it.
That third step was where reality began.
A functional website does not automatically become a visible website. A beautiful tool does not automatically appear in Google. Publishing a page is only the beginning.
My SEO journey did not begin with a perfect strategy, an expensive tool, or years of technical experience.
It began with confusion.
Then came impressions without clicks, keywords I had never expected, pages competing with each other, titles that needed rewriting, and many small decisions that gradually changed how I understood search.
This is the story of what I learned while growing Letras Diferentes from one simple idea into a collection of search-focused tools.
I originally treated SEO as a publishing checklist
At first, my understanding of SEO looked something like this:
Add a keyword to the title
Repeat it in the article
Write a meta description
Add some headings
Publish the page
Get backlinks
Rank on Google
Technically, those activities can be part of SEO.
But they are not a strategy by themselves.
I was treating search optimization like a checklist placed on top of the product. Later, I realized that good SEO had to influence the product itself.
It needed to shape:
Which tools I built
How I named them
What each page promised
Which questions I answered
How pages connected to one another
What visitors saw before and after using a tool
Whether the page satisfied the search that brought them there
That realization changed everything.
SEO was no longer something I did after creating a page.
It became part of deciding what the page should be.
My first mistake: targeting one broad keyword
The main idea behind the project was represented by one obvious phrase:
letras diferentes
It was relevant, popular, and closely connected to the product.
Naturally, I focused heavily on it.
The problem was not targeting the phrase. The problem was believing that one broad keyword could represent every type of visitor.
Someone searching for stylish Instagram letters does not necessarily want the same thing as someone creating a Free Fire nickname.
A user looking for cursive text may have a different goal from someone looking for an invisible character.
They may all use Unicode-based tools, but their search intentions are different.
That led me to build dedicated experiences around specific needs:
Stylish text for social profiles
Gaming nicknames and symbols
Small text
Cursive letters
Strange and unusual text styles
Invisible spacing characters
Fonts for Instagram and Discord
Football-inspired names and supporter text
The website became easier to understand because each page had a clearer job.
This taught me my first major SEO lesson:
A website should not force every search intention through one page.
One strong homepage is useful. A focused group of pages is much more powerful.
My second mistake: building pages around keywords instead of problems
After learning about keyword research, I began seeing phrases as opportunities.
A keyword had search volume, so perhaps I needed a page for it.
But a keyword alone does not tell the complete story.
The better question is:
What is the person trying to accomplish after typing this phrase?
For example, someone searching for an invisible character is probably not looking for a long history of whitespace in Unicode.
They may want to:
Create a blank-looking game nickname
Separate words in a username
Hide an Instagram highlight title
Send an empty-looking message
Copy a character that a platform will accept
That changed how I developed the invisible character generator.
Instead of only presenting a blank symbol, I added:
Multiple character sizes
Copy buttons
Quantity controls
A field for testing pasted characters
A character counter
Examples for different platforms
Explanations of what to try when one option fails
The page became stronger not because it contained more keywords, but because it handled more of the user’s journey.
That distinction matters.
A page can be perfectly optimized around a phrase and still disappoint the visitor.
Search engines may bring someone to the door. The product must give them a reason to stay.
Search Console became my most honest collaborator
Before using real search data, I made many assumptions about what people wanted.
Google Search Console challenged those assumptions.
It showed me the exact language people were using to discover the site.
Some searches matched the words I expected. Others revealed new opportunities, including variations related to:
Personalized letters
Letters for Free Fire
Nickname generators
Small letters
Text modifiers
Name modifiers
Social media fonts
That information was more valuable than guessing.
It showed me that people do not always describe a product using the terminology its creator prefers.
I might call something a Unicode text generator.
A visitor might call it:
A name modifier
A font changer
A nickname creator
Pretty letters
Different text
Fonts to copy and paste
None of these phrases is necessarily wrong.
They are different ways of expressing the same underlying need.
I began treating search queries as product feedback.
If Google repeatedly showed a page for a particular search but users did not click, I asked:
Does the title clearly match the query?
Is the description too generic?
Is another page more appropriate?
Is the page ranking for an intention it does not fully satisfy?
Does the search result make the benefit obvious?
Search Console stopped being only a performance dashboard.
It became a record of how real people understood the product.
Impressions initially made me feel more successful than I was
Seeing impressions grow is exciting.
It means Google is beginning to show the website.
But an impression is not a visit.
A page can receive thousands of impressions while attracting very few clicks. It can also rank for many irrelevant variations that do not bring useful visitors.
At first, I celebrated visibility without examining its quality closely enough.
Later, I began paying attention to the relationship between:
Query → Position → Impression → Click → Page experience
Each stage exposed a different problem.
High impressions but a low click-through rate
The page might need a clearer title or description.
A good position but limited impressions
The keyword might have low demand, or the page might need to cover related intentions.
Clicks but poor engagement
The result may have promised something that the page did not provide quickly enough.
Several pages appearing for the same query
The website might be confusing Google about which page is most relevant.
This taught me not to judge progress through one number.
SEO performance is a chain. Improving one weak point can help the entire journey.
I rewrote titles more often than I expected
A page title can look good to its creator while remaining unclear to a searcher.
Some of my early titles tried to include too much:
The primary keyword
Several secondary keywords
The word “free”
Multiple platforms
Copy-and-paste language
Branding
Extra promises
The result was often a title that felt crowded.
I gradually learned to prioritize three things:
The main task
The specific audience or use
A clear benefit
A simplified structure looked like this:
Primary task + relevant use case + clear action
For example:
Letras Pequenas: Gerador de Fontes Pequenas para Copiar
The title communicates what the page contains and what the user can do.
It does not need to explain the entire page.
That was another important lesson:
The purpose of a search title is not to contain every keyword. It is to earn the correct click.
Keyword repetition was not the same as relevance
Like many beginners, I worried about whether the primary keyword appeared enough times.
Should it be in the first paragraph?
How many headings should include it?
What percentage of the article should it occupy?
These questions are not meaningless, but they can become a distraction.
When I focused too heavily on repetition, paragraphs began sounding unnatural. The page became less enjoyable for the person it was supposed to help.
I started using a broader language model around each topic.
A page about gaming names could naturally discuss:
Nicknames
Player names
Symbols
Guilds
Profiles
Spacing
Free Fire
Copying and pasting
A page about Instagram text could discuss:
Bios
Captions
Profile names
Highlights
Stories
Stylish text
Social identity
This created relevance without repeating the same phrase in every paragraph.
The goal was not to hide the keyword.
It was to stop making the keyword carry the entire article.
Internal links taught me to think in journeys
Initially, internal linking felt like another mechanical SEO task.
Find a phrase. Add a link. Repeat.
But random links do not necessarily help the visitor or the website.
I began thinking about what the user might logically need next.
Someone creating a gaming nickname may also need:
Symbols
An invisible separator
Small letters
Unusual character styles
Someone improving an Instagram bio may also want:
Cursive text
Small text
Invisible spacing
Decorative symbols
A useful internal link continues the task instead of interrupting it.
For example:
<p>
Para separar partes do seu nome sem mostrar um espaço comum,
experimente um
<a href="/caractere-invisivel/">
símbolo que permanece visualmente em branco
</a>.
</p>
The anchor explains why the destination is relevant.
This is better than inserting a list of unrelated keywords at the bottom of every page.
Internal linking became less about distributing authority and more about designing movement through the website.
I learned that design and SEO were not separate departments
Letras Diferentes is a visual project.
People visit because they want to change how their text looks.
That means design is not only decoration. It is part of the search experience.
Each tool needed to feel connected to its purpose.
A gaming page could use a stronger, more energetic interface. An Instagram tool might feel lighter and more polished. An invisible-character page needed clarity because the output itself could not be seen.
However, visual design could not come at the expense of usability.
I had to consider:
Can the tool be used comfortably on a phone?
Is the copy button easy to find?
Does the result appear immediately?
Are headings understandable?
Is the text readable?
Do animations slow down the task?
Does the design make the tool feel trustworthy?
Can visitors identify the next step without instructions?
A page that looks impressive but makes the user work too hard is not fully optimized.
Technical SEO, content SEO, and interface design eventually began pointing toward the same objective:
Remove unnecessary friction between the search and the solution.
Structured data helped me describe the product more clearly
As the website developed, I began adding structured data to appropriate pages.
For an online generator, a basic application description might look like this:
<script type="application/ld+json">
{
"@context": "https://schema.org",
"@type": "WebApplication",
"name": "Letras Diferentes",
"applicationCategory": "UtilitiesApplication",
"operatingSystem": "Any",
"url": "https://letrasdiferentes.online/",
"description": "Ferramenta online para criar letras personalizadas, nomes, bios e estilos de texto para copiar e colar."
}
</script>
Structured data is not a shortcut to rankings.
It does not rescue weak content or a confusing tool.
But implementing it forced me to answer useful questions:
What exactly is this product?
What category does it belong to?
What does it allow the user to do?
Which page represents the application?
Which questions are genuinely answered?
The code was useful, but the clarity behind the code was even more useful.
My backlink strategy also had to mature
At one stage, I viewed backlinks mainly as a number.
More profiles, more mentions, more websites linking back.
But not every backlink has the same value.
A random profile created only to place a URL may do little for the project. A useful technical article, product listing, tutorial, or genuine community contribution can create much stronger context.
I started publishing the story behind the product instead of only promoting the product itself.
My second explored the failures behind engineering an invisible-character tool.
These articles do more than create links.
They demonstrate:
Technical understanding
Product decisions
Problem-solving
Real development experience
The human story behind the website
That is the kind of visibility I want to build.
Not merely a trail of URLs, but a trail of useful work.
The biggest challenge was resisting constant changes
SEO encourages experimentation, but it can also create anxiety.
A title changes, and I want to check the ranking immediately.
A page gains impressions, and I want to rewrite everything.
A competitor uses a phrase, and I wonder whether I should add it too.
The problem is that search engines need time to crawl, process, compare, and reassess a page.
Changing too many elements too quickly makes it difficult to understand what actually worked.
I had to learn a more disciplined cycle:
Observe → Prioritize → Change → Document → Wait → Compare
Not every low-performing query requires a new page.
Not every ranking fluctuation requires an edit.
Not every competitor should influence the product.
Sometimes the right action is to leave a strong page alone long enough to collect meaningful data.
For a founder, patience can be one of the hardest SEO skills.
SEO forced me to become more than a content writer
Working on Letras Diferentes changed my understanding of digital skills.
SEO was not a single isolated discipline.
It required me to combine:
Content writing
Copywriting
Keyword analysis
Interface design
Basic development
Analytics
User psychology
Branding
Technical troubleshooting
Product strategy
A title change required copywriting.
A new landing page required search-intent research.
An internal-linking decision required information architecture.
A tool improvement required development.
A low click-through rate required both data analysis and empathy.
This is why I no longer think of SEO as “writing for Google.”
It is the process of aligning a useful product with the language and expectations of the people searching for it.
What I would do differently if I started again
I would still make mistakes, but I would begin with a clearer framework.
- Define one purpose for every page
Each URL should answer a distinct need.
- Study the search result before writing
The existing results reveal what Google believes the query means.
- Build the tool and content together
The explanation should support the product, not compensate for a weak product.
- Measure queries by intention
A high-volume phrase is not valuable when the page cannot satisfy it.
- Create internal links around the next logical action
Navigation should reflect user journeys.
- Document important changes
Without a record, SEO experiments become guesses.
- Protect the human voice
A page should sound like it was written to help someone, not to satisfy a keyword counter.
- Build authority through useful experiences
A real product story is more memorable than another promotional profile.
The numbers mattered, but the learning mattered more
Watching the website begin to receive more impressions and clicks was encouraging.
Seeing pages appear for queries I had intentionally targeted felt like evidence that the work was beginning to connect.
But the most valuable result was not one position, one keyword, or one traffic milestone.
It was learning how to think more carefully.
I learned to question my assumptions.
I learned to read search data without reacting emotionally to every change.
I learned that a page can be technically optimized and still fail the person using it.
I learned that SEO is strongest when it improves the product rather than simply decorating it.
Most importantly, I learned that expertise can be built publicly.
I did not wait until I knew everything before starting Letras Diferentes.
I started with what I knew, discovered what I did not know, and improved one decision at a time.
Final thoughts
Letras Diferentes began as a creative text tool.
SEO transformed it into a much broader learning journey.
Every page taught me something different:
How people describe their needs
Why search intention matters
How design affects trust
Why technical details influence visibility
How small improvements can compound
Why honest product stories create stronger authority than empty promotion
I am still learning.
The website is still evolving.
There will be more experiments, pages that underperform, titles that need improvement, and assumptions that search data will prove wrong.
But that no longer feels like failure.
That is the process.
Building in public has taught me that progress rarely comes from one perfect strategy. It comes from repeatedly looking at what people need, what the data reveals, and what the product can do better.
That is how I am building my project.
One page, one query, and one lesson at a time.
Top comments (0)