Hi, I'm Dmitry

Full-stack web developer with a passion for UX and building products.

Vancouver, Canada and remote. He/him.

SEO story of a pet project

September 15, 2020

I launched MyGuitarTuner over 6 months ago – a side project just for fun.

It’s been visited by a large number of people, and it’s still growing. I didn’t promote it, just put it out there and played with SEO.

When a maker friend asked me what I did and if I could share any tips, I decided to write this article. I hope it’ll help other makers with their launches and set the right expectations for SEO growth.

Summary

Disclaimer: I’m no SEO expert. I didn’t know anything when I started, it’s a naive dev’s explorations in the world of SEO.

Key take-aways for me were:

  • It took really long for Google to index and get first views (1-2 months)
  • Just putting the site out there and relying on SEO alone can be very slow. Think months and years.
  • Having niche-specific pages can be very helpful for a few reasons. More about it later.

Monthly users report, last 90 days

Getting first views can be slow

Here’s the complete list of things I did when I launched it:

  • added the website to search engines for indexing
  • made sure the site has decent Lighthouse stats
  • made sure the robots.txt and sitemap look good
  • listed the app on AlternativeTo.com

That’s it.

Unsurprisingly, it took a long time for search engines to start showing the site.

Since then, I didn’t do anything to promote the app except for changing meta tags and text content and getting feedback from Google’s Search console.

How long it takes to gain momentum

Here’s the complete history of SEO impressions and clicks, from the very beginning (click to enlarge):

Overall SEO performance, from the beginning

How niche-specific pages helped to grow

This simple idea helped to get most of the traffic:

Having multiple landing pages, targeting different niches.

For example, MyGuitarTuner supports multiple instruments: guitar, cello, ukulele, etc.

Instead of making it a one-page app, I splitted the app into many separate pages, each having specific text content for that instrument.

Why it’s a good thing to do?

  • It helped to find underserved niches with low competition.

    E.g. there’s myriads of guitar tuners out there (good luck competing with them!), but only a few tuners for violins or cellos.

    I didn’t even know that violin tuners were a thing and people needed them!

    And yet most of my traffic comes from the violin and cello pages. Turns out there’s an entire niche of music teachers who use them.

  • It allowed for multiple parallel experiments with meta-tags and text content, speeding up the painfully slow feedback loops.

    E.g. when I found that one page did better then others in terms of click rate, I used wording of that page for other pages too.

    If I saw an improvement, it increased the probability that the new wording ranks better, more clickable (or both).

    Overall, I did around 15 changes/experiments with niche pages, optimizing their content, title and description tags. The feedback from the Goggle Search Console was delayed by a few days to a week.

Niche pages stats

Niche pages performance

That’s it! I hope it was helpful to get a better expectation for SEO indexing and growth timeline.

About MyGuitarTuner

TLDR: the most an advanced online microphone tuner for musicians.

  • it analyzes pitch of the played note in real-time (nothing new so far)
  • but it does so with one of the best available algorithms, with a very fast Rust implementation running in a WebWorker
  • …and it has an interactive playable guitar neck
  • …and it’s visual
  • …and…

You get the picture, it’s a labor of a geek’s passion.