Week 11 of building Convertify a free image converter (Rust + Axum + libvips, Next.js SSG frontend) in public. Solo, no funding, 52-week run.
Here's the honest headline: **indexing jumped from 100 to 143 pages, I shipped image compression in Rust, and my clicks are still stuck.
What I shipped
1. Internal linking between blog and converter (the debt I kept dodging)
I launched a blog in week 10 (3 posts, full schema, the works) and immediately got 18 views/week on one post. Great except that traffic had nowhere to go. The blog talked about HEIC; it never pointed at the page that actually converts HEIC.
So week 11 I built a RelatedArticle component and wired it both ways:
- Blog -> converter: inline contextual links in the body + a CTA block at the end.
-
Converter -> blog: a "Learn more" card driven by a
related_blog_slugcolumn on each page.
The component went through a small evolution. First version showed one article. Then I made it multi-article it pulls 2–3 related posts, matches them by topic via slug, caps at 3, and (the part that bit me) guards against the literal string "NULL" sneaking in from a missing DB value. Nothing fancy, but it closes the loop: informational traffic can finally flow to the transactional pages.
2. Image compression a free TinyPNG alternative, in Rust.
This was the fun one. TinyPNG does ~3M visits/month. That demand is enormous and I was leaving it on the table. libvips already gives me everything I need, so the backend was fast to stand up:
- JPEG -> mozjpeg. libvips can hand JPEG encoding to mozjpeg, which does trellis quantization and smarter Huffman tables. Same visual quality, meaningfully smaller files than baseline libjpeg.
- PNG -> imagequant. This is the same lossy-PNG approach pngquant (and TinyPNG) use: quantize a 24-bit PNG down to an optimized palette. Huge size drops on PNGs with limited color, transparency preserved.
The UX is a quality slider (1-100%) and a before/after size readout, so you actually see the reduction. I shipped three landing pages /compress-jpg, /compress-png, /compress-webp positioned as the free alternative.
The interesting signal: Compress JPG Free pulled 24 views in its first week, landing straight in my top pages. Different user intent than conversion (people converting a format vs people shrinking a file), and it showed up immediately. That alone made the cluster worth it.
3. Blog post #4
"Why HEIC Files Won't Open" published, 8 FAQs, FAQ schema embedded in the BlogPosting. Continuing the content momentum and feeding the new internal-linking machine.
Also did a cannibalization audit across blog + landing pages. Came back clean: intent is cleanly split (blog = informational, landing = transactional), so they're not fighting each other in search.
The numbers
Google Search Console (3 months):
- Indexed: 100 -> 143 (+43, second-biggest jump ever)
- Not indexed: 65 -> 39
- Avg position: 40 -> 39.5
Google Analytics (7 days): 31 active users (+29%), 24 new (+33%), 52 sessions (+11%).
Traffic sources had a surprise: l.threads.com sent 13 sessions in a week Threads quietly became a real referrer. chatgpt.com has been a steady trickle for over a month now too (LLMs citing the blog posts).
Housekeeping wins
Cleared a stack of small debts that were quietly rotting: fixed the 404 source (broken internal links), cleared the redirect errors flagged in GSC, and consolidated my project docs into a single source-of-truth file instead of three drifting documents.
Next week (12)
- Two new tools that are low-effort / high-ROI: resize + crop and background removal.
- A compression-focused blog post to feed the new cluster.
- The real work: on-page pushes on my top-impression pages to finally break something into the top 20.
If you've shipped a programmatic-SEO site and watched impressions climb while clicks flatlined I'd genuinely like to hear how (or whether) you broke out of it. That's the wall I'm at.
Convertify is built with Rust + libvips + Next.js. Following along week by week.
Top comments (0)