I’ve been testing a few different internal linking workflows over the last couple of months on a couple of niche sites, and I’m curious what others are seeing in real projects.

My setup is pretty straightforward:

– WordPress sites in fairly competitive affiliate niches
– A mix of informational and commercial pages
– AI used mainly to suggest link opportunities, not to auto-publish anything blindly
– Manual review before anything goes live

What I’ve noticed so far is mixed.

On one site, improving internal links between related informational posts seemed to help crawl depth and page discovery pretty quickly. A few pages that were sitting on the second or third crawl cycle started getting picked up more consistently after I tightened up the internal linking structure.

On another site, though, the impact was much smaller than I expected. I added links from stronger pages to weaker money pages, but rankings barely moved. That made me wonder whether the issue was really internal linking, or if the content itself just wasn’t strong enough to compete.

A few things I’m testing:

1. **Contextual links vs. widget/footer links**
Contextual links still seem to matter far more. The AI tools I tested were decent at finding opportunities, but they weren’t always good at understanding actual topical relevance.

2. **Anchor text variation**
I’m being careful not to over-optimize anchors. In my experience, a natural mix performs better than repeating exact-match phrases everywhere.

3. **Cluster structure**
Pages grouped into clear topical clusters seem to benefit more than random “add more links” approaches.

4. **Commercial pages**
This is where I’m least convinced. Internal links help, but they don’t seem to overcome weak intent match or thin content.

5. **Automation limits**
AI is useful for scale, but it still misses nuance. It can suggest links that are technically relevant but awkward in the actual paragraph flow.

My current opinion is that AI-assisted internal linking is useful, but only if it’s treated as a workflow accelerator rather than a ranking hack. It helps you execute better structure faster, but it won’t fix poor topical coverage or weak pages.

I’m also a bit skeptical of tools that claim “one-click internal linking optimization” as if that alone will move rankings in a meaningful way. In practice, the sites I’ve seen improve usually had stronger content architecture first