Allan de Wit

building · seo on autopilot

[Allan de Wit @dewitallan]

SEO has always been a hassle. I know it's crucial, but I'd start strong and eventually neglect it.

That's why we build https://t.co/xe6pdB2xTF, to put SEO on autopilot and focus on what we love: building awesome products.

[Allan de Wit @dewitallan]

Made a little animation for RankYak with @Remotion. Turned out nice. Super fun to work with too.

[Lars Koole @larskoole]

This is exactly what we try to avoid and solve with @rankyak_com.

Here's a simple case study of our own site.

We noticed a big lift after the December Core Update. The articles we'd been writing were actually helpful and Google finally recognized this. We spend a lot of time, effort, and money producing solid articles, so the recognition felt good.

Then a new Core Update came and we started losing traffic.

The reasons were obvious:
- We only produced new articles, never updated anything
- Internal links were always pointing backwards (article #1 could never link to article #2 or #999 since they didn't exist yet)

Step 1: Fix internal linking

We updated our internal linking pipeline and relinked every article, giving priority to articles within the same topic cluster.

Within a single day of republishing, our rankings started climbing again. You can see where the dip reverses in the screenshot. That's the day we republished.

Bonus: articles that Google previously ignored suddenly started getting indexed. Better link structure meant Google could actually discover them.

Step 2: Accept the hard truth

Relinking fixed the bleeding, but we had to be honest with ourselves. Some of our older articles were outdated. Written months ago, never touched since. Meanwhile RankYak itself kept evolving: better image generation, smarter enrichments, YouTube embeds, improved content quality across the board.

Our older articles didn't have any of that. They were stuck in time while the product moved forward.

So we asked ourselves: if we expect our users' content to stay competitive, shouldn't ours?

Step 3: Build Refresh and Rewrite

We started building two new features internally: Refresh and Rewrite.

Refresh takes an existing article that's ranking but could do better. It updates the content, adds missing enrichments, makes it current. A tune-up.

Rewrite is for articles that were ranking at some point but completely fell off. Full regeneration with everything RankYak has learned since.

Right now we're running both on our own site first. Not everything has been rewritten yet, not everything has been refreshed yet. We're monitoring the results carefully. Too early for hard numbers, but we haven't dipped since.

If everything works out the way we expect, we roll these features out to all users.

Step 4: The bigger picture

From day one, RankYak has used a topical map approach behind the scenes. Every project gets topics and keywords mapped out before a single article is written. The goal was always to build a cohesive topical entity, not just spam blog posts.

What changed recently is that we made this visible. We've built a full topical map view in the dashboard so users can actually see their topics, keywords, and how everything connects. No more black box.

Step 5: Where we're heading

Once your topical map is fully covered, there's no point in writing more new content just for the sake of it. The direction we're heading is an "optimize mode" where all energy goes into refreshing, rewriting, and keeping your existing articles at the top.

Because at the end of the day, none of this matters if the article doesn't answer the user's question. That's priority number one. Not word count, not keyword density, not publishing frequency. Does someone searching this keyword find what they need? If yes, Google rewards that. If no, no amount of SEO tricks will save you.

[Allan de Wit @dewitallan]

@larskoole @rankyak_com 👏 Cooking up the frontend for it right now

[Allan de Wit @dewitallan]

For those working on SEO: what's your biggest headache or challenge right now?

[Allan de Wit @dewitallan]

Just launched RankYak on @ProductHunt! 🚀

Would love your feedback and support:
https://t.co/mvkjx4AmMM

[Allan de Wit @dewitallan]

Is a free trial necessary?

[Allan de Wit @dewitallan]

Want people to read stuff?

After animating the text instead of showing it instantly, we found that users are actually reading it thoroughly instead of quickly scanning it

[Allan de Wit @dewitallan]

@minniti_ Thanks! We saw a big boost in signups after it. Onboarding completion dipped slightly, but overall the trade-off is totally worth it. Way more users getting in the door and trying the product

[Allan de Wit @dewitallan]

Quick tip: add an option to sign up with Google, it removes a huge amount of friction

More than 80% of all our signups are with a Google account

[Allan de Wit @dewitallan]

@KonradSkrobecki content + consistency = key

[Allan de Wit @dewitallan]

@faraz0x I like the "dogfooding" approach, build something you can also use yourself

[Allan de Wit @dewitallan]

Shipped a new article preview for RankYak

Way easier to read the generated article and get a quick glimpse of the structure

[Allan de Wit @dewitallan]

@forgebitz Very nice landing page! Gonna give it a try as well, thanks for sharing

[Allan de Wit @dewitallan]

@apoorvdarshan Mostly depends on your audience, adapt to their behavior.

I was surprised that most of my users do everything on mobile, even the more 'tedious' tasks.

[Allan de Wit @dewitallan]

@mmcgarvey03 Yeah it can be tricky. I like to start with a hyper-focused MVP, just the core problem/solution as a stable base. Spending money on ads is scary, but it's a very quick way to validate and get insights to guide further development. I do SEO for the long run and scale ads as I grow