Growth strategy
When to rebuild a page vs patch the conversion leak
A decision framework so engineering and growth stop fighting the roadmap.
Your landing page has a 1.2% conversion rate. Half the team wants to rebuild from scratch. Half wants to "fix a few things." Here's a framework for making this decision.
The patch-first principle
Default to patching. Always. A rebuild takes 4-8 weeks and resets all institutional knowledge. A patch takes hours and gives you data much faster.
When to patch
Patch when the page structure is sound but specific elements underperform:
- The page loads fast and looks professional, but conversion is low
- Heatmaps show people reading content but not clicking the CTA
- Bounce rate is normal but the conversion action isn't happening
- You can identify 3-5 specific issues from an audit
What patching looks like:
- Rewrite the headline for clarity
- Move the CTA above the fold
- Add social proof near the CTA
- Simplify the form
- Fix performance issues
Each patch is a testable hypothesis.
When to rebuild
Rebuild when the page has fundamental structural problems:
- Built for a different product/audience than what you sell now
- Information architecture doesn't match the user's mental model
- Tech stack is so outdated that small changes require major engineering
- You've patched 5+ elements and conversion still hasn't moved
That last sign is the key indicator.
The rebuild trap
Most rebuilds launch to... the same conversion rate. Why? The new page was designed based on opinions, not data. The old page's problems were never properly diagnosed.
Before rebuilding:
- Audit the current page thoroughly
- Talk to users — ask bouncers why they left, converters why they stayed
- Study competitors — understand expected conventions
- Write the content before designing — copy drives layout, not the other way around
The hybrid approach
Rebuild the critical section (usually above-the-fold) while preserving the rest. The top of the page does most of the heavy lifting. Rebuild that with a clear hypothesis and strong copy — often 80% of the lift at 20% of the cost.
Making the call
Run a structured audit first. Get the data. Identify whether problems are element-level (patch) or structural (rebuild). If you're not sure, start patching. The data from patches will tell you if a rebuild is necessary.
