Welcome to our first invalidity analysis report from Planda Portal. We believe in radical innovation and transparency: planning applications are complex, councils have varying requirements, and the system so far has been opaque. That's why we're sharing our findings, our journey, and the data that shapes our roadmap.

This is the first in a series of reports where we'll break down trends, celebrate wins, and show you what we're learning. The planning industry has historically lacked the data infrastructure to improve systematically so we're changing that. We're making our findings available to you, our users, and the wider planning community.

If you'd like to discuss this data or have questions, reach out to us at hello@plandaportal.co.uk.

What we learned

We've now submitted enough applications to understand how our users work, where mistakes happen, and what can be improved. Our goal is simple: make the application creation process easier, spot mistakes early, and speed up the approval process. Here's what the data showed us:

Our applicants received a 15% lower invalid rate than the previously reported 50% industry standard (data released by MHCLG). The Planda Portal invalid rate is closer to 35% at the time of writing this report.

Our AI flagged many of these issues ahead of time. 96.4% of applications that received feedback on invalid block plans had been flagged by our validator beforehand. Because our system doesn't block submissions, it just flags potential issues, some applications with issues still slip through. And that's fine. We learn from them, and find new ways to nudge users to make the right decision.

What happens when users take our advice? The numbers improve significantly. 85.8% of applications that passed all Planda Portal checks got validated by councils. Overall, our AI reduces the chance of getting an invalid letter by 40%. And we're getting better.

What we did about it

We want to learn from the trends and improve our AI systems, UX, and feature set to reduce invalid rates even further. Our product team refines the roadmap using data from user requests, wider market signals, reports like this one, and user behaviour. After analysing our user’s usage, we made some immediate improvements.

Launched the new site picker: Our location plan tool now lets you select your site without manual drawing. It automatically includes access to and from public paths, and allows you to edit the boundary afterwards to include everything you need, addressing the most common reason for invalid location plans (sometimes the best UX is removing steps entirely).

Improved the AI validator: This is the first time planning applicants have met an AI system that flags issues before submitting to councils. We've learned from user actions and improved both the UI and backend to learn even faster from how people interact with it. The new AI validator is in beta now - we expect it to roll out to everyone in early May.

Built a CIL generator: Pre-populated CIL forms save you time and prevent common mistakes. We pre-populate as much information as we can, meaning less manual data entry for users and fewer council rejection letters. Users are already loving it!

What's next

We want to reduce the invalid rate to 25% in the next two months.

We are already working this, and on new ways to improve the applicant’s experience on our platform, and improve the quality of applications received by councils.

Smarter application type selection: Many users don't know which application type to pick, leading to wrong fees. We already pre-populate fee calculations, but we can do better. We're building checks that cross-reference your project type against uploaded documents and flag mismatches before you pay.

Fun fact: In our first weeks, a user contacted us worried because they couldn't figure out how to calculate their fee. The problem? The system had already done it for them correctly. They were looking right at it but didn't believe auto-calculating fees was possible. That is what we want Planda Portal to do: to push the boundaries of what good UX means in planning applications.

Council-specific validation: We're training our models to handle council-specific requirements and building a dashboard for councils to verify these checks. This will close the gap between councils and applicants even further.

Our goal

We've kept invalid application rates below 50%, our ongoing goal, but there's still work to do. Every invalid letter is friction and every resubmission is lost time. We're analysing application outcomes constantly, tweaking the platform, and learning from each submission.

Stay tuned for the next report in a few months. We'll keep you updated on trends, changes, and what we're building next.