Gökhan Arkan

Building NestScore: A Property Evaluation Tool for London House Hunters

I’m buying a flat in London. If you’ve ever done this, you know it’s equal parts exciting and overwhelming. In the space of a few weeks, you might view a dozen properties across different neighbourhoods, each with their own quirks, trade-offs, and red flags. By the third weekend, they start blending together. Was it the Hackney flat that had the damp smell, or the one in Earl’s Court? Which one had the dodgy boiler? Did any of them have decent broadband?

I needed a system. Spreadsheets weren’t cutting it. I wanted something that would let me score properties consistently, weight what matters most to me, and actually compare options side by side. So I built NestScore. The fact that you can export your data to spreadsheets later is just a bonus, and yes, I realize that might contradict what I said about spreadsheets weren’t cutting it. But frankly, I don’t see myself filling out an Excel table on my phone during a property viewing.

The Problem with Property Hunting

Here’s how most people evaluate properties: they walk through, get a “vibe,” maybe take some photos, and leave with a general sense of “yeah, that was nice” or “definitely not.” Three viewings later, the details blur. You’re comparing a Victorian conversion in Islington against a new build in Stratford, and your brain is trying to weigh period features against a concierge service without any consistent framework.

Estate agents don’t help. They’ll tell you every property is “a fantastic opportunity” and “rarely available.” The EPC certificate tells you about energy efficiency but nothing about whether you can get decent mobile signal inside. Rightmove shows you the glossy photos but not whether the neighbours play drum and bass at 2 AM.

Over the years, I’ve looked at plenty of properties myself and watched friends and family go through the same process. The pattern is always similar: you fall in love with a place during the viewing, then discover the practical issues after you’ve moved in. The boiler that’s older than you are. The windows that let in every sound from the street. The “excellent transport links” that turn out to be a 15-minute walk to the station. These things add up, and they’re easy to miss when you’re distracted by a nice kitchen.

I wanted something structured. A checklist I could run through at every viewing, scoring each aspect consistently. Then I wanted to see those scores aggregated, weighted by what actually matters to me, and compared across properties.

The 37 Questions

The heart of NestScore is 37 questions across 10 categories. Each one represents something that either I’ve regretted not checking, or that friends and family have complained about after moving in.

Location & Transport covers the obvious: tube/rail access, bus routes, cycling infrastructure, parking. But it also asks about traffic noise, because that’s the kind of thing you don’t notice at 11 AM on a Tuesday but definitely notice at 7 AM on a Monday.

Local Amenities goes beyond “is there a supermarket nearby” to ask about variety: shops, restaurants, green spaces, healthcare. Some areas look great on paper but feel like ghost towns after 8 PM.

Safety & Crime includes both the objective (what does the crime data say?) and the subjective (does the area feel safe walking at night?). Data tells part of the story; your gut tells another.

Building & Structure asks about age, construction type, roof condition, damp issues. These are the expensive problems that estate agents conveniently forget to mention.

Utilities & Services covers water pressure, heating systems, boiler age, electrics. Anyone who’s lived with storage heaters or a temperamental boiler knows why this matters.

Internet & Connectivity might seem excessive with four questions, but for anyone who works remotely, the difference between 20 Mbps and 200 Mbps is the difference between productivity and frustration. Mobile signal matters too; not all London flats get reliable 4G indoors.

Energy Efficiency translates directly to running costs. With energy prices where they are, the difference between a C-rated and E-rated flat is hundreds of pounds a year.

Interior & Space covers layout, storage, natural light, ceiling height. The things that affect daily life but are easy to overlook when you’re distracted by staged furniture.

Outdoor Space asks about balconies, gardens, and communal areas. Lockdown taught everyone how much this matters.

Legal & Financial rounds things out with leasehold terms, service charges, and ground rent. The unglamorous details that can turn a good deal into a money pit.

Weighted Scoring

Not every category matters equally to everyone. For me, Internet & Connectivity gets high weight because I work from home. Someone else might weight Outdoor Space heavily because they have kids or dogs.

The scoring is simple. Each question has options with scores from 0 to 100. Category scores are averaged from their questions, then the overall score is a weighted average of categories.

Default weights reflect my priorities, but they’re adjustable. If transport doesn’t matter because you work from home full-time, lower it. If you’re buying with children, bump up Safety and Outdoor Space. The framework adapts to your priorities whilst keeping the evaluation consistent.

No fancy algorithms. No LLM integrations. Just the old good weighted averages you can reason about.

Sharing Without Servers

My wife and I both need to evaluate properties, but we notice different things. We needed a way to share.

The obvious solution is a backend with user accounts and syncing. But that means servers, authentication, maintenance. For a personal tool, that felt like overkill.

Instead, NestScore encodes the data directly into the URL. Serialise to JSON, compress, encode as URL-safe base64. The resulting link contains the entire property evaluation. Open it, decode it, import it. No server, no accounts, no sync conflicts.

There’s also a QR code generator. Compression keeps most properties under 2KB. Show your partner the QR code, they scan it, done.

URLs can get long, and properties with extensive notes might hit browser limits. But for the typical case, it works well. Sometimes simple is best.

Building It

I built the MVP with Copilot over a couple of hours. My wife was watching some high school drama on Netflix I didn’t follow much, and by the time she’d finished three episodes, I had a working MVP. The stack is nothing exotic: React, TypeScript, Tailwind, IndexedDB for local storage. Everything runs in the browser with no backend.

For those interested:

Sample Properties

When you first open the app, you can load four sample London properties to explore:

Each has all 37 questions filled in. Load them, compare them, see how the weighted scoring shows different strengths.

What’s Next

The app covers my core needs, but there’s more I’d like to add:

Try It

If you’re hunting for property in London, give it a go. Load the sample data, poke around, see if the framework matches how you think about properties.

The questions are opinionated, based on what I care about. If your priorities differ, adjust the weights. If you think I’ve missed something, the code is open source.

Property hunting is stressful enough without keeping everything in your head. A bit of structure helps.

Now, back to Rightmove. That Victorian conversion in Dalston isn’t going to evaluate itself!

URLs in this post: