What Running a Food Business Taught Me About Software

Before Vanda's Kitchen, I understood business software as a builder. After Vanda's Kitchen, I understood it as a user. The difference in perspective was profound, and it directly shaped everything we build at neart.ai.

As a builder, I evaluated software on architecture, scalability, and feature completeness. As a user, I evaluated it on one criterion: does it save me time or cost me time? If opening the software, doing what I need to do, and closing it takes longer than doing the same task manually, the software has failed. This sounds like a low bar. Most business software fails to clear it for small business users.

The bookkeeping tools available to a small food business were the first frustration. I needed to record purchases from multiple suppliers, track cash sales at markets, handle mixed VAT rates on food products, and produce reports that told me whether I was actually making money. The consumer-grade tools were too simple — they could not handle the VAT complexity. The professional tools were too complex — they assumed accounting knowledge that a food business owner does not have. The gap between the two was exactly where I needed to operate.

The scheduling and order management tools were similarly mismatched. Tools designed for restaurants assumed a fixed location with table service. Tools designed for catering assumed large events with long lead times. A food business that operates across markets, online orders, wholesale accounts, and occasional catering does not fit neatly into any existing category. I ended up using a combination of spreadsheets, notebooks, and memory — the universal fallback when software fails.

Compliance management was another gap. A food business needs to track food safety records, allergen information, supplier certifications, and environmental health requirements. The available tools were either expensive systems designed for large food manufacturers or paper-based templates that required manual entry and offered no analysis. There was nothing in between.

What struck me was not the absence of technology — it was the absence of thoughtful technology. The capabilities existed. Cloud software, mobile apps, automated data capture, intelligent categorisation — all of these were well-established. But nobody had assembled them into a product that worked for a small food business operator who needed things to be simple, accurate, and fast.

The most valuable lesson was about context. Software built by people who have never run the type of business they are building for misses the obvious things. It requires three clicks where one would do. It asks for information that does not exist. It presents data in a way that makes sense to a software designer but not to a business owner standing in a kitchen at 6am. Context — genuine understanding of how the software will be used, by whom, under what conditions — is the ingredient that most business software is missing.

This lesson became foundational for neart.ai. Every product we build starts with the user's context, not the feature list. How does a sole trader actually manage their bookkeeping? When do they have time to do it? What information do they need, and in what form? The answers to these questions shape the product more than any technical consideration.

Running a food business also taught me about the emotional relationship between a business owner and their tools. When your software works well, you feel in control. When it does not, you feel anxious. When it produces an error you do not understand, you feel helpless. These emotional states matter because they determine whether the software gets used consistently or abandoned after the first frustration. Building software that makes people feel capable rather than confused is a design objective, not a marketing aspiration.

Vanda's Kitchen continues to inform every product decision at neart.ai. It is the constant reference point for whether something is genuinely useful or merely technically complete.