In the Age of AI, Data Is What Decides 

The advantage isn't sitting where it used to, and the companies that haven't noticed yet are going to find out the hard way.

Where the Advantage Used to Sit 

For a long time, software value lived at the surface - in the interface, in the features, in how readily people could navigate a workflow without needing to understand what sat underneath it. That was a reasonable place for value to live when interacting with systems required structure: forms, fields, process. The software guided the user through it, and how well it did that guidance was largely what you were paying for.

 When the interface stops being the work 

Now the interaction is changing: you don't need to navigate a system in the same way when you can ask it questions, or when you can move across functions without thinking about where things live, or explore without being routed across application or data boundaries. This means the surface level matters less than it did, and what sits underneath it - the data, and not just that it exists but that it's consistent, connected, and genuinely reflective of how the operation runs — is becoming the thing that actually determines whether the technology is useful or merely impressive.

The truth is that- once you remove the friction of the interface- the quality of everything below it becomes impossible to hide. It’s like the difference between a car that looks good and one that is good, and this is where the separation is happening: between systems that present information and systems that hold something closer to the truth.

 When data becomes the decision 

In shipping, that difference is hard to ignore and harder to paper over. Nothing in a shipping operation runs in isolation - commercial decisions affect technical performance, technical issues hit cost, compliance sits across everything simultaneously - which means the data that underpins decisions isn't a set of discrete workflows but a connected operation, and if it doesn't reflect that, it shows up quickly, not as a technical failure but as hesitation: a pause before a decision, a second check, a quiet doubt about whether the number being presented is actually right.

AI doesn't resolve that hesitation. It amplifies it, because now you're not just looking at outputs, you're relying on them - and if the underlying structure is solid, that's powerful, but if it isn't, the cracks appear faster and with less forgiveness than before.

The conversation is shifting accordingly: less about what a system can do, more about what it knows; less about features, more about whether the data can support a decision without requiring translation, stitching, or someone quietly running a check in a spreadsheet before signing off. This isn't a new idea - it's just been easy to overlook, because for a long time the interface did a creditable job of smoothing over what sat underneath it. That layer is thinning now, and what's left is what actually matters.

 What Actually Holds Up Under Pressure 

The companies that get this right won't necessarily look the most advanced from the outside, but they'll have something considerably harder to replicate: a version of the truth that holds up under pressure, across teams and systems and decisions, without needing to be verified against something else before it can be used. In a world where you can ask software anything, that's the part that decides whether the answer means anything at all.

Let's talk through your current setup

John Wills

COO, Shipnet