What This Is Judged On
Whether people complete the task. Not how the screens look in a portfolio.
For business software that distinction matters more than it does for marketing sites. A beautiful interface where the most common action takes six clicks is a worse product than a plain one where it takes two. The people using it are trying to finish something, not admire it.
That is the standard we design to, and it changes what the work involves: more time understanding the job, less time on visual exploration.
Design As A System, Not A Set Of Screens
The single decision that most affects cost and consistency.
Screens handed to a developer produce individually built screens. Similar things end up behaving slightly differently, the interface drifts as it grows, and changing anything global means touching everything.
A system defines the reusable parts and the rules for combining them: what a button is, what a table does, how forms behave, what the spacing scale is. Then screens are assembled from parts rather than drawn.
That is cheaper to build, cheaper to extend, and it stays coherent. It is also what makes a brand change a single edit rather than a project. Ask for it explicitly, because the default deliverable in a lot of design work is still a set of pictures.
What Actually Improves Business Software
Reducing the number of decisions per screen. Most business interfaces show too much because everything was requested by somebody.
Making the common path fast and the rare path possible, rather than treating them equally.
Sensible defaults. Every field pre-filled correctly is a decision removed.
Honest empty states. What a screen shows before there is any data is where new users form their impression, and it is nearly always an afterthought.
Error messages that say what to do, not what went wrong.
What Moves The Price
Number of distinct screens, and how much they share.
Whether a system exists already or is being created.
Research depth. Talking to actual users changes the outcome and costs time. Frequently worth it for internal software where the users are down the corridor.
Accessibility requirements, which are much cheaper designed in than retrofitted.
Mobile. Not a resize. If people use it on phones, that is a design constraint from the start.
Where Design Goes Wrong On Software Projects
Designing the happy path only. Real screens have errors, empty states, long text, missing data and slow connections. A design that ignores those gets those decisions made by a developer under time pressure.
Beautiful and unbuildable. Designs that assume data will always be short, lists always brief and names always fit.
No handover conversation. Design and build treated as separate phases produces an interface that looks like the design and behaves differently.
Optimising the wrong screen. Effort on the dashboard everyone sees once, none on the form used forty times a day.
Questions To Ask Whoever Designs It
- Will we get a system or a set of screens?
- What happens in the error and empty states? Ask to see them.
- How does the busiest screen behave with real data volumes?
- Have you spoken to anybody who does this job?
- What are the five most frequent tasks, and how many steps each?
A Brief You Can Send Anyone
What the software does: [and for whom]
The most frequent tasks: [with rough daily counts]
Who uses it: [technical confidence, training, how often]
Devices and conditions: [desk, phone, on site, poor connection]
What exists now: [screenshots of the worst screen are useful]
Accessibility requirements: [any obligations]
What we are told most often: [the recurring complaint]
How We Work
Understand the job first, design the system rather than the screens, cover the unhappy paths, and hand over with the developers in the room.
Where the build follows directly, web application development covers that side, and having the same team across both removes the gap where consistency is usually lost.
Get in touch with the screen your team complains about.