Skip to content

You talk to the engineers, not an account manager hello@arktik.id

Photo by Duskfall Crew on Unsplash

Design & User ExperienceNovember 4, 20268 min readTika Aurora

Why an Inconsistent Interface Reads as an Untrustworthy Business

Customers cannot read your code, so they judge the business by whether the second screen behaves like the first one did.

Part of our complete guide: Design Systems: Keeping One Product Consistent as It Grows

Your customers cannot open the code and judge whether the work is tidy. What they can judge is what shows on screen: whether the second page behaves the way the first one did. When a button moves, a label changes name, and a date format flips halfway through, the conclusion lands on the company behind the screen rather than on whoever designed it.

Consistency is the cheapest trust signal a growing business owns. Inconsistency sends the opposite message without anybody deciding to send it, and that is why tidying up an interface is worth paying for.

What users actually notice when a screen changes

People who buy custom software usually cannot judge the technical quality of it, and that inability is one of their larger anxieties. The people using your application sit in the same position with far less time, so they read the surface instead.

The cues are small. The primary button sits bottom right on one page and top left on the next. The menu says Orders while the detail page says Transactions. A date reads 04/11/2026 in the list and 11 November 2026 on the receipt.

The expensive version of that happens on the pages a customer pays from, fills in booking details on, or presses submit on. Somebody who hesitates there does not message your admin to ask what is going on. They close the tab, and they do not come back.

The four places inconsistency usually creeps in

Inconsistency tends to arrive through naming, button behaviour, layout, and the tone of the writing.

  • One thing has three names across three screens, say customer, client and account holder, all reading from the same table.
  • A save button on one form saves immediately, while the same button on another form opens a confirmation first, so users stop trusting buttons.
  • The primary action moves corners between pages, so people hunt for it again every time they enter a new section.
  • One error message is polite and explains what to do next, while another blames the user or shows a code that means nothing to anyone outside the team.

Those screens were built months apart, by different hands, with no record anyone could open about how the previous screen worked.

Why it gets worse as the product grows

Every new feature is an opening to invent a new pattern. Whoever builds it needs a button, a table and a notification message. With no obvious place to check which version is already in use, they make their own, and the choice is reasonable on the day it gets made.

Growth adds screens faster than anybody's memory. A product with a handful of screens can stay consistent out of one person's head. A product with dozens cannot, and it certainly cannot after a developer changes. What closes that gap is one place a developer can open to see which button, table and message pattern is already in use, so the next person reuses it instead of inventing another. We set out how to build that in stages in the guide to design systems for scalable products.

Consistency as a trust mechanism, not a style preference

A predictable interface gives a user one mental model that they reuse across the whole product. Once somebody has learned how to cancel one order, they know how to cancel another without learning twice and without pausing before they press anything.

Customer portals, booking systems and online shops all depend on that kind of repetition, and so do the internal systems sitting behind them. The fuller argument about design quality and business results sits in the guide to the business value of UX design.

So we treat consistency as an operational matter. What we look for is the step where a user stops because they are unsure.

Internal tools have the same problem with a different bill

Staff cannot abandon an inconsistent internal system, so they route around it, and the workarounds are visible from across the room once you know the signs. Somebody keeps a shadow spreadsheet on their laptop because the report format in the system is not the format the manager asks for, which puts that part of the process back where it started. One person remembers the correct click order, and every new hire asks them. The same data gets entered twice because nobody is certain which page is the official one.

The bill arrives as long training time and data nobody quite trusts, which makes it harder to see than lost sales and usually means it gets left alone longer. Some of the patterns that recur in systems like this are covered in UX for complex internal systems.

How to find your own inconsistencies in an afternoon

You do not need a design audit to start. Pick one real task that gets done often, say a customer ordering through to the payment being recorded. Give the walkthrough to the person who runs that task every day, because they know the real click order, and have them screenshot every step.

Then line the screenshots up in order on one screen, or print them and spread them on a desk. What you are hunting for is the same job done two ways: two button styles for equivalent actions, two names for the same object, two positions for the primary action, two registers of error message. The list is usually longer than the person who commissioned the product expects, and it goes to whoever gets to decide which version wins.

Once that decision is made, write it down with the reason behind it. We record decisions like these next to the code, so the next person inherits the thinking rather than a pile of files. Without that record, a screen built six months from now guesses again from scratch.

What to fix first when you can't fix everything

Consistency work is bought one agreed stage at a time. Scope and price are written down before the stage starts, so you can stop after any stage without an argument and nobody has to commit to an open-ended redesign to find out whether it was worth doing.

Few businesses need to tidy an entire product at once, so the first stage takes the screens that carry the most weight: the ones on the money path, and the ones people open every day. The money path means the pages where a customer pays, fills in booking details, or sends the request you invoice against. Daily screens mean the pages your admin and counter staff keep open from opening until close. The admin corner somebody visits three times a year can wait, and leaving it untidy for now is a defensible decision.

On whether the fix covers its own cost, the arithmetic we normally use is in the ROI of design systems in scaling products.

Walk one real task in your product with us

There is no need to prepare design vocabulary or a long list of complaints. Name the one task your customers or staff run most often, and we will walk it together in the product you have now. We will point out where the screens contradict themselves and which part deserves attention first. The first call is free, there is no deck, and you are talking to the engineers who would do the work.

If all you need is a few simple agreements, we will say so. If there is real work in it, Discovery gives you a written plan and a fixed price, and that document is yours whatever you decide afterwards. Use the contact form, which opens WhatsApp, or email hello@arktik.id.

ui consistencydesign systemsproduct designinternal systemsdiscovery

Tika Aurora

Tika membantu bisnis menemukan solusi digital yang tepat. Dengan fokus pada hubungan klien dan pertumbuhan pasar, ia menulis tentang bagaimana teknologi, strategi, dan kemitraan dapat menghasilkan dampak nyata bagi bisnis.

Walk one real task in your product with us

A free first call with no deck, straight with the engineers. Name the task your customers or staff run most, and Discovery gives you a written plan with a fixed price that stays yours whatever you decide.

More from Design & User Experience