DEV Community

Cover image for Reflow Testing: Can a Peer Finish Your SaaS Demo After Zooming In?
Uriel Bitton
Uriel Bitton

Posted on

Reflow Testing: Can a Peer Finish Your SaaS Demo After Zooming In?

Reflow testing should follow a real task after the interface becomes narrow or enlarged. In a public SaaS demo, check whether a peer can still read the instructions, reach the controls, correct an error and complete the work after zooming in.

A screenshot of neatly stacked cards answers only part of that question. The trouble may appear when a dialog opens, a validation message wraps or a fixed toolbar covers the final action.

You can make a useful public review by choosing one harmless workflow, recording the viewing conditions and showing where the task stops. Keep the claim limited to what you checked; a reflow walkthrough is not a complete accessibility audit.

Understand the requirement you are checking

W3C’s explanation of WCAG 2.2 Reflow describes presenting content without losing information or functionality or requiring scrolling in two dimensions. For vertically scrolling content, the reference width is 320 CSS pixels. It gives 1280 CSS pixels at 400% zoom as an equivalent starting condition.

Some content needs a two-dimensional layout for meaning or use. W3C includes examples such as maps and data tables. That exception does not make every part of a dense application exempt, and individual table cells still need consideration.

Use the documented criterion to frame the check, then describe your actual environment. A browser's reported zoom, the viewport and the interface state all help another developer understand the observation.

Do not treat a phone screenshot as interchangeable with desktop browser zoom. They may share a narrow layout, but your public claim should say which condition you actually used. If you inspect both, keep separate notes.

Choose a task that crosses an interface boundary

Use a fictional review app for this worked example. A reviewer opens a sample document, leaves a comment, marks one section as needing changes and returns to the document list. No real customer records or messages are involved.

This task crosses several states: a list, a document detail view, a comment form and a confirmation dialog. That makes it more useful than inspecting the landing screen alone.

Write the expected result before testing: the comment is saved against the intended sample section, the status changes as requested and the reviewer can return without losing their place. Do not turn that into a script requiring one specific navigation route.

Choose sample text long enough to reveal wrapping problems. A one-word document title and a two-word comment may leave awkward constraints hidden. Use ordinary fictional content rather than a meaningless wall of characters.

Record whether the task begins at default zoom or after zoom has already changed. Also try changing zoom partway through the task if preserving the active state matters to your application.

Record the conditions a peer needs to repeat

A useful note identifies the browser and version, operating system, browser zoom, starting viewport and product build or date. It also identifies the account role and the exact sample workflow.

For example, a test plan might specify a desktop viewport 1280 CSS pixels wide at default zoom, followed by browser zoom at 400%. This is a proposed check, not an observation that this fictional application passed it.

If a browser or tool reports dimensions differently after zooming, state what you actually read rather than silently converting physical screen pixels into CSS pixels. The distinction matters when comparing notes across devices.

Include the starting page and sample record name in the request. A peer should not need to search a large demo catalog before reaching the state you want reviewed.

Keep the public reproduction note free of private tokens, real emails and account links that expose customer work. A screenshot can document a clipped control with fictional data just as clearly.

Follow the form into its difficult states

At the enlarged view, read the document heading and the instructions above the comment field. Check whether their full meaning remains available rather than assuming a tooltip or cut-off label is sufficient.

Enter a comment and deliberately leave a required field incomplete. Inspect how the error appears. Can the reviewer identify the field, understand the message and reach it without losing the rest of the task?

Open the status dialog. A fixed-height dialog might initially show its title and options while hiding the final action below the visible area. Check whether its contents can be reached and whether surrounding overlays interfere.

Then complete the action and return to the document. Make sure the confirmation and changed state remain understandable. A successful click is weaker evidence than seeing the intended result in the right sample record.

Finally, inspect any persistent toolbar or help widget while scrolling. A component that occupies modest space at default zoom can cover a substantial part of the enlarged working area. Describe the actual overlap instead of reporting only that the page looks crowded.

Treat a table exception narrowly

Suppose the fictional app includes a comparison table beside the document. A horizontal scroll area may be needed to preserve meaningful column relationships. That does not explain why the comment form next to it also forces sideways scrolling.

Review the sections independently. Ask whether each needs that layout for its meaning or function, rather than using the presence of one complex table to excuse the entire screen.

Look inside the table as well. Can a long status label wrap? Can the reviewer understand which document a row concerns? Does moving through the table hide the surrounding instructions?

Your public review can identify a question for deeper assessment: “The comparison table keeps a horizontal scroll area, while the adjacent form reflows. We still need to review how long cell text behaves.” That is more accurate than a blanket pass.

If you are uncertain about the exception, record the uncertainty and seek an informed accessibility review. Avoid inventing a compliance verdict from a single visual walkthrough.

Share a failure that another developer can investigate

A useful fictional bug report would say: “At the recorded desktop zoom, opening the status dialog places the final action below its scrollable region. The reviewer can select a status but cannot reach the action that saves it.”

Attach a screenshot of that state and describe the expected behavior. Include the steps before the dialog, because a sticky header or retained page position may affect what becomes visible.

Do not claim the CSS cause until you have inspected it. An observation about an unreachable button does not establish whether a fixed height, overflow rule or another element caused the problem.

When you fix the issue, repeat the whole task. A newly visible button is encouraging, but the comment, validation state and return path may still need checking. Compare the same conditions so the before-and-after images explain a specific change.

Make the public request narrow and honest

A post in Buildside’s founder community can connect a technical change to the work it supports. Explain the sample task, the viewing condition and the unresolved question before asking peers for feedback.

For example: “I’m reviewing the sample document flow after browser zoom. Can you reach the status dialog’s final action under these conditions? Please share the first point where the task becomes blocked.”

That invites a useful report without asking strangers to certify the whole product. It also gives developers a clear reason to inspect something beyond the polished release screenshot.

Close the public follow-up with the checked scope. Name the task and conditions, show the repaired state and leave untested browsers or other accessibility concerns visible as remaining work. The next useful progress update is a repeatable result another person can assess.

Hey I'm Uriel Bitton. I write about building in public strategies and growing startups.

Subscribe for more stories on growing your audience by building in public.

Join us on Buildside: the social network for founders building in public.

Sources

Top comments (0)