Email guides

How to View Mailgun Inspect Dark Mode Previews in 2026: How QA Can Troubleshoot Email Display Issues

For developers and QA teams debugging HTML email appearance, Mailgun Inspect’s paired light and dark previews can isolate display differences within the selected client. This guide explains what those previews prove, how to investigate common visual problems, and when to run a separate controlled receipt test.

The logo is clear in light mode but disappears against the dark preview.

Use Mailgun Inspect’s Light & Dark Mode Dual View to compare the same email in both modes for the same selected client. Treat the result as a rendering check only: it does not prove that the message arrived, or that every recipient’s device will display it the same way.

For HTML email developers: isolate color, image, and layout changes that appear in dark mode.
For email QA: keep preview evidence separate from proof of receipt.
For indie developers: review your own product’s email before release without overstating what the preview validates.

Last updated November 5, 2026. We checked the feature name and scope against Mailgun’s official release note.

Start with the evidence the preview provides

Mailgun’s release note describes Light & Dark Mode Dual View as a way to compare the same email in light and dark mode for the same selected client. To use the comparison, select that client’s two modes. This makes it easier to spot differences without switching between unrelated message versions or clients. It does not turn the preview into a real delivery test, and it does not establish that every client or device will render the email identically.

How do you compare the light and dark versions side by side? Open the email in Inspect, choose the client you want to review, and select both light and dark mode for the comparison. Keep the email content and template version fixed while reviewing the two views. For the documented feature scope and operation, see Mailgun’s release note for the dual-view feature.

The comparison is most useful when you can point to a concrete change: a logo becomes hard to distinguish, body text loses contrast, or a call-to-action blends into its surroundings. Record the exact area and mode where you see the issue. Then check the original template before deciding what caused it. A preview shows an appearance; by itself, it does not explain the underlying CSS, image asset, or client behavior.

Observation in the preview What it supports What to inspect next
Logo or image looks different in dark mode A visible difference between the selected client’s two previews Image asset, transparent edges, nearby background, and template markup
Body or helper text is harder to read A legibility problem in the displayed combination Text color, background, hierarchy, and the affected content area
Button is less noticeable The action may be harder to identify in that view Label, boundary, surrounding content, and intended button treatment
Both previews look correct No visible issue in those selected preview conditions Run a controlled receipt check if actual received rendering matters

This table is a triage aid, not a pass/fail standard for all email clients. Mailgun’s Inspect documentation describes the product’s preview context; use the limits of that context when writing QA conclusions.

Compare paired views before changing the template

Start with a stable test input. Open the exact email or template version under review, select the client you need to inspect, and enable its light and dark views. Confirm that the subject, content, image assets, and template version match what the team intends to release. If the underlying email changes during review, a side-by-side comparison no longer isolates mode as the variable.

Then move through the message in order, from the header to the footer. Note where the appearance changes, not just whether the whole email “looks different.” A logo can remain visible while its edge disappears into the background; body text can be readable while a secondary label becomes faint; a button can retain its color but lose a clear boundary. Specific observations give the developer something reproducible to investigate.

What should a QA note capture when the previews differ? Record the selected client, the mode, the template version, the affected area, what the preview shows, and what the design is supposed to show. This makes it possible to distinguish an actual defect from a deliberate visual variation. Keep the screenshots or other review evidence with the same version reference used for the test.

This is the core of email dark mode QA: compare equivalent views, describe the visible result, and keep the conclusion inside the evidence boundary. Don’t infer that a particular CSS rule caused a change simply because the preview looks different. Check the template and assets before attributing a cause.

Diagnose logo and image changes

A logo that becomes difficult to see is often the most obvious signal, but “the logo disappeared” is not yet a diagnosis. Look at the asset and the area immediately around it in both previews. Does the image itself appear altered? Does a transparent edge become hard to distinguish? Does the surrounding background make the logo look less legible even though the image remains visible?

Separate what you can observe from what you still need to verify. The preview can show that the logo is hard to distinguish in one view. It cannot, by appearance alone, tell you whether the issue comes from the image file, its transparency, the template background, or how the selected client renders the message. Review the source asset and markup, then make one targeted change at a time so the next comparison has a clear purpose.

Check other brand images and decorative elements as well. A background image, icon, or colored panel can change the visual balance around the logo even if the asset itself looks unchanged. Compare the image with its surrounding content rather than judging it in isolation. If the brand treatment depends on a specific background, consider whether the image remains identifiable when that context is less prominent.

How should you investigate a logo or background that looks wrong in dark mode? First identify whether the visible problem is in the image, its edges, or the surrounding area. Next inspect the image file and the template markup. Make a focused adjustment, then repeat the same client and mode comparison. Don’t generalize a finding from one preview to clients you have not tested.

Restore readable text and background relationships

Review primary copy, secondary text, labels, and footer content separately. They serve different purposes and may not remain equally easy to read in both views. A page can look generally acceptable while an important instruction or small supporting label loses clarity. Note the exact text and background combination that causes difficulty.

When you revise a color, compare the new result in both modes. A change made to improve dark-mode readability can affect the light version, so don’t approve a fix by checking only the view that first exposed the problem. Preserve the intended hierarchy too: key instructions should remain distinguishable from supporting copy without relying on color alone.

Use accessibility guidance as a reference for evaluating text contrast, but don’t treat a preview screenshot as a complete accessibility assessment. The W3C explanation of WCAG contrast requirements explains how contrast is assessed. Apply the relevant criteria to the actual text and colors under review; avoid declaring a message compliant based only on a visual impression.

Can one dark-mode preview prove that text is readable everywhere? No. It shows how the selected client’s preview presents that message under the chosen mode. It does not establish the result across all clients or devices. Keep the finding scoped to the tested client and view, and add other targeted checks when your support requirements call for them.

Check whether the button still reads as an action

A button is more than a fill color. Review its label, shape or boundary, position, and separation from nearby copy in both modes. Ask whether the recipient can recognize it as an action and understand what will happen after selecting it. If the color remains visible but the label is difficult to read, the button still has a problem.

Compare the button with its immediate surroundings. A call to action may blend into a panel, or nearby links and text may compete with it. Note the precise concern and the intended design separately: for example, “the button boundary is hard to distinguish in the dark preview” is more useful than “the button is broken.” That distinction helps developers choose a focused fix and helps QA retest the same expectation.

After changing the treatment, compare both views again. Also confirm that the label itself is correct and that the surrounding content does not suggest a different action. A visually prominent button with an unclear label is not a successful correction.

Separate preview evidence from a real receipt check

A preview and a controlled test receipt answer different questions. Inspect can help you review how a message appears in the selected preview conditions. A controlled receipt check helps establish whether a test message arrived in a specific receiving environment and how it appeared there. Neither result should be substituted for the other.

If the side-by-side preview looks right, do you still need to test actual receipt? Yes, if release acceptance depends on the message reaching a mailbox or being reviewed in a real receiving environment. Send a controlled test message to an inbox you can inspect, then record the receiving environment and the template version. Don’t label preview-only evidence as successful receipt or real-inbox acceptance.

When you investigate a missing message, keep delivery and rendering separate too. Mailgun’s event types documentation describes events recorded during message processing. An event can inform delivery troubleshooting, but an event record does not itself prove that the message is visible in the recipient’s inbox or that it renders as intended. Keep the evidence tied to the question it can actually answer.

For test email receipt acceptance, define the receiving environment before sending. Use a controlled address and check the actual message in that environment. Record whether it arrived, where it appeared, and what it looked like. This result applies to that test setup; it is not a guarantee about every recipient’s mailbox or every client’s dark-mode handling.

Keep a review record that can be repeated

A useful acceptance record lets another person reproduce the observation without guessing which version was checked. Keep the template or message version, selected client, light and dark preview results, affected area, intended appearance, change made, and retest conclusion together. If you also performed a receipt check, keep its environment and result as a separate entry.

Use a short review sequence:

  • Identify the exact template or message version under review.
  • Select the client and compare its light and dark previews.
  • Record each visible issue by area and mode.
  • Check the original markup or assets before assigning a cause.
  • Apply a focused fix and repeat the same preview comparison.
  • Run a separate controlled receipt test when actual receipt or received rendering is part of acceptance.
  • State what was not tested so the conclusion does not imply broader coverage.

These checkpoints turn a subjective visual review into a traceable test without claiming more than the evidence supports. The final result might say that the logo and primary action were legible in the selected client’s two previews, while actual receipt was checked separately in a named test environment. If no receipt test was run, say so plainly.

Choose the right test route

Use this decision split to keep test effort proportional to the question:

  • If the question is whether one message looks different between light and dark views for the same selected client, use Inspect’s paired preview.
  • If the question is whether a real message reached a mailbox, send a controlled test and inspect the receiving environment.
  • If the question is whether a specific received message looks right in that environment, review the received message itself; don’t substitute a preview.
  • If the question is whether the email works across a wider client set, define that set and test each required environment. A paired view for one selected client is not a complete compatibility result.
  • If the test involves personal or sensitive data, use a controlled test dataset rather than exposing real user information. Our guide to data practices for email testing covers ways to keep test data separate from real personal information.
  • If you need an address for a temporary test rather than a long-lived account, choose it according to the test’s actual receipt and follow-up needs. Our guide to choosing a disposable address explains when a temporary inbox is suitable and when a durable address is safer.

A temporary inbox can be one option for receiving a controlled test message, but it does not reproduce every recipient environment or prove universal dark-mode rendering. Don’t use a temporary address for an account that needs long-term recovery, ongoing correspondence, or reliable access beyond the test. Inboxtmp is worth considering when the task is limited to temporary test receipt; pair it with Inspect’s preview for appearance checks, and keep the two results distinct. When actual arrival is the acceptance requirement, verify the message in the receiving environment rather than relying on the preview alone.

Confirmed this is a short-term task? Use an address now

Use a three-hour inbox to isolate one-off messages and finish the task before it expires.

Create a temporary address