Design Case Study Carousel Cards with Numbers Readers Can Check

Put two comparable records on the workbench

Imagine a fictional task app called Handrail. Its illustrative review sheet counts requests that needed a follow-up to identify the next owner. The team also introduced a manual owner check between two four-week review periods.

The changing totals and manual owner check prevent attributing this hypothetical difference to Handrail. Keep that alternative explanation visible.

For the earlier work of deciding what a source can substantiate, use the customer-story video guide. This page concerns how the accepted evidence survives on separate carousel cards.

Build the result card before the cover

A weak result card would say “Follow-ups cut in half” above the count changing from twelve to six. That compares raw counts while hiding the smaller number of requests reviewed. A reader could infer a larger change than the comparable shares show.

A more complete illustrative result card contains:

• A plain heading: “Owner-identification follow-ups.”
• The baseline: “12 of 80 requests, 15 percent.”
• The later review: “6 of 60 requests, 10 percent.”
• The difference: “5 percentage points lower in this hypothetical example.”
• A visible condition: “Different request totals. Manual owner check introduced.”
• A source line: “Illustrative four-week review sheets A and B.”

Calculate the difference from the shares: 15 percent minus 10 percent equals five percentage points. Do not label that five percent. Avoid adding a relative-reduction headline merely because it sounds more dramatic.

The source line and condition must remain readable with the figures. If the card cannot hold them, simplify the visual or divide the evidence across clearly connected cards. Do not reduce the qualifications to decorative gray type.

Use different rules for a quote card

A quote card establishes what a person said. A result card presents a measured or calculated observation. One should not acquire the authority of the other through similar styling.

For an actual approved quotation, keep the exact words, speaker identification and scope needed to understand it. If the speaker says the workflow felt clearer, do not place an unverified percentage beside their portrait and imply they supplied that measurement.

Confirm the authorized name, role, quote, logo use and destinations with the responsible customer reviewer. Keep that approval attached to the final card version.

Use this design distinction:

Do not invent a quotation to fill a quote-card template.

Make the source note belong to its card

A final bibliography can identify the full source, but a card still needs local context. “Source: customer data” is too vague to explain which period or definition produced a number.

Use a concise source label that matches a private evidence record. The public card might name the approved customer report and period; the internal record can hold the original file, calculation, reviewer and permitted use. Keep private contact details and unapproved source material out of the public deck.

For the hypothetical Handrail sequence, a panel that shows period B should also identify period B. A screenshot of that panel should not look like a company-wide result with no time boundary.

Use the carousel-caption guide for longer methods. Essential qualifications still belong with the claim.

Trace the complete five-card argument

An original sequence for this example would give each card a different evidential job:

1. Cover: identify an illustrative review of owner-identification follow-ups, not a customer success headline.
2. Baseline card: show twelve of eighty requests, its four-week window and source A.
3. Later card: show six of sixty, the same definition and source B.
4. Comparison card: show fifteen percent versus ten percent, the five-percentage-point difference and the introduced manual check.
5. Interpretation card: explain that the example demonstrates a way to present evidence, not a causal product result.

The cover must remain true before someone reaches card four. Do not promise “the feature that fixed handoffs” when the later panel admits that a manual process changed too.

For native slides, inspect each card independently. For timed video, also allow enough viewing time to compare the figures. The image-versus-video guide helps choose the delivery without changing what the evidence says.

Review a correction across the whole deck

Suppose the hypothetical later count is corrected from six to nine, while its denominator stays sixty. The later share becomes fifteen percent. The original five-percentage-point difference disappears.

That correction affects the later-result card, comparison card, cover if it claimed improvement, post caption and any graphic or approved quote context that referred to the earlier result. Changing one numeral is not a complete repair.

Keep a card-to-source record so the editor can identify those dependencies. Recheck the actual exported images or video, not only the text field that was edited. Renew the relevant approval when the public claim changes.

Present the evidence without manufacturing it

Use Prolifik Image Carousel to assemble accepted card copy and permitted visual material. Inspect the actual slides and video separately when both will be used. The creation workflow cannot establish a missing result, approve attribution or supply customer consent.

Keep exact chart labels in controlled artwork or editable text rather than trusting generated art to reproduce the numbers.