How can text become corrupted after character conversion?
Text can be saved in one encoding, interpreted as another, and then converted again. The result can contain visible odd sequences where the original diacritics or symbols should have been. These remnants often appear after imports, exports, copy/paste, or integrations that disagree about character encoding.
The detector looks for sequences characteristic of this type of alteration and shows the context in which they occur. It does not claim the original character can always be reconstructed: a bad conversion can lose information, and the same observed sequence can require context for a safe remediation.
Examples showing why context matters
These examples are fictional and explain the kind of situation the detector can flag. They are not measured store findings and do not justify an edit without checking real product data.
“Café” in a title or text[can indicate bytes for “Café” were interpreted under the wrong encoding]This illustrates a common mojibake pattern; confirm the original form before editing.
A Romanian word contains several unexpected characters after importCheck the source and correct spelling before repairNot every suspicious sequence can be fixed with the same replacement rule.
A commercial symbol becomes multiple odd characters after importCompare with the product source or original feedThe finding locates the problem without inventing the lost symbol.
What does PROVEN check?
Checks sequences characteristic of text altered by an incorrect character conversion.
Why does it matter?
Text may contain incorrectly displayed characters, and the original form is not always safely recoverable.
Why is this useful to a merchant?
Corrupted text can affect titles, descriptions, or other visible values and make product information hard to read or unprofessional. Grouping suspicious sequences can also help identify a common import or conversion source.
For SEO, the practical value is readable, accurate text rather than an invented score. The page explains the phenomenon and detector without claiming every suspicious sequence creates a search penalty.
What can PROVEN do?
PROVEN can flag the sequence and context; repairing the characters requires your review.
How PROVEN handles this check
The workflow separates a reproducible finding from the editorial decision about the final value.
- Detect the suspicious sequencePROVEN looks for patterns characteristic of text altered by character conversions and preserves the observed value.
- Show context instead of guessingThe finding provides location and surrounding context so the correct form can be checked against product source data.
- Prepare repair under reviewWhen context supports a correction, PROVEN can prepare the edit and the merchant confirms the final characters.
Why might your review be required?
Mojibake can look recognizable, but automatic replacement without context can turn valid text into incorrect text. The original form should be confirmed when conversion may have lost information.
Check the import source, product data, or a trusted reference before applying a repair at scale.
What this page does not tell you
- The detector flags suspicious sequences; it does not prove the original encoding for every case.
- A sequence can resemble mojibake and still be intentional in a particular context.
- PROVEN should not guess letters, symbols, or diacritics when multiple reconstructions are plausible.
- Examples are illustrative and are not findings from your store.
- This page does not promise measurable SEO effects from repairing encoding.
Frequently asked questions
What is mojibake?
It is a common term for text displayed incorrectly after character data is interpreted or converted under the wrong encoding.
Can every corrupted string be repaired automatically?
No. Some patterns have an obvious correction, while other conversions lose information or permit multiple plausible reconstructions.
Why can it appear after an import?
Files, feeds, or applications can disagree about encoding declarations or interpretation, and an incorrect conversion can alter the text.
How should I verify the repair?
Compare against a trusted product source and apply the change only when the correct form is confirmed.
Sources and independent context
The sources below explain the relevant platform or standards. They provide context for interpretation but do not by themselves show that your store has this finding.
- Unicode Standard — U+FFFD substitution in conversionUnicode documents conversion errors, unconvertible sequences, and U+FFFD as a way to signal data that cannot be interpreted correctly.
- Shopify — Product details pageShopify documents where product titles and descriptions are managed, fields where corrupted text can become visible to shoppers.