“Did You Mean” suggestions catch typos and bad transliterations before the shopper bounces. They’re typically generated by combining edit-distance algorithms (Levenshtein, Damerau-Levenshtein) over the store’s actual product vocabulary, weighted by query frequency so popular intended queries float to the top.
For ecommerce, Did-You-Mean is one of the highest-ROI search features to ship: the shopper has already declared intent, and the only barrier is a one-character typo. Sites that auto-correct silently (and surface a “searching for X — show results for Y instead” banner) tend to outperform sites that show a list of alternatives.
Watch out for over-correction: medical, technical, or international stores often have legitimate words that look like typos to a generic spellchecker. Always train the corrector on your own catalog and search-query history, not a general English dictionary.