Fancy Dress SEO
From a broken migration to 56% more search traffic - before the season even started.
I work with a UK children's fancy dress retailer on an ongoing engagement. When we started, a recent migration had quietly broken the site: about half the pages weren't resolving, roughly 300 pages were indexed from a site that should have had far more, and the site-wide average position was 28. Four months later, year-on-year organic clicks were up 56%, average position had moved from 28 to 11, and around 200 keywords across 100 product pages had shifted from positions 11-15 into 7-9. The full case study is here.
And that's the point about fancy dress SEO: the technical foundations decide the season before the season arrives.
Year-on-year organic clicks over the same 14-week window
Google Search ConsoleSite-wide average search position in four months
Google Search ConsolePages returning 200 OK - up from around 45% at the start
Crawl dataMobile pages with Good Core Web Vitals - up from zero
CrUX / Search ConsoleTiming is the central problem
Halloween is the commercial peak of the year for most costume retailers, and a Halloween category page needs to be indexed, established and ranking by early October. Google doesn't rank new pages overnight - a new category page in a competitive niche can take months to develop authority. So the pages get built in summer, the internal linking gets wired early, and the meta gets year-versioned so the same URL retains its equity season after season instead of starting from zero every year. That's exactly the architecture we're running into Halloween: lightweight seasonal category pages, merchandised from existing stock, live months before the peak.
Product pages are naturally thin - the fix is structural
A costume page has no specifications, no dimensions, no ingredients. On smaller costume sites, product pages often have to work as pseudo-category pages - capturing the occasion, the pairing and the search language of someone who's already decided what they're going as. That's a harder content problem than most of ecommerce, and it's won page by page: we target a handful of high-potential pages a week rather than spraying thin edits across the catalogue.
Entity analysis finds what competitors cover and you don't
On the category side, I run entity-level competitor gap analysis - software reads the top-ranking pages for a category and extracts the concepts Google associates with them, showing exactly what the winning pages cover that yours don't. Piloted on a flagship category, it turns "write better category copy" into a specific, prioritised brief.
Image search pulls real weight here
Costume shoppers want to see exactly what they're buying. Descriptive image filenames before upload, proper alt text, product image structured data - it's unglamorous, it gets deprioritised in favour of content, and in this niche it moves numbers on both Google Images traffic and conversion.
Out-of-season URL management
What happens to seasonal URLs in November? Keep live, seasonally de-emphasise or redirect - the same questions as fashion ecommerce at higher amplitude. The answer depends on the site's authority and the equity in each page, but the default is simple: seasonal URLs that took months to establish don't get deleted. They get maintained, year-versioned and re-merchandised.
A costume site that only performs in October has left its ceiling untested - the work that decides Halloween happens in spring and summer. If it hasn't started by August, you're already late for this season. And if your last migration left the site quietly broken underneath, that's the reason the spike never reaches its potential.