Life In A Dollhouse

The homepage <h1> is “Ella’s Top Picks”. There is no value proposition, no scale statement, no maker story and no press above the fold — just a 631 KB collage image and a product grid.

Evidence — single <h1> on / reads “Ella’s Top Picks”. The meta description carries the positioning; no on-page element does.

A visitor who does not already know Ella cannot tell this apart from a dropshipper in the first five seconds. In a category full of generic import listings, provenance is the product.

Fix — State above the fold what this is and who makes it — designer-led original 1:12 designs, made by a named person in Minnesota, press logos beside it. Demote “Ella’s Top Picks” to an <h2>.

Two pages exist. /pages/featured-in holds real content — 2,578 characters. /pages/new-featured-in is effectively empty at 1,462 characters, indistinguishable from a blank page’s chrome. The nav links to the empty one, twice. The populated page has zero inbound links from the homepage.

Also empty — /pages/project-gallery at 1,470 characters, reachable from the nav as “Projects”. Two of seven main-nav items lead nowhere.

CNBC and House Beautiful coverage — the highest-trust asset a solo maker can own — is invisible.

Fix — Repoint the nav to /pages/featured-in, redirect the empty duplicate, and pull the two strongest logos onto the homepage and every product page over $100.

249 products across 19 collections. Sorting exists — Featured, Best Selling, Price, Alphabetical, Date. Filtering by room, colour, size or price does not. Search results carry filters but no sort control.

Evidence — product_type is blank on 181 of 249 products (73%), the field both Shopify filtering and Google Shopping key off. Only Rug (65), Sofa (2) and Gift card (1) are typed at all.

Compounding it — variant option names are inconsistent across the catalogue: Accent Color vs Accent color, Hood/Range color vs Hood/Range Color, Subway Tile Grout vs Subway tile grout color, Knob Color vs Main Color/Knob Color. Even with a filter app installed today these would render as separate, half-empty facets.

Someone furnishing a pink-and-green nursery has to scan seven paginated pages by eye.

Fix — Backfill product_type on all 249 products, normalise option names to a fixed vocabulary, then enable filtering by room, colour and price.

All 19 of 19 collections have an empty description.

“Dollhouse Rugs” and “Printable Dollhouse Wallpaper” are exactly the high-intent, low-competition queries this shop should own, and the landing pages for them are wordless. No context for shoppers, no text for search engines.

Fix — 80–150 words per collection on scale, materials and how to choose. Highest value first: Rugs (65 products), Wallpaper, Tile, Dollhouses.

Cyber Monday Sale and Cyber Monday Serene Dollhouse are still published and crawlable nine months out of season.

Fix — Unpublish or repurpose. Minutes of work, immediate credibility gain.

118 of 249 products — 47% — have no dimensions anywhere in the description. In a hobby defined by a ratio, this is the most damaging gap on the site.

Evidence — a search for any numeric measurement (", in, inch, cm, mm) matched only 131 of 249 descriptions. There is no size guide or scale reference page anywhere on the site.

Also thin — 20 products carry descriptions under 60 characters. 50 products have a single photograph; the catalogue average is 2.83.

A 1:12 miniaturist will not buy a $14 sofa without knowing whether it fits a ten-inch room box. This is the leading driver of both pre-purchase abandonment and “not as described” returns — which the refund policy already refunds in full.

Fix — Add height × width × depth to all 249 products as a structured field rendered in a fixed position on the page. Photograph a scale reference — a common object, or a printed 1:12 rule — in at least one shot per product.

No breadcrumbs. No related products. No “you may also like”. No room sets. No bundles.

Evidence — the raw product-page HTML contains zero breadcrumb, related-product or cross-sell markup.

On a catalogue with a $80 median and a $1.00 entry price, average order value is the whole game. A shopper who lands on one $1.50 plate from search has no path to the other 78 accessories. This is the largest single order-value lever available.

Fix — “Complete the room” modules driven by room tags, plus curated room-set bundles — kitchen starter, nursery starter — that turn one $14 purchase into a $90 one.

The buy box offers Quantity and Add to cart. No “Buy it now”, no accelerated wallet button.

Worth knowing — Shop Pay Installments is enabled and eligible. The $2,750 dollhouse offers 2 × $1,375, and financing plans run to $30,000. That is a real asset for the high-ticket line, and it is not mentioned anywhere in the copy.

Fix — Enable the dynamic checkout button. On printables, where the buyer wants the file in thirty seconds, it matters most.

Ten products are out of stock, including two of the three one-of-a-kind dollhouses — the Seaside Tiny Home at $1,250 and the Serene Tiny Home at $1,150.

The two highest-consideration pages on the site, reached by the most motivated visitors, end in nothing. No notify-me, no waitlist, no “commission something similar” hand-off.

Fix — Back-in-stock capture on every sold-out item. Every sold-out dollhouse routes to the custom enquiry — a $1,250 disappointment is the warmest commission lead this business will ever get.

No reviews app. No star ratings. No testimonials. No customer photography. The only third-party scripts on the page are shop.app and s.skimresources.com — an affiliate link monetiser worth reviewing, since it does not obviously belong on a first-party store.

For contrast — MiniatureCrush leads its homepage with “Over 80,000 Reviews and Over 250,000 Sales”.

A shopper is asked to spend $2,750 with a brand showing no evidence that anyone has ever bought anything. Ella has roughly 19,000 Instagram followers and years of customer photos: the proof exists, it is simply not on the site.

Fix — Reviews with photo capture, backfilled from Instagram and past customers, with ratings surfaced on product cards as well as product pages.

The refund policy states the shop “does not guarantee careful handling during shipping” once an item leaves. The shipping policy runs about eighty words: processing 1–7 days, custom 2–5 weeks, “refer to the shipping times listed during checkout”. No costs, no carriers, no international terms — despite the store offering seven currencies.

Also — returns open with a 7-day window and an email to a Gmail address. Who pays return postage, and whether restocking fees apply, are both unaddressed.

Disclaiming damage liability on fragile, hand-built, four-figure goods is the single most conversion-hostile sentence on the site.

Fix — Rewrite it to promise the opposite — packed by hand, insured, replaced if it arrives damaged. Add real shipping costs and an international section. Photograph the packing process; in this category it sells.

Custom kitchens run $475–$495 and dollhouses reach $2,750. The custom enquiries page has no form. It has four portfolio photographs and one instruction: email a Gmail address.

The page states no lead time, no price bands, no deposit terms, no process, and no answer to “what can I actually ask for”.

Compounding it — the contact page has no form either. Every page on the site carries the same four inputs, a search box and a currency selector, and the contact page is no exception. Every enquiry in the business routes through one personal Gmail address.

This hits the buyer with the most money and the most anxiety. A four-figure commission from a solo maker demands more reassurance than any other purchase here, and receives the least. Each unqualified email costs Ella a back-and-forth to price; each hesitant buyer who never writes is invisible.

No email capture, no popup, no welcome offer, no loyalty programme, no wishlist. No marketing app of any kind is installed. /blogs/news.atom returns 404 — there is no blog.

A category built on inspiration, seasonal builds and long consideration cycles, with no owned audience channel. Traffic arrives from Instagram, browses, and leaves without a trace.

Fix — Email capture with a lead magnet the shop can already produce — a free printable wallpaper sheet, or a 1:12 scale reference chart. Then a content engine: room tours, build diaries, how to wallpaper a dollhouse. Ella’s grandmother taught her wallpaper installation; that is a post, a video and a rugs-and-wallpaper cross-sell in one.

Digital downloads start at $6.50 and form a distinct product line — wallpaper, tile, wall mouldings. Nothing on those pages explains print settings, paper stock, sheet coverage, or how to apply the sheet at 1:12. The shopper is buying an image of an image with no way to judge how it prints.

Fix — State sheet size and coverage, specify print settings and paper, include an application guide inside every download, and show one photograph of the printable installed in a real room box.

Canopy 2.10.0 predates Online Store 2.0. No sections on every page, no native metafield rendering, no app blocks, no native filtering. The platform for reviews, filters, dimension fields, cross-sells and back-in-stock capture is simply absent.

Fix — Migrate to an Online Store 2.0 theme. Most of this report’s Phase 2 depends on it, and several Phase 2 items then cost nothing to implement.

The {width}x requests are a genuine bug: unresolved image-size placeholders are being fetched as literal URLs, returning about 2.8 KB of nothing each. Product photography is stored at 3024 × 4032 and served at a fixed 1024 pixels to every device.

This is an image-led brand on a slow, image-heavy site. Mobile shoppers arriving from Instagram — almost certainly the majority — pay for it directly.

Fix — Re-export the hero as WebP (631 KB down to roughly 60 KB), lazy-load below the fold, add responsive srcset, and fix the placeholder bug. Nearly all of this comes free with a modern theme.