Furniture and home décor, based in Lahore, shipping nationwide across Pakistan. Almost all our orders are cash on delivery — card adoption is low in this market, so COD isn’t optional for us.
We migrated from Ella to Impulse 9.2.0 in July. Traffic has held up, but conversion hasn’t come back to where it was. I’ve already found one real migration bug: the collection header block came across disabled in templates/collection.json, so every collection page lost its H1 and breadcrumbs. Nothing showed as broken in the theme editor — it only turned up in an SEO crawl. That’s fixed now.
What I’d like other eyes on:
Product page trust. Our average order is high for this market and it’s COD, so the customer is agreeing to accept a large delivery at their door from a store they may not have bought from before. What signals are missing that would make that feel safe enough to commit?
Mobile above the fold. Most of our traffic is mobile on patchy connections. Does the first screen do enough work before someone bounces?
Anything that reads as broken or half-finished after the migration that I’ve stopped seeing because I look at it every day.
Not looking for “add more reviews” — I’d get more out of specifics on what you’d change first and why.
I checked the live store rather than relying on the theme editor. The first thing I’d change is the homepage hero: “Made for living / Where spaces begin” currently has no button or link, so on mobile the first screen gives no obvious next step. Add one clear CTA to a high-intent collection (for example, ready-to-ship or best-selling furniture) and put “Cash on delivery across Pakistan” in the hero/announcement rather than waiting until the product page.
On the product page, the COD line beside Add to cart is useful. The remaining trust gap is what happens after the click. I’d add a short three-line block under the button, using only policies you actually follow:
how the COD order is confirmed (call/WhatsApp/SMS),
the real delivery window and when shipping cost is confirmed,
what happens if a large item arrives damaged or does not match.
For a high-ticket item like the Aesthetic Decorate Mirror, the current copy is still very generic. I’d add the exact weight/material thickness, whether mounting hardware is included, packaging/damage handling, and installation availability if you offer it. Those details answer the risk a COD buyer is actually evaluating.
I would test this on one collection and one high-AOV product before changing the whole store. Track mobile Product view → Add to cart → Checkout → Order for the same traffic source. If the first two improve but completed orders do not, the next issue is likely checkout/delivery certainty rather than the theme layout itself.
If conversion fell while traffic held after Ella to Impulse, I’d treat it as leftover theme wiring, not another homepage pass.
On Impulse’s product template, add a few Sales point blocks near the buy buttons. Archetype documents those as icon plus short text for shipping or returns style friction. For high ticket furniture, connect them to product metafields with Connect dynamic source so a sofa and a small décor piece don’t share one blanket promise. There’s also a Trust badge block if you want one image row for COD or delivery proof you actually use.
On mobile, Impulse has a Mobile layout control for the main image (full width vs 75%). Worth checking whether the first screen is mostly photo, with title, price, and ATC pushed below the fold.
Same class of bug as your disabled collection header: open the product templates and look for disabled blocks, empty tabs, and app embeds that never got turned back on after the theme switch. App blocks and embeds are per theme, so Ella ones don’t carry over by themselves.
At checkout, native COD is a manual payment under Settings > Payments. Fill Additional details so the COD option itself explains what happens next. Shopify’s estimated delivery dates mainly show in checkout, not as an Impulse product page estimate, so don’t wait for an auto ETA on the PDP for Pakistan COD.
I haven’t opened your live store. This is an editor checklist only.
Hey @ideagen, one gap neither checklist above catches: open DevTools Console + Network on a live product page, collection page, and the cart drawer separately. Ella and Impulse load different JS bundles, so a leftover snippet or app embed still pointing at an Ella-specific script/asset will often 404 silently in the background without breaking the visible layout, exactly the kind of thing you stop noticing day to day. If you want a full fresh-eyes pass, our free audit tool at audit.ecomswiftllc.com is built for catching this kind of leftover-after-migration issue too.
Your collection header find is worth pulling on further, because the breadcrumbs only came back halfway. On every product page I loaded the breadcrumb renders as Home followed by a divider and nothing after it, and on collection pages it’s Home / Collections / with the collection name missing. Fourteen products and five collections, all identical, so it’s the template rather than a stray page. There’s also no BreadcrumbList markup left anywhere, product, collection or homepage, so the trail Google reads is gone as well as the one shoppers read. Same shape as your H1 problem: the block is enabled so the editor shows nothing wrong, it just isn’t outputting its last segment.
Separately, and not a migration thing, about a fifth of the products I sampled have no Search engine listing set, so Shopify is auto cutting the body copy at exactly 320 characters to make the meta. Those all open with the literal words Product Details and stop mid word. The Serenya Bookcase snippet ends on “and kee”. The other 36 in my sample are hand written and good, so it’s a gap to close rather than a rewrite.
Worth saying the rest of your Google facing layer is stronger than most stores that post here, and every product image I loaded had real alt text written for it, which almost nobody does. That’s why the breadcrumb thing reads as leftover wiring rather than something you never got to.
One more migration-specific check - the inverse of Ecom_swift_LLC’s leftover 404s: apps that never made the jump. Theme files do not migrate between themes, so anything an app wired into Ella by writing into its files (the manual-snippet generation: review widgets, trust badges, WhatsApp buttons, COD fee logic) did not come along to Impulse. The app is still installed, still billing, its admin dashboard shows green - but its storefront piece silently stopped rendering in July. This is easy to miss precisely because nothing errors: the code is not broken, it is absent.
On a COD furniture store the two I would verify first, because both are conversion-load-bearing for high-ticket COD: whatever renders your COD messaging and fees around the cart, and your reviews/trust widget on product pages.
Quick audit that takes fifteen minutes: keep the old Ella theme unpublished (do not delete it), open the same product page in Ella’s preview and on live Impulse side by side, and list everything that renders in Ella but not in Impulse. Then re-add each missing piece through the app’s current app block or embed in the Impulse editor - Impulse supports app blocks, so most modern apps can be re-placed in a minute - rather than pasting the old snippet across. The old Ella copy doubles as your record of which vendors ever touched the storefront, which you will want the day you clean it up properly.
Others have given you a lot of good advice and things to check. I would add a few thing that stands out to me. And some may have no connection to the theme change.
For me, your logo takes too much space, leaving also a lot of blank space around, but the issue is it pushes all other content down. Size of the logo on checkout is fine.
I did find an address and phone but you do not have clear contact information page or about us page or contact page. Just on checkout, there is a link with a contact popup. For me that alone is enough to close your site.
The homepage is just list of products one after another, no other content, nothing to explain your business, some story that customers can connect to. Why would people buy from you when they can find the same products elsewhere?
Collection page has too big a banner/breadcrumbs pard, filters are useless, and you could have longer, more meaningful descriptions. Also is it OK to use other sections there too: icon/text for customer reassurance, FAQ…
For your questions:
*Product page: it is clean but missing several blocks and sections. For blocks, more icons/customer reassurance, when available featured review, accordions with basic shipping, return, payment info. Sections yes reviews, but also FAQ, some general image/text about how you work, and agianicons/reassurance.
*Mobile has the same issues as desktop: big logo, just product carousels, no real filters, too many images on the product page and no sticky add to cart
WEll it looks clean, not sure what store looked like before but in general it looks good, clean, nice colors and font.
You asked what to fix first after the Ella to Impulse move, with traffic still there and conversion not back. On the live homepage the H1 is still waseeh.com. Collection pages breadcrumb Home / Collections / Furniture, and the furniture title repeats “Furniture” twice. That is leftover Impulse structure, not a finished category page. If you want the collection headings/breadcrumbs and the COD/PDP trust items cleaned up, say if we should take it.
(1) Contact, about, FAQs, and reviews return HTTP 200 today, but I found zero links to them on the live homepage or PDP. The June 7 Wayback snapshot shows Ella linked all four; Impulse only wires policy pages. Rewire your menu settings to restore them: for a COD store, a reachable human is the conversion signal.
(2) Your return and shipping copy used to render in PDP tabs; live PDP has only the COD banner and ‘Shipping calculated at checkout’. The archived page shows ‘You can return your item within 7 days’ and ‘Orders are processed within 1-2 business days’. Pull that copy back onto the PDP near add-to-cart using Impulse blocks.
(3) Sticky add-to-cart existed in Ella’s mobile assets; Impulse ships none. Restore it.
(4) Breadcrumbs truncated everywhere and BreadcrumbList schema vanished. Maybe not very important, but worth trying it out. You can get a sense from Clarity sessions of June if it’s worth it
Run a before/after diff on homepage, collection, and PDP to surface what broke silently. Fix: prioritize 1 and 2, then audit conversion by device × landing-template × channel to localize the residual drop
Traffic staying steady while conversion fell immediately after the theme migration points to a storefront regression, so I wouldn’t change acquisition yet.
I checked the public Vilinda Dresser flow and found one concrete issue: the White variant maps to media labelled “Merionne Dresser,” which appears to be a different product. That can create doubt at the colour-selection step, especially during a COD journey.
I’d correct that mapping first, then test both colours through product → cart → COD handoff on mobile and compare each checkpoint with the previous Ella flow.
If useful, I can post the next five checks here in priority order.
@ideagen ,
One thing I’d check first is whether the migration changed the customer journey not just the design. For a high ticket COD store trust and clarity are especially important.
I’d compare the old and new product pages for things like delivery/return information, contact details, COD messaging, reviews, and sticky Add to Cart on mobile. Also check app embeds and theme blocks since these can easily get missed during a theme migration.
I’d fix those conversion related gaps first before making bigger design changes. The fact that traffic stayed stable but conversion dropped makes this worth investigating.
Start with the first screen on mobile, because right now the homepage is doing a lot of talking. The H1 is still just the domain name, and the hero copy is all brand language with no clear next step.
So a shopper on a slow connection has to work too hard to figure out where to go. I’d suggest one obvious button above the fold that sends people straight into the highest-intent collection, plus a short line that addresses the major questions: delivery timing and return policy.
The other thing I’d check is whether the product pages are carrying enough trust near the buy button for a first-time buyer. Again, I’d put the delivery promise, COD wording, and return or exchange terms much closer to the add-to-cart area. If the page makes someone hunt for those details, they’ll often just silently bounce (which is the reason I built Carti fwiw).
Eleven replies in and everyone has been looking at the theme, so I went at the catalogue instead. That part came through the migration untouched, which is exactly what makes it easy to stop seeing.
Your storefront publishes 6,249 products. 3,700 of them, 59%, have every variant unavailable. I pulled the public catalogue and then spot-checked twenty at random against their own product pages, and the markup agreed every time. So roughly three product pages in five cannot be bought from. On a COD store, where you are asking a first-time buyer to accept a large delivery at the door, landing on two dead pages before a live one is its own trust problem, separate from anything the theme is doing.
Second thing from the same pull. 475 products carry the product type “Not Sale FB”. Those are live and buyable, and they sit inside the collections a shopper browses by price: 405 of them in Above 500 & Less 1000, 276 in Above 3900. Only you know whether those are meant to sell on the site. If they are not, each one is a COD order you have to phone and cancel after the customer has already said yes.
That type field is carrying more than it can. There are 13 spellings across the catalogue, with kitchen item and kitchen Item sitting as two separate buckets, and that field is what your collections and filters are built out of. Bathroom Mats and Kids Beds currently hold nothing at all.
Last one, on the trust question specifically: /pages/track-your-order is a 404, and /pages/faq 404s while /pages/faqs works. Order status is the page a nervous COD buyer goes looking for straight after they commit.
@ideagen — twelve replies, and every one of them is a storefront review. A review can list what could be better; it can’t tell you which item is actually costing orders, and only your numbers can — brihati came closest with “audit conversion by device × landing template × channel.” Before the fix list, three questions that narrow the July drop to one cause:
Which step fell? June vs August, mobile only: product view → add to cart → checkout → order. A fall at product view points to trust or dead pages; at checkout, to the COD handoff.
Did the traffic mix hold, or only the volume? Losing collection H1s moves organic landing pages. Same sessions arriving on different pages is a different funnel.
Were sold-out products visible in collections before July? wei_arvio’s 3,700 unavailable items only explain a drop if Impulse started showing them and Ella didn’t. Otherwise they’re a ceiling, not the cause.
Answer even one and I’ll tell you what it rules out — cheaper than fixing twelve things.
