UK home-fragrance brand
Case study · Meta Ads + e-commerce · UK
Client confidential · Full case study on request
I ran the ads. Then I fixed what they were selling.
Paid and organic Meta for a handmade candle brand, and the store and the maths underneath them.
2.2×
more earned per candle than per pack of melts
78%of units sold were wax melts
3.6×ROAS an ad needed just to break even on a candle order
5product pages rewritten live
1new store designed and built
Grow online sales for a one-person candle studio in England.
Hand-poured candles, wax melts and workshops, sold through a GoDaddy store. I covered every channel:
Meta AdsInstagramFacebookOnline storeEmail
02 · The ads
I ran Meta across the whole funnel.
FindProspectingNew audiences, with product and making-of creative.
RemindPixel retargetingProduct viewers and abandoned carts, brought back.
SellCatalogue and shopping adsThe store's products, straight into the feed.
CaptureLead form adsInterest captured for launches, with creative tested each round.
What we were selling
Fresh
Warm
Nutty
Fruity
Floral
Herbal
Fresh
Warm
Nutty
Fruity
Floral
Herbal
The ads were sending people to a store that wasn't ready.
Thin product pages
No scent required at checkout
Bare search titles
An empty Gift sets category
Delivery and returns hidden
04 · The fix
I fixed the destination before buying more traffic.
Product pages that answer the shopper
Before
A scented candlePrice · Add to Cart
After
A scented candle
Natural soy wax, glass tumbler, 30–40 hour burn, made to order.
Scent and mood · How to use · Options · Gifting
Search results that earn the click
Before
store.co.uk › wax-melt-gift-boxesRound wax meltsNo description written.
After
store.co.uk › wax-melts › round-wax-meltsRound Soy Wax Melts, Pack of 6 | Brand nameA pack of 6 hand-poured soy wax melts in your choice of scent. Flame-free home fragrance, made in the UK.
An email list worth sending to
Before
A list padded with duplicates and silent bounces, and unclear who had opted in.
After
Bad contacts removed and the sign-up box rewritten to say what people get.
7catalogue fixes, from gift sets to cross-sells
Scentmade a required choice before Add to Cart
GA4set up for the store in GBP
18live changes logged with dates
05 · The insight
Melts drove the volume. Candles carried the margin.
Units
Revenue
Wax melts
Candles
A candle earns about 2.2× more per unit than a pack of melts.
06 · The maths ads depend on
So I worked out what an ad can afford to pay for a sale.
Postage and packing are paid once per order, so bigger baskets leave the ad more room.
BasketBreak-even ROASRoom for the ad vs melts only
Melts only3.8×1×
One candle3.6×1.8×
Candle + 2 packs of melts3.6×2.9×
What a cold ad had to beat
UK Meta benchmarks: cost per 1,000 views ÷ click ratethe cost of one click
× 2% of visitors buyingthe cost to win one order
Against what a candle order can pay its adabout 9× too much
Click cost needed to break evenabout a ninth of the benchmark
Cold ads: ~0.4× return vs 3.6× needed
At UK benchmark prices, cold ads cost more than a single-candle order can pay. So the money had to go where it can work: people who already visited, bigger baskets and repeat buyers. That's why I paused spend and rebuilt the basket first.
Every ad sale had to clear 3.5–3.8× ROASBelow that, an order doesn't pay for its ad. Bundles give the ad 2.9× more room to win the sale.
My call: build bigger baskets, then scale spendConfirm real costs with the owner first, then advertise bundles, not single packs.
07 · The rebuild
Then I designed and built her a new store, made to grow the basket.
Free UK delivery over a basket thresholdNudges bigger baskets
Shop by scent familyFresh, fruity, warm, herbal, floral
Scent notes and reviews on every cardAnswers before the click
Workshops and the maker's storyTrust a handmade brand needs
27 client changes, one sign-off packMapped from a review call and handed over for client sign-off
What this says about how I work
I fix the destination firstbefore buying more traffic.
I let sales data pick the heroand the basket to sell.
I spend when I know the ceilingprofit per order sets the limit.