3,180 Pounds in Ads, 11,398 Pounds in Revenue, and a Checkout That Was Leaking

Business Overview

Ammar Fragrances is a UK fragrance ecommerce brand built from scratch. Mid range price point with a 35 pound average order value, and a product mix that includes sample sets, which matter more than they look. Sample packs are how you sell a fragrance to someone who cannot smell it first.

The brand sells nationally, with demand heavily concentrated in London.

The Challenge

Traffic was not the problem. The store attracted 9,300 users and 11,000 sessions in the month. The problem sat right at the end of the funnel. 715 people started checkout and only 296 finished.

Nearly 60 percent of buyers who had already decided to purchase were walking away at the final step. That is the most expensive kind of loss, because the traffic has already been paid for.

Objectives

• Grow revenue from paid media at a profitable return
• Reduce cost per conversion on both platforms
• Find and fix the checkout drop off
• Identify which products were actually pulling demand so budget could follow them

Initial Metrics

Initial Metrics by Channel

Audit Findings

• Checkout was the single biggest revenue opportunity in the business. 419 buyers reached checkout and left. At a 35 pound average order value that is roughly 14,600 pounds of intent lost in one month, which is more than the month’s actual revenue.
• 98 percent of traffic was new users. Only 200 of 9,300 users were returning. Fragrance is a naturally repeating category, so this was a retention gap rather than an acquisition success.
• Google Ads was doing the efficient work at 8 pounds per conversion. Facebook was roughly three times less efficient per conversion, but it had generated 2,400 add to carts and 1,200 initiated checkouts. That audience pool was sitting unused.
• Demand was concentrated in a small number of products. White Crystal alone accounted for 1,486 units of product movement, followed by the 5 Best Seller Perfume set, Blue Lotus and the 5 Perfume Samples set. Budget was spread across the catalogue rather than weighted toward the products the market had already chosen.
• Geography was concentrated. London produced 3,000 of 9,300 users, followed by Kingston at 282, Sheffield at 253, Leeds at 226, Birmingham at 203 and Nottingham at 191. No geographic bid adjustments were in place.
• Traffic was 87 percent mobile across iOS at 3,800 users and Android at 3,000, with 1,000 on Windows. The checkout friction was worst on exactly the devices carrying most of the traffic.
• Average engagement time of 1 minute 18 seconds on a considered purchase suggested the product pages were not doing enough selling before asking for the order.

Strategy

Stop buying more traffic until the traffic already arriving converts. Fix checkout first, because that is the cheapest revenue in the business. Then use the add to cart audience Facebook had already built. Then weight budget toward the products and the cities that had already proven demand.

Buying more clicks into a leaking checkout is the most common and most expensive mistake in ecommerce media buying.

Technical Improvements

• Checkout simplified for mobile first, with guest checkout enabled and form fields reduced to the minimum
• Delivery cost and delivery timing surfaced before checkout rather than inside it, since unexpected shipping cost is the most common reason a committed buyer abandons
• Payment options widened so the final step was not the reason to leave
• Event tracking verified across both platforms so add to cart and checkout initiated could be trusted as remarketing audiences

Campaign Structure

• Google Ads split by intent into brand, category, product name and competitor, so each could be bid separately
• Shopping given its own budget with White Crystal and the sample sets prioritised, because the data had already identified them as the demand drivers
• Facebook moved from broad conversion campaigns to a two tier structure of broad prospecting above dedicated remarketing

Audience Strategy

Abandoned checkout became the first remarketing tier with the highest bid, since 1,200 people had already initiated checkout. Add to cart formed the second tier at 2,400 users, and page viewers the third.

Sample sets were used as the entry offer for cold traffic. In fragrance the real objection is that the customer cannot smell the product, and a sample pack removes that objection at a low price point while creating a buyer who can be sold a full bottle later.

Bidding Strategy

Geographic bid uplifts applied to London and the five secondary cities that were already producing traffic, rather than bidding evenly across the country. Mobile bid adjustments corrected once the mobile checkout had actually been fixed, in that order, because raising mobile bids into a broken checkout would only have increased the loss.

Creative Changes

Ads led with the proven best sellers rather than presenting the whole range. Sample pack creative was used for prospecting and full bottle creative for remarketing, which matches the offer to where the customer is in the decision.

Optimization Process

Weekly search term work on Google. Monthly product level review so budget followed the products pulling demand. Continuous monitoring of the checkout funnel, because checkout completion rate had become the most valuable number in the account.

Results

Key Learnings

Final Business Impact

The account stopped being a traffic buying exercise and became a system where paid media, checkout and retention each carried their share of the work. The brand went from launch to a 4x return on ad spend, and the largest remaining growth lever had been identified and quantified rather than guessed at.