From 1.8× to 5.2× ROAS: The Elara Commerce Story
When Elara Commerce came to us, they were spending £40,000 per month on paid media and generating a 1.8× ROAS. They knew the performance was poor but couldn't identify the root cause — their agency had been explaining it away as 'market conditions' for two quarters. Here's exactly what we found, what we did, and what it produced.
The Diagnosis: What Was Actually Wrong
The paid media account was running broad-match campaigns with generic creative. No audience segmentation beyond basic age and gender targeting. Zero creative testing framework — the same four ads had been running for six months. All spend was allocated to new customer acquisition with nothing invested in retention.
The attribution setup was last-click, which was making Google Shopping look like the sole driver of revenue and masking the contribution of Meta's top-of-funnel activity. This had led to a gradual budget shift away from Meta — which was quietly driving awareness — towards Google, which was capturing demand but increasingly struggling to generate new demand at scale.
Email and SMS were almost entirely unused. Klaviyo was connected but only had a basic welcome series active. There was no abandoned cart flow, no post-purchase sequence, no winback campaign. The retention layer — which typically accounts for 25–35% of revenue at healthy e-commerce operations — was effectively non-existent.
Phase 1: The Paid Media Rebuild (Months 1–3)
We rebuilt the Meta and Google Ads infrastructure from scratch over four weeks. New campaign architecture using the three-tier funnel structure outlined in our PPC framework post. Dynamic product ads introduced for both retargeting and broad audience prospecting. Custom audience lookalikes built from LTV cohort data rather than all-customer data.
Creative testing framework launched: 40+ new ad variants in month one, testing hook formats, offer mechanics (free shipping vs. percentage discount vs. bundles), and social proof formats (customer reviews vs. UGC vs. brand content). The first clear winner — a UGC-style video ad featuring a genuine customer review — outperformed the previous control by 2.3× on conversion rate.
Attribution was rebuilt to a position-based model with a 30-day window, providing a much more accurate view of channel contribution. The immediate effect: Meta's value became visible, and budget was reallocated accordingly.
Phase 2: Building the Retention Engine (Months 2–4)
While the paid media rebuild was running, we simultaneously built out the full Klaviyo retention infrastructure. Welcome series (5-email, 7-day sequence with progressive brand story and social proof). Abandoned cart (3-touch: browser push at 15 minutes, email at 1 hour, SMS at 24 hours). Post-purchase (5-email LTV extension sequence triggered by product category, timing cross-sell recommendations based on co-purchase data).
The abandoned cart sequence alone recovered £34,000 in its first 60 days. The post-purchase sequence, which focuses on extending LTV rather than immediate re-purchase, showed its impact at the 90-day mark: repeat purchase rate among customers who received the sequence was 34% higher than among those who didn't.
SMS was introduced as a channel in month three — initially just for abandoned cart and flash sale notifications. The 15-minute abandoned cart SMS alone drives a 14% recovery rate, making it the highest-performing single touchpoint in the retention stack.
The Results: Month-by-Month
Month 1: ROAS moved from 1.8× to 2.4×. The structural changes to the campaign architecture and the initial creative improvements drove the lift, even before the algorithm had enough data to fully optimise.
Month 3: ROAS at 3.8×. The paid media algorithm had accumulated sufficient conversion data for Target ROAS bidding to activate effectively. The retention email flows were contributing 12% of total revenue. The lookalike audiences built from the LTV cohort data were consistently outperforming the previous broad audiences.
Month 6: ROAS at 5.2×. Year-on-year revenue up 91%. Email and SMS now accounts for 28% of total revenue. Average order value up 18% due to the upsell and cross-sell sequences. Customer acquisition cost down 43% due to improved audience quality and creative performance.
The transformation wasn't magic and it wasn't luck. Every change was data-informed, every result was measured, and every iteration was based on what the numbers were telling us. The lesson is simple: most underperforming ad accounts aren't suffering from budget problems. They're suffering from structure and strategy problems. Fix those first.
Want Personalised Growth Advice?
Stop reading — start growing. Let our team analyse your business and give you a tailored roadmap for compounding digital growth.