Executive Summary
When more conversions doesn’t mean more revenue
Postgoldforcash.com (PGFC) operates in the postal gold-buying market, offering customers a simple way to sell unwanted gold from home. Users complete an online form to request a secure pack, send their gold for valuation and receive a cash offer, with over 90% of returned packs accepted.
While the form drives lead generation, true revenue comes from customers returning their packs. PGFC partnered with us to improve form performance without reducing user intent.
A four-variant CRO test was launched using Varify, exploring layout changes, field removal and reassurance messaging. Although one variant increased submissions by up to 33%, the control delivered more returned packs.
The test revealed a key insight: reducing friction can increase volume but weaken commitment, making full journey measurement essential for meaningful optimisation.
The problem
Increase form completions without reducing intent
PGFC needed to increase the number of users entering its gold-selling journey by improving form performance. However, success could not be judged on submissions alone, as revenue depends on customers returning their gold packs.
Reducing friction risked attracting lower-intent users who were less likely to follow through. The challenge was to balance ease of completion with maintaining strong user commitment.
Analysis from Microsoft Clarity highlighted hesitation around form length and the phone number field, even when optional. At the same time, testing needed to remain robust, with equal traffic distribution across variants to avoid bias.
Crucially, online and offline data had to be connected, ensuring any uplift in submissions translated into real commercial impact rather than misleading gains.
The solution
Test, measure and follow the full customer journey
We ran a controlled multivariate test using Varify, splitting traffic evenly across four form variants. Each version tested a specific hypothesis around reducing friction and improving completion rates.
The control remained unchanged. A scaled variant improved layout so all fields and the CTA were visible immediately. Another removed the phone number field entirely, while a further variant retained it with explanatory copy to reassure users. All phone fields were clearly marked optional.
Performance was tracked weekly through Varify and GA4, measuring click-through rates and submissions. This was paired with detailed pack return data, including return rates and value.
The test ran for two months to allow offline data to mature and avoid seasonal bias, ensuring results reflected full user behaviour, not just initial actions.
The results
Submission gains revealed a hidden trade-off
The test showed clear uplifts in form submissions across several variants. Removing the phone number field performed best, delivering increases of up to 33% compared to the control. The scaled form also showed smaller gains, while the phone explanation variant consistently underperformed.
However, pack return data told a different story. As results matured, the control generated the highest volume of returned packs, despite lower submission rates.
This revealed a key insight. Reducing friction increased volume but weakened user commitment, leading to fewer completed journeys. The presence of the phone field appeared to act as a subtle qualifier, attracting higher-intent users.
By measuring both online and offline performance, the test identified the approach that delivered the strongest commercial impact, not just the highest conversion rate.
Up to +33%
Form submission uplift
4 Variants
Tested simultaneously via Varify
2 Months
Rigorous monitoring period
90%+
90%+ - Cash offers accepted
