Tried But Not Bought: How a Fashion Retailer Used IDIs and Purchase Analytics to Understand Why Virtual Try-On Was Boosting Engagement But Not Conversion

Executive Snapshot

Client

Fashion Retailer, United Kingdom

Situation/Challenge

The client had integrated an augmented reality virtual try-on feature into its mobile app following vendor assurances that similar deployments elsewhere had lifted conversion rates meaningfully. Six months in, try-on usage was strong but purchase conversion among try-on users was not materially different from users who browsed without using it. The product team disagreed about whether the feature was working and needed more time, or was working on the wrong problem entirely.

Objective

Conduct structured IDIs with customers who used the try-on feature but did not purchase, paired with purchase analytics comparing conversion patterns between try-on users and non-users across product categories and price points.

Constancy Researchers Solution

Primary Research & VoC through In-Depth Interviews (IDIs) combined with Data Analytics & Business Intelligence, 34 IDIs with non-converting try-on users across two customer segments, paired with analytics examining conversion rate differences by product category, price band, and try-on usage pattern.

Impact

IDIs revealed that try-on users valued the feature for exploration and wishlist building rather than immediate purchase decision-making, a fundamentally different usage mode than the conversion tool the vendor had positioned it as. Analytics confirmed that try-on usage correlated strongly with return visit and purchase on a second session, not same-session conversion.

Client Outcome

The retailer repositioned the feature as a discovery and save tool, redesigning the post-try-on customer journey around wishlist capture and return visit activation, and achieved a 28% lift in purchase conversion attributable to return visits from try-on sessions.

The Situation / Challenge

Virtual try-on in fashion retail solves a specific problem, the inability of online shoppers to visualise how a garment or accessory will look on them before purchasing. When it works well, it should reduce the uncertainty that causes hesitation and, in turn, improve conversion.

The client’s product team had integrated the feature expecting a direct uplift in same-session purchase conversion, matching the pattern vendor benchmarks described. When that uplift did not materialise, the team split.

The risk of continuing without that understanding was straightforward: the team would either persist with a feature deployed in the wrong context, or abandon one that was actually generating deferred value in a way the existing analytics had not been designed to measure.

Key Challenges

  • No direct research into what customers were actually doing with the virtual try-on feature and what role it was playing in their decision-making process.
  • No purchase analytics checking whether try-on usage correlated with conversion on return visits rather than within the same session.
  • A product team split between two competing explanations, neither grounded in direct customer evidence.
  • Vendor benchmark data describing same-session conversion lifts that the client’s own data was not replicating.
  • Risk of either persisting with a misdeployed feature or abandoning one generating deferred value that existing analytics were not capturing.
  • Marketing leadership pressure to make a definitive decision about the feature’s future before the next app development cycle.

AR try-on features in retail do not necessarily behave the same way across different customer segments, product categories, or shopping contexts. A feature designed to close the purchase gap in one retailer’s conversion funnel can play an entirely different role in another’s, and the only way to know which is happening is to ask the customers who used it and did not buy.

Constancy Researchers Solution

Constancy Researchers ran structured IDIs with non-converting try-on users alongside purchase analytics segmented by visit session, product category, and price band, checking both what customers said they were doing with the feature and what the data confirmed they did next.

In-Depth Interviews (IDIs) with Non-Converting Try-On Users
  • Conducted 34 structured IDIs with customers who had used the virtual try-on feature during the prior six months but had not purchased during that same session, split.
  • Found that try-on users across both groups used the feature primarily for exploration and wishlist-building rather than immediate purchase decision support, describing it as a way to hold their place on items they intended to return for.
Same-Session vs. Return-Visit Purchase Analytics
  • Analysed purchase behaviour across all app sessions involving try-on interaction, separating same-session conversion from return-visit conversion in the 14 days following a try-on session.
  • Found that try-on users converted at a rate 31% higher than non-try-on users when return visits within 14 days were included, but showed no meaningful same-session conversion.
Product Category & Price Band Conversion Segmentation
  • Segmented the analytics by product category and price band, identifying whether the return-visit conversion pattern was consistent across the range or concentrated in specific areas.
  • Found the return-visit conversion pattern was strongest in mid-to-premium price band garments, confirming that the feature was performing a consideration-extension function for higher-commitment purchases rather than an.
Post-Try-On Journey Redesign
  • Identified that the current post-try-on interface offered no persistent save or wishlist pathway, meaning customers who were using the feature for deferred purchase consideration had no product-supported.
  • Recommended redesigning the post-try-on journey around wishlist capture and return visit activation, adding a save-to-wishlist prompt, a try-on item summary in the customer’s account, and a time-based.
Repositioning & Measurement Framework
  • Delivered a repositioning recommendation framing the feature as a discovery and consideration tool rather than a same-session conversion driver, shifting the success metric from immediate conversion rate.
  • Built a measurement framework tracking try-on to wishlist save rate, return visit rate from saved items, and purchase conversion at each stage of the newly designed post-try-on journey.

The engagement replaced a damaging internal disagreement with a specific, evidence-grounded explanation of what the feature was actually doing, and a redesigned customer journey built around how customers were genuinely using it.

Impact

  • IDIs confirmed try-on users were using the feature for exploration and deferred consideration, not immediate purchase
  • Return-visit analytics found try-on users converted 31% higher when 14-day return visits were included, explaining the
  • The strongest return-visit conversion pattern was concentrated in mid-to-premium price band garments.
  • The post-try-on journey was found to lack any persistent save or wishlist pathway for deferred purchase
  • Wishlist capture, item summary, and reminder notifications were added to the post-try-on journey.
  • The feature was repositioned as a discovery tool, with success metrics shifted to 14-day return-visit conversion.
  • A 28% lift in purchase conversion attributable to return visits from try-on sessions was achieved following
  • The internal disagreement between product team camps was resolved with a shared, evidence-based understanding of the

Client Outcome

Conversion Lift

A 28% lift in purchase conversion attributable to return visits from try-on sessions was achieved following the journey redesign.

Feature Understanding

IDIs and analytics jointly confirmed the try-on feature was performing a discovery and consideration role rather than a same-session conversion one.

Journey Redesign

Wishlist capture, item summaries, and reminder notifications were added to support the deferred purchase behaviour IDIs had identified.

Metric Realignment

Success measurement shifted from same-session conversion to 14-day return-visit conversion, matching the feature's actual customer usage pattern.

Internal Alignment

The product team reached a shared, evidence-based position on the feature's role, ending a disagreement argued from assumption on both sides.

Category Targeting

The strongest return-visit pattern in mid-to-premium garments directed further investment toward the category where the feature had most impact.

Analytics Upgrade

A return-visit measurement framework was built into the app's analytics, making the feature's deferred conversion value visible on an ongoing basis.

Vendor Benchmark Context

The mismatch between vendor benchmarks and the client's data was explained, allowing the team to use industry comparisons more critically.

Market Positioning

The retailer was repositioned as a merchant that validates feature performance against its own customer behaviour rather than vendor benchmarks.

Case Studies

Learn how our success stories power data-driven growth across industries

Speak with an Analyst

    Download TOC