This is Part 2 of a three-part series on my published research into customer experience, purchase intention and re-visit intention in e-commerce. Part 1 covered the study and its headline findings. This part digs into what the data actually tells us about designing e-commerce interfaces — including one finding that runs counter to conventional UX wisdom.
The Four Dimensions of Digital Customer Experience
Rather than measuring “customer experience” as one vague construct, we measured four distinct dimensions, each with its own items and reliability:
- Security (α = 0.871) — whether users feel their data and payment are safe.
- Interface (α = 0.854) — the design, usability and navigability of the site itself.
- Image (α = 0.823) — the brand impression and perceived standing of the retailer.
- Trust (α = 0.925) — confidence in the retailer to deliver on its promises.
This decomposition is what makes the study useful for designers. Knowing that “experience matters” is a platitude. Knowing which specific dimension moves which specific outcome, and by how much, is actionable.
Finding 1: Interface Matters Nearly Twice as Much for Coming Back as for Buying
This is the finding I think about most. The interface dimension had a modest effect on purchase intention (β = 0.128) — but a substantially stronger effect on re-visit intention (β = 0.214). Nearly 1.7 times stronger for repeat visits.
The interpretation is intuitive once you see it. A first purchase can survive a mediocre interface if the price is right, the brand is trusted, and the product is what you want. Motivation carries the user through friction. But coming back voluntarily is a different decision entirely — it’s a choice made without that specific purchase motivation driving you forward. If the interface was annoying, you remember, and you look elsewhere.
The practical implication: UI/UX investment pays off disproportionately in retention, not acquisition. If you’re building a business case for design work and framing it purely around conversion rate, you’re undercounting its value by a wide margin. The interface is what determines whether a customer becomes a repeat customer.
Finding 2: Security Drives Buying — But Has No Effect on Returning
Here’s the counterintuitive one. Security had a clear, significant effect on purchase intention (β = 0.190; p < 0.01). But on re-visit intention, its effect was not statistically significant at all (β = 0.073; p > 0.05). H2 was only partially accepted because of this.
Why would security matter for buying but not for returning? My reading is that security functions as a threshold, not a scale. Before you hand over your card details for the first time, security perception is a live, active concern — it gates the transaction. But once you’ve completed a purchase without incident, security stops being a differentiator. It has been validated, it recedes into the background, and other factors take over the decision to return.
This maps neatly onto how designers should treat security signals in an interface: they are a hygiene factor for acquisition. You need them, prominently, at the point of first transaction. But loading a returning customer’s experience with security reassurance adds nothing — they’ve already cleared that hurdle. The design effort is better spent on interface quality and brand experience, which do keep working.
Finding 3: Image and Trust Dominate Both Outcomes
Across both purchase intention and re-visit intention, the two strongest dimensions were the same: image (β = 0.381 and 0.420) and trust (β = 0.274 and 0.345). Both grew stronger when predicting re-visit intention than purchase intention.
These are also the two dimensions that correlate most tightly with each other (r = 0.787), which makes sense — brand image and trustworthiness are deeply entangled perceptions. What matters for design is that neither is purely a marketing concern. Image and trust are constructed largely through the interface: visual polish, information clarity, honest error handling, transparent pricing, delivery-promise reliability. Designers build trust whether they intend to or not.
This aligns with the wider literature we reviewed. Hausman and Siekpe found website design plays a critical role in turning online visitors into real customers. Sam and Tahir identified website design, trust, empathy, perceived risk, information quality and usability as factors positively affecting purchase intention. Shergill and Chen emphasised that website design affects the online purchasing decision directly.
What This Means for E-commerce UX Design
- Front-load security signals at first transaction, then let them recede. They gate acquisition but contribute nothing to retention.
- Treat interface quality as a retention investment. Its effect on returning is nearly double its effect on first purchase.
- Design for image and trust deliberately. These are the heaviest levers on both outcomes, and they’re built through the interface, not just the marketing.
- Remember the purchase-to-revisit link. Purchase intention explained 75% of re-visit intention variance. An experience that converts tends to retain.
These findings sit comfortably alongside the heuristic principles and usability foundations covered in Part 3 of my UI/UX Playbook — with the difference that here, we have regression coefficients rather than received wisdom.
Previous: ← Part 1 — The Study and Findings · Next: Part 3 — CX Management for E-commerce →
Based on Meriç, G. & Yıldırım, F. (2021), International Journal of Commerce and Finance, Vol. 7, Issue 1. Get in touch.