The mobile-first design framework has, since its initial articulation by Luke Wroblewski in 2011, served retail commerce as the canonical answer to the question of how to accommodate the growing mobile fraction of the visitor base. The framework’s core proposition — that the mobile experience should be designed first and the desktop experience derived from it as a progressive enhancement — was, at the time of its formulation, a reasonable response to the technical and market conditions of the period. The mobile fraction was growing but was not yet dominant; the desktop fraction was substantial but no longer the majority case; the appropriate design discipline was to ensure the mobile experience was not, as had been the prior pattern, treated as a degraded version of the desktop.
The conditions that made the framework reasonable have, in 2026, decisively changed. The mobile fraction of retail commerce traffic has, on the data I have available across more than a hundred retail sites, exceeded eighty per cent on the median catalogue; the desktop fraction has, correspondingly, fallen to a minority position from which it is unlikely to recover. The mobile-first framing is, in this environment, no longer the appropriate framing; the framework’s recommendation that the mobile experience be designed first is now, in operational terms, a recommendation that the experience that serves the great majority of the audience be the design starting point. The recommendation is correct as far as it goes; it is also, in 2026, no longer interesting, because there is no longer a meaningful alternative.
The interesting question is, instead, the mobile-only question: should the desktop experience be designed at all, or should the available design budget be allocated entirely to the mobile experience and the desktop experience be derived as a secondary consideration rather than as a parallel investment? This post is an account of the considerations that bear on the question and the conditions under which the mobile-only answer is, in my view, the correct one.
What the desktop fraction now contains
The visitors who arrive on a retail site via desktop in 2026 are, in their behavioural patterns, a different population from the visitors who arrived via desktop in 2018. The composition has changed in ways that bear on the design decision.
The first change is that the desktop visitors are, on the demographic data, substantially older than the average. The age skew of the desktop fraction has tracked the age skew of the population that retains a meaningful preference for desktop browsing for retail; the cohort is concentrated in the 45-and-over age band, with a corresponding concentration in higher average order values and longer session durations than the mobile cohort. (For a more detailed account of this demographic split, see The Demographic Behaviour project.)
The second change is that the desktop visitors are, on the conversion data, substantially higher-converting per session than the mobile visitors. The conversion rate of desktop sessions is, on the median across the sites I have measured, approximately 1.6 times the conversion rate of mobile sessions; the average order value is approximately 1.4 times. The combined effect is that desktop sessions, despite being a minority of total sessions, contribute approximately a third of total revenue on the median catalogue.
The third change is that the desktop visitors are, on the qualitative interview data, more likely to be in a “research” mode than the mobile visitors. Desktop sessions show, more frequently than mobile, the patterns of comparison shopping, of consultation of multiple browser tabs simultaneously, and of considered review-reading before purchase. The desktop session is, in some sense, the retail-research environment; the mobile session is the retail-discovery environment.
The mobile-only question, considered carefully
The mobile-only proposition, in its strongest form, is that the desktop experience should be derived directly from the mobile experience as a styled-up presentation rather than as a separately-designed layout. The proposition has the advantage of consolidating the design budget on the experience that serves the great majority of the audience; it has the disadvantage of producing a desktop experience that is, in many implementations, visibly compromised against the higher expectations of the desktop audience.
The proposition is, in my view, correct for some retail catalogues and incorrect for others, with the determining factor being the revenue contribution of the desktop fraction. For a catalogue whose desktop revenue contribution is below approximately fifteen per cent — typical of fashion brands targeting Gen Z, of beauty brands with strong social acquisition, of any catalogue whose audience composition is concentrated in the under-30 cohort — the mobile-only approach is, on the data, the correct one. The desktop fraction is small enough that the experience compromise is operationally acceptable, and the budget consolidated on mobile produces a meaningfully better mobile experience than the budget split would have produced.
For a catalogue whose desktop revenue contribution exceeds approximately thirty per cent — typical of luxury, of high-consideration categories such as automotive accessories or homewares, of any catalogue whose audience composition is older — the mobile-only approach is incorrect. The desktop fraction is large enough that the experience compromise produces measurable revenue regression, and the budget consolidated on mobile loses more than it gains. The retailer in this case should retain the parallel-design pattern; the mobile-first framework remains operationally appropriate, with the understanding that “mobile-first” now describes the dominant experience rather than the lesser one.
For catalogues whose desktop fraction sits between the two thresholds — between fifteen and thirty per cent — the question is genuinely open and the appropriate answer depends on factors specific to the catalogue. The factors include the audience’s responsiveness to desktop-specific design improvements (which can be measured through controlled experiments), the brand’s tolerance for visibly compromised desktop experiences (which is partly a brand-positioning question), and the operational capacity of the design team to maintain two distinct experiences (which is partly a budgeting question).
The implementation considerations
For retailers who arrive at the mobile-only conclusion, the implementation requires more discipline than the framework’s name suggests. The mobile-only design must, in operational terms, accommodate the desktop derivation without producing the worst-of-both-worlds outcome in which the desktop experience is the mobile experience styled up to a larger viewport without any of the affordances the desktop audience expects.
The most consequential of the affordances is the use of horizontal space. A desktop layout that simply scales the mobile layout to fill the available horizontal space produces, in the great majority of implementations, a layout that visibly fails to use the space appropriately — content stretches to readable-paragraph widths well above the comfortable maximum, the visual hierarchy collapses, the design feels visibly inflated rather than appropriately scaled. The mitigation is to introduce explicit desktop-only layout rules at the top of the responsive scale, with the rules constrained to layout rather than to the design’s substantive content; the substantive content remains the mobile content, while the desktop layout uses the available horizontal space to present the mobile content with appropriate desktop affordances.
The second consideration is the desktop-specific interaction patterns that have no mobile equivalent: hover states, keyboard shortcuts, drag-and-drop interactions. The mobile-only approach, taken literally, would omit these interactions entirely; the more pragmatic approach retains them in the desktop derivation while ensuring that the mobile experience does not depend on them. The pragmatic approach requires some additional design and engineering work; the work is, on the conversion data, justified for catalogues whose desktop fraction exceeds the mobile-only threshold.
An advisory close
The mobile-first framework served retail commerce well through the period of its formulation, and the framework’s residual recommendations remain operationally sound. The framework’s framing — that the mobile experience should be designed first — is, in 2026, no longer the interesting question; the interesting question is whether the desktop experience should be designed at all as a parallel investment, or should be derived from the mobile experience as a secondary consideration. The answer to the question depends, decisively, on the desktop fraction’s revenue contribution to the catalogue; the framework’s framing does not, in itself, settle the answer either way.
It is recommended that retailers reviewing the design budget allocation between mobile and desktop do so with explicit reference to the desktop fraction’s revenue contribution rather than to the framework’s general recommendations. The two are, in 2026, frequently divergent; the divergence is decisive for whether the budget allocation produces the conversion outcomes the audience composition would predict.