The Mobile Majority Arrived. Averages Are Now Lying to You.

The milestone headlines arrived this month on schedule. ComScore’s tracking now shows Americans spending more of their internet time on mobile devices than on PCs. The industry has been promising this crossover in conference decks since the first iPhone summer. Now it is a measured fact with a respected logo attached, which means it will appear in your leadership’s inbox attached to a question shaped like a mandate. What are we doing about mobile?

This post is about answering that question well, and the answer begins with a warning we have issued in other contexts: the industry’s average is not your strategy, and even your own average is lying to you.

The number that should drive your roadmap is not comScore’s, and it is not the single mobile-share figure on your dashboard’s first screen. It is the split inside your own analytics, section by section and task by task. That is where the decisions live, and the averages conceal exactly the decisions that matter.

Why the site-wide average misleads

A pattern typical enough that most readers will recognize their own site in it. A university’s site-wide mobile share reads 38 percent. Respectable, and rising.

Segmented, that average dissolves into a landscape. The campus-visit page runs near 70 percent mobile, because prospective families check logistics from cars and kitchens. Admissions deadlines and financial-aid basics run high fifties. The faculty research database runs 15 percent, because its users are at desks doing desk work. The events calendar spikes mobile on weekends and evenings, per the second-screen rhythm our London analysis documented.

One site, one average, and priorities that point in opposite directions depending on which section you are standing in. Build the roadmap from the 38 and you will over-invest where desks dominate and under-invest where thumbs do. Build it from the segments and the sequence writes itself. Visit logistics and deadline content first. The research database in due course. And the forms, below, wherever their completion data screams.

The three segmentations that produce decisions

• Mobile share by content section. The landscape view above. Landing-page groups, the reporting structure our keyword-era reform installed, crossed with device category. An afternoon in any analytics account, and the output is a ranked mobile-investment list grounded in your audience’s actual behavior rather than the year’s headline.

• Mobile share by traffic source. Email clicks now skew heavily mobile. Our responsive-email coverage told you why. Social referrals skew more so. Search splits by query type, with the local and immediate intents our Venice analysis described running phone-first. If your campaigns drive mobile-majority traffic to desktop-assuming landing pages, the mismatch is a measurable leak, and this cross-tab measures it.

• Mobile conversion against desktop, task by task. The money view. Completion rates for each key goal, split by device. Where mobile completion craters relative to desktop (the multi-field form, the PDF-gated download, the payment step) you have priced the cost of the current experience. That is the number budget meetings actually respond to. Where mobile converts at parity, take the win and look elsewhere. The gap, not the share, is the work order.

Reading the results without the reflexes

Two interpretive cautions from running this exercise across a client base.

First, low mobile share on a section is not always neglect’s evidence. Sometimes it is task truth, like the faculty database, and pouring mobile investment into desk-work content is the inverse error. The question per section is whether the share is low because the audience prefers desks or because the experience punishes phones. The conversion-gap view usually answers it. Preference shows parity at low volume. Punishment shows craters.

Second, the trend line outranks the level. Sections sitting at 30 percent mobile and doubling annually are tomorrow’s 60. The roadmap should serve the direction of travel, not just the snapshot, and client dashboards commonly show the same direction at different speeds.

The mobile majority is not an event that happened this month in comScore’s panel. It is a gradient moving through your own sections at measurable rates, and the measurement is the map.

Two worked findings, to show the method’s teeth

Composite examples from this quarter’s segmentation passes, shapes typical enough to borrow.

A health system’s site-wide mobile share read 41 percent. The find-a-doctor flow read 63, and its mobile completion rate ran at barely a third of desktop’s. Session recordings traced it to a filter interface that required precision tapping on targets sized for cursors. One template fix, prioritized by exactly this cross-tab, moved the mobile completion rate most of the way to parity within a month. The math, thousands of monthly mobile searches for physicians times the recovered completion delta, made the before-and-after slide the quarter’s easiest budget defense.

Second shape: a B2B manufacturer assumed its buyers were desk-bound, and its 22 percent mobile average seemed to confirm it. Segmented by source, the email-driven traffic ran 55 percent mobile. The nurture program’s carefully built sends were landing on spec-sheet pages that phones rendered as pinch-and-squint exercises. The responsive-email lesson from our 2012 coverage had been learned. The landing-page half had not, and the source segmentation was what exposed the seam.

Neither finding required new tooling, new spend, or a consultant. Both required looking at data already collected, one level deeper than the average. That is this post’s entire method and most of measurement’s entire value.

The meeting this prepares you for

When the milestone headline reaches your leadership, the prepared answer is neither ‘we have a responsive site’ nor a request to fund everything mobile at once. The first is an architecture, not a strategy.

It is the segmented picture. Here is where our audience is already mobile-majority and how those experiences perform. Here is the conversion gap we are fixing first, with its revenue or mission number attached. Here is the section where desktop still legitimately dominates and why we are leaving it alone this cycle.

That answer converts a mandate-shaped question into a sequenced plan, which is what the question was always looking for. It takes one analyst-afternoon to assemble from data you already collect. The industry spent seven years predicting this milestone. Spend one afternoon locating it precisely in your own house, and let the averages go back to what they were always for. Headlines, and other people’s roadmaps.