The Page Experience Update Lands: Thirteen Months of Notice, Now Due

June 15 arrived, and with it, on schedule as always, the page experience update. Core Web Vitals (the LCP, FID, and CLS we translated when they were named thirteen months ago) began counting toward Google’s rankings. The rollout runs gradually through August per the announced plan.

Arriving in the same package, and deserving its own celebration in certain publisher offices: the end of AMP’s exclusive grip on the Top Stories carousel. The mobile carousel that since 2016 required the AMP format as its price of admission now admits any page that meets the performance bar. Performance on merit, format optional. Publishers who resented the format tax will rightly read it as the argument conceded. The goal was always speed. The format was always a means. The means is now, officially, yours to choose.

Two enforcement events, thirteen months of notice, zero surprises. The familiar pattern, running again. The work here: set expectations against the panic cycle. Map where the fixes actually concentrate. And restate the pattern that has earned its ‘once more with feeling.’

Expectation-setting, against the panic cycle

The calibration, from Google’s own statements and the early rollout data, put plainly enough to survive forwarding to an anxious stakeholder. This is a tiebreaker-weight signal phasing in gradually, not a cliff. Relevance and content quality continue to dominate ranking. A fast page about the wrong thing ranks for nothing. The vitals operate at the margins, mattering most where competing results are otherwise comparable. The slow June-through-August rollout means no site gets a switch flipped on it.

The sites that spent the runway should expect modest, favorable, hard-to-isolate effects, which is what tiebreakers deliver. The sites that ignored the runway are not falling off a cliff. They are giving up a margin, compounding, in every close matchup. That is both less dramatic and more expensive than the panic version.

But the calibration comes with its counterweight, because ‘modest ranking weight’ understates what actually changed this month. Being visible matters apart from the weight. Search Console’s vitals reports now connect, in every executive’s mental model, to rankings and revenue. The scores are public. Any competitor or board member can run them. The field-data literacy we have been teaching is about to matter in rooms where it was optional. CrUX pass rates by template over 28-day windows, not Lighthouse lab scores in a meeting. The signal’s weight is a tiebreaker. The scoreboard’s weight is organizational, and the scoreboard is the durable change.

Where the work concentrates, from the remediation queue

• LCP: images and server response, in that order. The February format work paying off exactly as scheduled. The modern-format rollouts, dimension attributes, and CDN negotiation move LCP by the hundreds of milliseconds that flip failing templates to passing. Server-response tuning (the caching-ladder review, the hosting-tier honesty) is the second lever where images alone don’t clear the bar.

• CLS: space reservation, which is design governance. The layout-shift failures in our queue are overwhelmingly the un-reserved. The ad slot that arrives late and shoves the article. The injected banner. The un-dimensioned embed. The fix is reserving the space before the content arrives, dimensioned containers, aspect-ratio boxes, the no-injection-above-content rule. It lives in templates and ad-ops policy. This is the metric that embarrasses layouts, and the remediation is a design-and-revenue conversation with a metric attached.

• FID: mostly already passing, watch the successor. Field FID passes broadly, given the metric’s known generosity. The script-weight work remains right on its own merits. The measurement community is visibly building a fuller responsiveness metric, so treat this slot of the scoreboard as provisional.

• The AMP decision, now actually a decision: publishers running AMP purely for the carousel can now weigh retiring it. The math is upkeep tax against measured benefit, with the exclusivity term deleted. No reflexive exits, but the quarterly review we always attached to that bargain finally has both columns free.

The pattern, once more with feeling

The eleven-year observation, restated at its latest confirmation, because each cohort of readers meets it at a different enforcement date. Google keeps its announced deadlines, and the entire strategic content of that fact is available to anyone who chooses to believe it.

The sequence this month completes ran exactly as its predecessors. Criteria published with names and thresholds, May 2020. Tooling shipped at the same time. Date announced with generous runway. Date kept, with a brief operational delay from April to June, announced, itself a form of keeping. Enforcement phased gently.

Mobile-friendly ran this same course. HTTPS ran it. Interstitials, the Speed Update, and now the vitals. Five for five across seven years, and the approach it rewards has not varied once. Treat announcements as project calendars. Spend the runway on the published criteria. Experience enforcement day as a reading day. The sites that did so are spending this month watching dashboards confirm work already done, which is the anticlimax we keep holding up as the entire point. Anticlimaxes are the goal. Drama on an enforcement date means the runway was spent elsewhere, and the runway is always, always announced.

The argument since 2010

The performance argument we opened in 2010, speed as strategy, argued to rooms that heard it as engineering’s hobby, has now completed its full institutional passage. Practitioner case, then measurable incentive, then named vocabulary. Then public scoreboard, and now enforcement.

What remains, and what we will cover in the years ahead, is the part no platform can enforce. The difference between passing the thresholds and actually being fast. The thresholds are floors, not finish lines, and the median institutional page has real room above them. The successor metrics already visible in the tooling. And the standing equity argument, the visitors on median hardware, which was always the better reason than the ranking ever was.

Pass the bar this summer. The runway was thirteen months and the work is February’s checklist. Then keep going, because the users the thresholds approximate were never approximations, and the sites we have always admired are fast for them, not for the tiebreaker. The scoreboard finally agrees with the sermon. The sermon, as ever, was aimed higher.