The iPad Question: A Third Screen Between Phone and Desktop

Steve Jobs sat in an armchair on Wednesday and browsed the web on a 9.7-inch sheet of glass. The room’s instant verdict, repeated in half the coverage since: it’s just a big iPhone.

Fine as a hardware take. Wrong as a question for anyone who runs a website. ‘Big iPhone’ undersells precisely the part that lands on our desks. The iPad, shipping in April from 499 dollars, runs desktop-class Safari on a screen size the web has never seriously designed for, held in postures the web has never seriously considered. And, once again and now unmistakably a policy, no Flash.

Our January 2007 post on the iPhone announcement advised watchful patience and cost-free preparation. Three years later, with that playbook’s results in hand, here is the iPad edition.

What’s actually new here

Not the browser. It is mobile Safari, WebKit, well understood from three years of iPhone work. What is new is the context. A 1024-by-768 touch canvas is large enough that sites will render their full desktop layouts. No mobile detection will intercept it, and none should. For the first time, desktop-designed pages will be routinely operated by fingers.

Every hover-dependent menu, every cursor-sized link, every interaction assumption our 2007 field notes flagged on a 3.5-inch screen now applies to layouts that were never on notice.

The posture matters more than it sounds, too. Apple’s entire pitch staged the device in laps and armchairs. A reading-first, leaning-back mode of use that favors long-form content, generous typography, and pages that respect a comfortable measure. The craft we have argued for on typographic grounds is about to get a device category optimized for enjoying it.

The Flash absence completes a trajectory into a policy. One device without the plugin was a quirk. A product line spanning phone to near-laptop, all Flash-free, all premium-audience, is a strategic fact. Sites still running Flash navigation or media now face a growing, influential audience that experiences them as broken. Our component-not-foundation advice graduates from suggestion to urgency. We suspect this argument is about to get considerably louder from Cupertino itself. The signals are not subtle.

The tidy worldview this scrambles

Here is the strategic wrinkle worth sitting with. The industry had just settled into a serviceable two-bucket model. Desktop sites for desks, mobile considerations for phones, with device detection sorting visitors between them.

The iPad fits neither bucket. It will receive desktop layouts and deserve touch consideration. It carries mobile Safari and affords long sessions. The honest reading is that the bucket model itself is the casualty.

Screen diversity is compounding, not resolving. Phones at several sizes. This new middle category, and competitors are inevitable if the sales materialize. Laptops, and the large monitors at the far end. Building a dedicated experience per bucket was already a maintenance treadmill at two buckets. At four and climbing, it is a losing architecture.

What is needed is an approach where layout adapts to the canvas it finds, rather than sorting visitors into fixed experiences. We will be candid that the industry, ourselves included, does not yet have a settled method for that. The pieces exist: fluid layouts, the media-query machinery in modern CSS. The coherent practice does not, quite. We say ‘watch this space’ often. This time we mean it literally. The space between 320 and 1024 pixels is where the next few years of front-end thinking will happen.

What to do before April, sorted by cost

• Free: the standing preparation, which three years of practice has made routine. Standards-based markup renders properly on WebKit by construction. Restrained page weight serves every network. And the Flash audit your site should have had since 2008 is now overdue rather than prudent. Nothing iPad-specific required. Everything iPad-relevant already on the list.

• Cheap: an interaction pass on your key templates asking the touch questions at desktop scale. Does navigation work without hover? Are click targets humane? Do forms behave? These fixes serve existing accessibility and usability audiences today. The 2007 lesson holds: every flaw touch exposes was already failing someone. And they will be tested by iPads in your analytics within a quarter.

• Premature: iPad-specific experiences, apps commissioned because the keynote was exciting, and any vendor pitch containing the phrase ‘iPad-optimized microsite.’ April will tell us adoption. Your analytics will tell you yours. The 2007 rule, let the audience arrive before building for it, has saved clients more money per sentence than any other advice we give.

The publishing wrinkle worth an editorial meeting

One second-order effect deserves its own note for the content teams in our readership. If the reading-first device thesis holds, long-form content gets a better stage than the web has ever offered it.

The industry chatter this month is full of publishers redesigning for the tablet: apps, editions, experiments. Most of that is magazine-economics anxiety most site owners can skip. But the underlying observation transfers. Audiences with comfortable reading hardware read more, longer, later in the evening.

Institutional content strategies have spent years optimizing for the skimming desktop visitor and the hurried phone glance. A couch-context device re-legitimizes the substantive piece. The annual report worth actually reading. The long explainer. The photo essay.

None of this requires tablet-specific production. It requires what we have argued on other grounds: content worth reading, set legibly, structured honestly. If the category succeeds, the organizations holding that inventory get a new prime-time slot for free.

Our bet, in writing as usual

We think this device category is real. Not because of Wednesday’s specifications, but because the couch-browsing behavior it targets already exists in laptops, awkwardly, and purpose-built ergonomics have a way of collecting existing behavior.

We would guess meaningful institutional-audience adoption within eighteen months. Donors, executives, faculty: the early-adopter profile institutional sites court. And we would guess the ‘in-between screen’ problem drags the whole industry toward adaptive-layout techniques faster than anyone currently plans. Both guesses are checkable. Both are filed.

Meanwhile the advice stands unchanged. Build for the web, not the gadget. A sentence we wrote about a phone in 2007, reuse about a tablet in 2010, and expect to reuse about something unannounced in 2013. The gadgets keep proving it for us.