We hosted a Mobile Web Summit here at Zillow last week, sponsored by Barton Ventures, with leaders from several companies: Avvo, DogVacay, Expedia, Glassdoor, GrubHub, Nextdoor, Orbitz, RealSelf, Trover, Trulia, Yelp and Zillow. It was a fantastic event – we spent the day exchanging questions, learnings and ideas around mobile web. And, we were lucky to have Rich Barton and Bill Gurley in attendance to discuss their insights. The content was timely, valuable and worth sharing with a broader audience.

As Rich pointed out, mobile web is the fastest growing platform for most of us, but has the lowest engagement. For many of our companies, mobile web gets the majority of mobile users (as high as 70 percent) but a small fraction of engagement (as low as 5 percent). Mobile web is not as sexy as apps, from either a development or usage standpoint – but, because of this, there is an opportunity to invest heavily in our mobile web experiences, and get disproportionate returns, since so many players are focusing their efforts primarily on apps and desktop experiences. Creating a great mobile web experience has its challenges, with browser friction and slower page load times, but we believe this area deserves to be more than an afterthought.

Moving away from apps?

Given the historically low engagement rates on mobile web, why are we now seeing more of our mobile traffic on that platform? One common suggestion is app fatigue: Users don’t want to keep downloading apps (interestingly, this trend seems to be higher with women than men). We’ve all seen that in the past couple of years, it’s become harder to acquire an app user – both in terms of cost and user motivation. In focus groups, we hear people say that they just don’t want another app on their phone (this is especially true in verticals with a lot of players, like real estate and travel).

Another theory is that mobile web is finally “good enough,” so people don’t see greater value in an app. The mobile browser has come a long way, and many scenarios are fully supported by it now. As Rich speculated, “Hypertext killed Microsoft” (to the tune of Video Killed the Radio Star); will hypertext-enabled mobile web do the same thing to the app world? Bill’s counterpoint to this theory is that strong brands in verticals with loyal users (like Amazon Prime) take users out of search, directly into their app.

When to drive users to apps

Given that we will always have a segment of our visitors that want to stay on mobile web and a segment that want the immersive app experience, how do we know how and when to drive people to our apps without overbearing upsells?

Companies with a goal of continued content consumption and creation and B2B engagement tended to err on the side of more aggressive app upsells, whereas companies with the goal of one-time conversion tended to keep users in the mobile web experience. But even for sites that don’t want to take over the mobile web experience with app upsells, it is still important to push upsells gently in order to stay high in the app store rankings.

Part of the strategy around app upsells is also context. If a company does not have strong brand recognition, they need to communicate their value first before pushing the app. One company found that if they could get a user to navigate to a second page on mobile web first, that user would be much more likely to download the app than if it were promoted immediately.

Transition points are painful

One of the biggest UX challenges to solve for on mobile web is the handling of transitions. Email to mobile web, mobile web to apps, desktop at the office to iPad on the couch. Users don’t want to think about the platform, and often don’t even remember what device they last used to view our sites. Our challenge is to understand the user’s intent (e.g., searching for homes in Ballard or hotels in Tokyo) and make sure they stay in that path across devices. To do this, we need to know who a user is; logins are easy but not always available, and inferring users isn’t entirely accurate and should only be used for less deep content.

SEO

Unlike with mobile apps, SEO and SEM make a big difference in driving traffic on mobile web. Some might argue that because of this, Google loves mobile web and Apple prefers apps. When users come to our sites through SEO and SEM channels, it’s important to give them context: What is our brand? How does the page they’ve landed on tie it to the rest of the site? Sometimes this even means a different UX for SEO-based visits.

We can attract new visitors through unbranded keyword search, and we continue to get branded search visits as well, because often search engines are better at navigating our sites than we are. Google has already started tagging “mobile friendly” sites and will be factoring mobile-friendliness into its algorithm as of April 21. This makes it more important than ever to focus on mobile web experiences.

Sending traffic to mobile web

In addition to SEO, there are other effective ways of driving more traffic to mobile web. Email is a major driver, and users are increasingly reading email on their mobile devices – it’s critical that emails are responsive since we don’t know what platform they will be opened on. None of us have found the perfect graceful solution to deep linking into apps from emails when opened on iOS, so we need to ensure a great mobile web experience for users.

Generating traffic from social networks works well for brands with users who want to share – but, interestingly, just having the “share” option available can be detrimental to converting users who want to keep their brand use private. One company even saw an increase in registrations when the Facebook login option was removed.

Creating marketing partnerships is another way to send traffic to mobile web, but only if those partners have a great mobile experience themselves.

Engagement

Engagement on mobile web was a big topic at the summit. We all have different metrics for understanding engagement, including time on site, conversion, scrolling, page views, page views per user and bounce rate – but none is foolproof.

We have all tried different tactics to increase these engagement measures, such as collapsing sections of a page (this also helps identify which content users are most interested in); testing tap vs. swipe (note that swiping and other gestures are currently a U.S. paradigm, so international brands should use them sparingly); testing infinite scroll vs. pagination (mixed results, leaning towards infinite scroll); and making information easier to enter (pre-populating, auto-suggesting and inferring information to reduce typing).

The biggest impact we’ve seen on engagement, however, comes from improving latency. Slow page load times are the biggest pain point on mobile web, and thus the biggest opportunity. Measuring and tracking performance can be done using third-party tools like Google PageSpeed Insights as well as in-house tools. Creating cultural initiatives and goals to improve latency goes a long way: At Zillow, we set up a purposefully slow Wi-Fi network to demonstrate what slow network connections feel like on mobile web; another company sends a signal to themselves whenever a page is taking too long to load, and uses that information to determine which pages to focus on. Yet another has a core team to champion performance with the rest of the company.

The design teams at some of the companies at the summit wanted to boost engagement on mobile web with animations, but everyone found that animations failed to work smoothly and consistently on mobile Safari, although they mostly worked on mobile Chrome. Even on Chrome, however, they didn’t seem to move the needle much, so the consensus was that animation wasn’t worth the effort.

What about tablet?

Poor tablet web, always being forgotten. Many of us had stories about designing intentionally for mobile and desktop, but not as much for tablet. This isn’t just an oversight: We see tablet web usage flattening out, despite the growth on other platforms. This may be due to the growing screen sizes of phones like the iPhone 6+. We also find tablet usage to be more browsing focused and less transactional. Despite this, it’s still important to have an intentionally designed tablet web experience. With lower usage, it can be difficult to get a team focused on this, and some companies have had to get creative in pulling teams’ focus into this area. Even though desktop experiences fit on a tablet screen, they are usually far from optimized for a touch device.

At the end of the day, we all agreed that the Mobile Web Summit provided a valuable opportunity to discuss the challenges and strategies around creating an optimal experience for users across platforms. Mobile web is still rapidly evolving – there is no one expert in this area, and there’s much we can learn from examining our diverse approaches, successes and stumbling blocks. Mobile web is a key strategic area for any company with a web presence, and cross-company sharing is a win for everyone!