// Commented out mbk Feb 2019 // //
Search this site:
Enterprise Search Blog
« NIE Newsletter

The Looming Mobile Phone Singularity and Search: Are You Ready?

Last Updated Feb 2009

By Mark Bennett, New Idea Engineering, Inc. - Volume 5 Number 5 - November 2008

Introduction

Nobody who knows us would ever accuse us of being hard core Apple fans; some might incorrectly accuse us of being quite the opposite. But we have to admit that the iPhone and Google's Android mobile platform are really pushing the capabilities of what handsets can do, and have certainly renewed interest in these shiny handsets. Although the buzz has died down a bit, looking at the numbers it's quite clear that mobile will become the dominant platform for casual information access, and this will significantly impact the search industry as a consequence. Think this doesn't apply to your business? Read on!

This isn't just a rerun of the mobile hype from a few years back. The http://www.itu.int ITU (International Telecommunications Union) puts the eclipsing of worldwide desktops by cell phones as having taken place sometime in the late 1990s, but back then the reality was tiny black and white screens, and the demos we saw of mobile apps were often just plain silly. But in the years since, the number of mobile phones has quadrupled, while the number of desktop PCs just crossed the 1 billion mark only a few months ago. These new mobile web browsers are pretty darn good and people are actually using them to surf the web. Very soon mobile devices will surpass desktops for users accessing online resources in many specific industries. Though this trend will vary widely by geographic region and market segment, the mobile screen will become the primary target.

Businesses involved in eCommerce and advertising already know this, but we think IT departments at most other business are assuming this will pass them by; they figure that since mobile search is for shopping sites and mapping applications, and since their business is just making tractor parts for example, then none of this applies. Or they will understand that some executives want to check their email on their BlackBerry, but they assume corporate mobile usage stops there - no employee would seriously want to browse and search from their phone, would they? If you ask it in that way, people might say "no", but remember that search is being embedded in practically every business application these days, so if employees are accessing 2 or 3 corporate apps from their phones, then by definition they will likely be using mobile search.

We also wonder if this disconnect is a generational factor. Consider the usage gap in usage of social networking sites and cell phone text messaging. Even in corporate search, we've noticed a difference in behavior between older workers and the under 30 crowd. Younger workers expect their employer's search engine to work really well or else they will simply abandon it, whereas older workers may be either satisfied with mediocre search, or use other means to find information. And think about social networking usage. We believe these usage trends affect perception, and in turn, decision making. If a manager doesn't think Web 2.0 trends and messaging are very important, then of course he's not going to think mobile search is a big deal either!

Here are some of the "PC in your pocket" platforms that can handle real mobile search:

  • Apple iPhone: which attracts a unique and juicy demographic
  • Blackberry: a market leader and favorite of CEOs and VPs everywhere.
  • Google's Android mobile platform
  • Symbian OS: which is going open source
  • Microsoft Windows Mobile: on the verge of greatness for nearly a decade
  • Palm OS: a nice stable platform that some of us still love, but is sadly fading.
  • A bunch of other generic phones with screens that can display rudimentary web pages and occasionally run Java

Is your company (and favorite search vendor) ready for mobile search? Even if it's not on your radar, we suspect it will be soon, and perhaps in the form of a very nonlinear "mandate" from upper management. We think this will be sudden and disruptive in some IT departments, possibly like a compressed version of the mid to late 1990s when folks scrambled to get web sites up.

 

Sidebar: Some justification for the Apple Hype

Just as Google caused a new wave of excitement about search in the past few years, now Apple has caused a similar buzz in the high-end cell phone market, which will benefit other players as well.
  • First, as widely reported, the 3G iPhone sales, and more importantly App Store sales, are through the roof.
  • Second, companies that did an iPhone port of their app are reporting doubling their installed based, entirely due to iPhone app sales, over all other platforms in their company's history. Keep in mind this is in little more than a month. Apparently iPhone users are much more willing to try something new.
  • Third, iPhone will soon be carried by some big brick and mortar retailers.
  • Fourth, rumors of companies moth-balling their entire fleet of Blackberries in favor of iPhones, in one fell swoop.
  • Rumors that VP and CEO level executives are overriding IT policies, to mandate iPhone support, regardless of technical challenges and standing policies. Of course iPhone does offer enterprise capable features now, so this is now more feasible.
Look for these trends to quickly wash over into mainstream phones. We also expect to eventually see much less expensive Android handsets.
 

Sidebar: Every Apple has some Seeds

Some companies and individuals are frustrated with Apple because of all the technical and contractual limits on this platform. There are rationales for all of these, but we doubt any other company could get away with this many restrictions.
  • From a developer standpoint, to create a compiled application, you'll be using only Objective C. You are not allowed to even try using Java. .net, C++, C#, Python, SilverLight, nor even Flash.
  • Apple places many other restrictions on the type of apps you can code, including restrictions on web browsers, media players, VOIP, etc.
  • Public applications are distributed through iTunes and are therefore subject to Apple's unilateral and opaque approval process. Other distribution mediums are available for private corporate apps.
  • No background applications, although Apple will release a system of interrupts to summon applications to the foreground in a controlled way.
  • Apple has finally rolled back some of their extremely restrictive non-disclosure clause, so that now developers can at least talk to each other about basic coding issues.
  • The platform also lacks some features that other mobile platforms have had for a long time, and was slow to implement others. Users are still waiting for Copy and Paste and stereo Bluetooth support.
  • To use the Apple supported tools, you will need to buy an Intel powered Mac. It's annoying enough that they don't support Windows, but they also don't support their own older Power PC based Macs. Despite all of Apple's promises about compatibility and "fat binaries" back when they switched from the Power PC platform to Intel, they have seemingly retreated and do not support iPhone development on those older Macs.
  • General complaints: Sound quality and cell coverage, though it seems the 3G has addressed at least the sound quality. Limited battery life and no way to swap batteries; no support for swappable memory cards.
Despite these restrictions, it is a very nice mobile platform, and worthy of support. And remember, you don't necessarily have to do a special compiled application just for the iPhone: it's possible to use web server-side technology to support the iPhone and many other mobile devices; you won't get all the bells and whistles, but it might be good enough.

Call to Action:

Assuming you believe us, what should you do?

  • Start the paperwork for a Mac and an iPhone. The good news is that in many companies the developers already have both of these in hand, often at least partially paid for out of their own salaries, which does say something for Apple's overall design. Smart people really to like it, and smart people are supposedly... well... smart! If you can live with non-Native iPhone apps (read on), you might skip this step. Also most iPhone applications will also run on the iPod Touch, aka "the iTouch", which is the media player part of the iPhone, big screen and Wi-Fi access, but just without the phone portion; this would let you test apps on your local Wi-Fi network without needing an AT&T service contract, etc. There are some compromises here... we won't go into, but it's another option.  
  • You might also want to get a BlackBerry and an Android phone while you're at it.  
  • Second, and this one won't cost you anything, download Safari for Windows. If you're not an Apple fan, and you haven't downloaded iTunes in the past year, you may not know that Safari is the default web browser on the Mac, analogous to Microsoft Internet Explorer on Windows. Well, they've ported it to Windows, and it also happens to be the base of the web browser on the iPhone. So, if you want to get some of idea of how a web application will look on the iPhone, you can try it on Windows! This isn't an exact test, you've got a bigger screen, a different OS, a different CPU and graphics chips, etc, but it's at least a start. Certainly if a web app doesn't work on Windows' Safari, it's sure as heck not gonna work on the iPhone. Oh, and "one more thing", you know how you've had to support and certify both IE and Firefox in your web apps in the past few years? Well, congratulations, you've now inherited a third web browser platform for your regression tests... not to worry, we hear triplets are three times the fun!  
  • You'll also want to download iTunes and setup an iTunes account, since 3rd party compiled applications are only available through the iTunes store. You can download and install the iTunes and Safari bundle for Windows, which will probably be the default download you'll find. Then you can download and play with some other folks' applications to get ideas, many of them are free. To Apple's credit, though they do take a cut of paid software, they also allow apps to be offered on iTunes for free, for which they make no direct profit, and we laud them for that. The good news is your enterprise and customer facing-search engines may already minimally be iPhone compatible. If web based search is your only responsibility, you may not need to break out the Objective C compiler.
  • Choose a strategy for supporting the iPhone; see the other sidebar, and possibly the other popular platforms.  
  • Our final suggestion is to start running reports on your public and enterprise web applications and which platform folks are using to access it. Even if you make no immediate code changes, this will help you watch the trends. Heck, if the main predictions of this post turn out to be bunk, you'll have the concrete evidence in six months to publicly shame us!

Choices for Specifically Supporting the iPhone

The way we see it, you have at least 5 basic choices for supporting Search on the iPhone:

  1. Do nothing, or only minor tweaking, depending on how your testing goes. We give Apple huge credit for having the best mobile web browser out there at this time. Many web sites designed for desktop IE and Firefox users will run on the iPhone. Heck, even lots of that wacky CSS and JavaScript nested-menu-nonsense is likely to work. As we said, iPhone is based on Apple's Safari web browser, which historically has done very well in standards compliance. And it supports Ajax - that's right - Ajax apps should work on the iPhone - we give the Apple engineers real props for that, and is something my Palm OS Treo simply can't handle.
  2. Do a somewhat more detailed tweaking of your search app to support smaller screens in general. Actually, with the iPhone's unique zoom-in-zoom-out features, making a small-screen version of your app will do more to help the Treo, Blackberry and Windows Mobile users, who don't have this type of "zoom" feature. But even iPhone users would appreciate being able to use your web app without constantly zooming in and out. In particular, put those 20 or 30 ancillary navigation links and menus at the bottom of the screen, after the primary content. Ease up a bit on the graphics, and make sure the thing still looks OK if you disable CSS. And to support the most users possible, don't require Java Script, Java or Flash. This type of UI will also please public web spiders, blind folks using screen readers, Lynx users and automobile-based dashboard web browsers. You can check for the web browser being used and automatically redirect mobile users to this UI, and also give it a machine name prefix of "m.".
  3. A third option, this time for a snazzy iPhone-specific version of the app, is to use Ajax. As we said, the iPhone is revolutionary in the mobile space for supporting this. You've probably used Ajax before, like on NetFlix when you add a DVD to your queue and get a popup window of other titles, or on Amazon, when you use their Wishlist and book preview features. And of course Google's GMail and Google Docs applications are hugely successful Ajax apps. In fact, a majority of iPhone apps are actually Ajax based, and not compiled Objective C apps - one good hint is their release date, if an app was released prior to August 2008, it is most likely Ajax based. Remember, you can do a bulk of your iPhone Ajax development on a Windows PC and do spot checks in the Windows version of Safari. If you can live with Ajax level functionality, you might not need to pony up for a Mac. I do wish the web in general would do a better job of clearly distinguishing and filtering on Ajax vs. Objective C based iPhone apps.
  4. The forth option, the "gold standard", is to write a native Objective C based application for the iPhone. This will allow the application to run locally and very quickly (aside from normal transactional network latency issues). The app will have access to local storage and superior interface options, including GPS location awareness which could be very handy for business apps and search, and even access the tilt and shake sensor. Native development is the option that will normally require a Mac. Our resident Unix Guru Nick says the open source community has put together an iPhone tool chain for Windows users, but we think if you're serious, you're going to want a Mac, and you'll want to use the Objective C IDE / developer GUI. Also, a reminder, an iPod Touch can stand in for an iPhone, if you don't need the telephone features and AT&T contract, and will be content with Wi-Fi-only access.
  5. Bug your search vendor to support it natively! We think all traditional search vendors should start working on an optimized iPhone-specific search client.
    Ultimately, the iPhone hardware is the real star here, despite Apple's attempts to micromanage and stifle developers. Having a generous color screen with good connectivity and a nimble UI makes up for a lot of sins. Your native search apps could leverage the real-time location awareness feature in a big way. And AT&T's "Visual Voice Mail", where all of your voice mails are displayed on the handset screen, including caller info, is very cool for busy people who deal with tons of voice mail. If you haven't seen that feature you really should check it out, it alone might make the iPhone more attractive to IT folks.

What about Other Mobile Phones?

Some corporations have already embraced the BlackBerry into their enterprise by setting up the BlackBerry Enterprise Server (BES) and the secure network infrastructure. Although the focus has been email, BlackBerries can also access web apps. Although it's possible to write native software for it, we think this might be overkill in many cases, and a well behaved web based app might do just fine.

We'd also suggest doing an employee survey to see what employees have in their pocket. Not every phone requires a specifically compiled application, using a well designed web app will do. But you'll still want to test that on the various screens and keyboards.

In Closing

The new phones are good enough that an unmodified web app may run OK. But as soon as you add Flash, Silver Light or other plugins to the mix, all bets are off. At a minimum, do some surveys or logging to find out what your users have in their pocket, and maybe validate your key apps on at least the top three. And do some reading... be ready for a sudden corporate mandate... it may never come, but better to be prepared.

What do you think? Do you believe in the impending Mobile Singularity, the convergence of all users and critical apps to portable platforms, surpassing even desktop deployments? Do you think your VP or CEO could pull rank and toss out all of your previous policies, mandating iPhone support by early 2009? Let us know!

References

Notes on market penetration of devices and number of users, usually quoted in billions :-)
ITU: International Telecommunications Union
http://www.itu.int
Mobile handsets surpassed PCs prior to 1999.
They crossed the 1 billion mark sometime in 2001/2002.
Crossed 2 billion in 2005
Almost 4 billion in 2008
Est 5 to 6 billion by 2013
PCs hit the 1 billion mark in 2008
Google I/O presentation said it happened prior to 2000 based on ITU
http://sites.google.com/site/io/an-introduction-to-android
http://sites.google.com/site/io/an-introduction-to-android/Google-IO-IntroductionToAndroid.pdf
(HTML page will prompt to download, see slide # 4)
Mobile had already well passed Internet even in 1999
http://www.itu.int/aboutitu/annual_report/1999/state.html
http://www.itu.int/aboutitu/annual_report/1999/graphs/state1.jpg
Cell phone users surpassed desktop internet users in the 1990s
http://www.itu.int/ITU-D/ict/publications/mobileinternet/sum.html
1 billion mobile vs. fixed phone lines
1 B cells in '01/'02, 2 B est 2005
1 B fixed in 2000/2001
21 years vs. 125 years
http://www.itu.int/osg/spu/newslog/lifestylesdigital.aspx
April '06 "Will Surpass soon?", mobile = laptop?
http://www.ipsos-na.com/news/pressrelease.cfm?id=3049
"By the end of 2006, there was a total of nearly 4 billion mobile and fixed line phone subscribers, plus over 1 billion Internet users worldwide."
http://www.itu.int/itunews/manager/display.asp?lang=en&year=2007&issue=07&ipage=Telecom-trends&ext=html
Aug 2007 mobile handsets pass 3 billion
ITU World Telecom / ICT Indicators (Scroll down to 11/27/07 post)
3.9B in 2008, 5.6B by 2013
http://www.itu.int/ITU-D/ict/newslog/Worldwide+Mobile+Subscriptions+To+Reach+56+Billion+By+2013.aspx
Well over half can display web pages of some sort
http://www.3g.co.uk/PR/Jan2005/8963.htm
300 M broadband in 2005
http://www.itu.int/osg/spu/newslog/digitallife+Going+Digital.aspx
4 B fixed phone lines in Sept '07
http://www.engadgetmobile.com/2007/09/05/planet-earth-now-home-to-four-billion-phone-lines/
June 2008 PCs hit 1 billion
http://www.reuters.com/article/technologyNews/idUSL2324525420080623?feedType=RSS&feedName=technologyNews
http://www.engadget.com/2008/06/23/worldwide-pcs-in-use-surpass-1-billion-next-billion-to-come-in/
ITU Database approx $200 US
http://www.itu.int/ITU-D/ict/publications/world/world.html