Fullhouse uses web and mobile analysis and recommendations in our strategy and planning and testing. We research and update this information on a quarterly basis. In conducting our research we found that no single source had all the information we were looking to gather and that non of the sources were completly accurate as they each just track a fraction of a percent of all internet traffic. So, we've compiled information from a number of reputable sources and have averaged them for each area we are reporting out on. Following you will find recommendations for:
1.0 : Desktop Browsers
Core List of Recommended Desktop Web Browsers
United States: Internet Explorer 7, 8 and 9, Firefox 3.6 and 4.0, Chrome 11, Safari 5 Mac
Worldwide: Internet Explorer 7, 8 and 9, Firefox 3.6 and 4.0, Chrome 11
Other Considerations
Internet Explorer 6 – IE6 has dropped to below 5% in the U.S. and Worldwide, and has been removed.
Chrome – Google releases major version updates much more frequently than other browsers (five in 2010), and has the fastest update conversion rate, quickly replacing old versions with the newest releases.
Mac Browsers – Current versions of Firefox, Chrome & Safari are all available for Mac; however, only Safari shows more than insignificant use on the Mac, likely because it comes preinstalled on all Mac computers.
Testing Approach for Multiple Browsers – The primary focus of testing with different browsers is to verify page integrity, appearance of fonts, images etc. When multiple browsers are supported for an application we recommend full functionality testing on the most common browser (IE 8), and for the remainder of the browsers:
-
Validate basic functionality
-
Generate error messages to validate display
-
Validate visual elements (such as photos)
-
Submit forms
-
Pop-up blocker
2.0 : Desktop Screen Resolutions
Recommended Screen Resolution: Most websites should be built to support a resolution of 1024x768 to cover a large percentage of user traffic.
Recommended Width for Website Content: While website widths up to 1004 pixels are acceptable, 960 is recommended as it does not carry any risk of horizontal scrollbars, and is a widely accepted standard for a 1024 pixel wide resolution.
Recommended Height for Content “Above The Fold”: An average of 190 vertical pixels of a user’s screen are devoted to toolbars, browser chrome and user interface elements. The remaining space is visible in the browser window when a page first loads. This space is considered “Above the Fold.” For a screen with a height of 768 – our recommended target resolution – the “Fold” is 578 pixels tall, measured from the top of the page.
Other Considerations
Mobile Devices – Screens with a resolution below 800x600 are considered mobile devices, and are not included in this data. The Other/Unknown data may be partially represented by mobile devices that do not report a resolution, but may also include desktop computers reporting unusual or erratic data.
3.0 : Desktop Operating Systems
Core List of Recommended Operating Systems
United States: Windows 7, Windows XP, Windows Vista, Mac OS X
Worldwide: Windows XP, Windows 7, Windows Vista, Mac OS X
Other Considerations
Testing on non-recommended OS – We will add additional OS to our testing plans by client request. For example, when testing Miller applications, the Miller Technographics guidelines should be followed which include Macs.
Testing Approach for Multiple OS – For most projects at Fullhouse, we execute our full suite of tests using the Windows XP operating system. Additional OS are tested in the following situations:
-
Applications with locally installed components: Testing on additional operating systems is recommended when something integral to the application is installed on the user’s workstation. This can be the application itself, or a component used by the application. This includes CD, DVD, screen savers, wallpaper.
-
Websites: The majority of web applications do not require multi-OS testing. In the following situations, testing on multiple OS would be considered:
-
if there is a download to the OS
-
if a browser needed as a part of multi-browser testing requires an OS other than XP to operate.
4.0 : Mobile Web Use
Mobile Web Use As a Percent of Total Web Use
While desktop browsers still serve a vast majority of web users, mobile devices are quickly gaining momentum. Nearly all of these devices run one of the primary Mobile Device Platforms (Android, iOS, & Blackberry). Unlike older mobile devices that could only display very rudimentary content such as WAP or limited HTML, all recent versions of these platforms support full HTML rendering, and can scale down and zoom in on desktop-sized websites for a better experience on the small screen. Still, it may make sense to build a mobile-targetted site to complement a client’s existing desktop site.
Other Considerations
Flash Support on Mobile Devices – Flash content is not supported on iPhone, iPad or iPod Touch, and support among other mobile devices is limited or non-existent. When viewed on one of these unsupported devices, Flash content does not appear, so use of Flash components should be avoided, or alternate content should be available, if mobile viewing is a concern.
6.0 : Mobile Screen Resolutions
Recommended Screen Resolutions for Mobile Devices
Most mobile websites should be built to support a resolution of 320x480 (or 480x320 in landscape orientation) to cover a large percentage of mobile user traffic.
Other Considerations
Mobile Screen Determination – Resolutions considered mobile are simply those resolutions under 800x600. Some mobile devices, such as iPhone 4, iPod Touch 4th Gen, iPad and other tablets are not included in this data set, as their resolutions are higher than 800x600. The data may also include non-mobile hardware with low resolutions. Unfortunately more accurate data was not available at the time of this report.
7.0 : Tablets
Core List of Recommended Tablets
Target OS for Mobile Sites for Tablets – iPad on iOS 4.2.1
Other Considerations
Tablets and iPad Competition – The iPad has a clear first-mover advantage in the tablet market, but competing devices running Android and BlackBerry, many with better spec sheets than the iPad, have been coming to market quickly and may disrupt Apple’s dominance, though any major threat has yet to materialize.
Tablets: Mobile or Desktop? – Based on their screen size and resolution, tablets may be grouped together with desktops, rather than mobile devices; However, because they run the same mobile OS as their smaller siblings they have been considered mobile devices for the purpose of this report.
8.0 : Flash Player
Recommended Flash Player Versions
The following recommendations were created based upon data gathered from available data and previous experience developing Flash sites.
Recommendation for Websites – For Flash applications that do not require version 10, version 9 is recommended in order to reach a wider audience.
Recommendation for Kiosks / Other – Flash version can be determined by the project team and or client. For usage data to help in making this decision see tables below.
Other Considerations
Data Accuracy – Data is provided from two sources, Adobe and StatOwl, to provide a more accurate view of Flash player install-base. Data from Adobe is generally higher than other sources, bringing into question the validity of the methodology used to collect data; however, since no worldwide data was available from other sources, both sources have been included.
Data Timeliness – More recent data was not available at the time of this report.
Flash vs. HTML/Javascript – HTML and JavaScript are supported in all modern browsers, while Flash requires a plug-in. By utilizing javascript libraries such as jQuery and MooTools, some functionality that was only possible in Flash just a few years ago can now be replicated without the need for a plug-in. In addition, Apple does not include any support for Flash in the iPhone, iPad or iPod Touch, and support among other mobile devices is limited or non-existent. When viewed on one of these unsupported devices, Flash content does not appear. Unless there is a specific need for functionality beyond what is supported by Javascript, such as 3D gaming, physics engines, or other advanced interactivity, HTML should be preferred over Flash for the broadest compatibility with users across all devices.
9.0 : Email Clients
Core List of Recommended Email Clients
Because of the diversity in email clients, we recommend analyzing each individual client’s mailing list to gain the most accurate information, since results can vary greatly depending on a client’s demographics. If no information is available, emails should be tested using these clients to cover a large percentage of users.
Email Clients: Outlook 2003/2000/Express and 2007, iOS 4, Hotmail, Yahoo! Mail, Apple Mail 4, and Gmail
Core List of Guidelines For Designing HTML Emails:
No Background Images: Several of the major email clients do not display background images; because of this, emails should be designed with only solid colors behind text, unless the text will be cut as part of an image. This means no gradients behind text, either.
Leave Extra Space For Text: Because fonts render at slightly different sizes in different email clients and web browsers, be sure to leave a few extra pixels below all text in the email. If the Photoshop file has graphics too close to text, it may ‘break’ the email when it’s displayed as HTML.
Max Width: 600px: Sticking to a maximum of 600 pixels wide gives good results in desktop and web-based preview panes, and your design should still be readable when scaled down for mobile devices.
Core List of Guidelines For Coding HTML Emails:
Supported Inline CSS: color, font, font-family, font-style, font-variant, font-size, font-weight, text-indent, text-decoration, text-align, line-height, letter-spacing, text-transform, table-layout, border-collapse, border, background-color . Any other CSS elements may be rendered inconsistently or ignored completely in different email clients, and should generally be avoided.
For a comprehensive list of guidelines, see www.campaignmonitor.com/design-guidelines
Other Considerations
Lack of Image/HTML Support – Email clients are recorded when images are loaded within an email. Some email clients block external images by default (Outlook 2007, Gmail), or are not capable of displaying HTML email at all (older Blackberry models, non-smartphones). As a result these are not reported. On the other hand, some clients display images by default (Outlook 2000, iPhone). This could give an inflated weighting over clients that need user interaction to load images, or are not capable of loading images at all.
Outlook 2000/2003/Express vs. 2007/2010 – All 3 older versions of Outlook used the same rendering engine (based on Internet Explorer 6), so there is no need to test more than 1 of those 3 versions. Outlook 2007 and 2010, however, utilize the rendering engine from Microsoft Word, which has even less css support than the older versions.
Big Movers – Apple had major gains in both iOS devices and their desktop Mail application, mostly at the expense of Outlook, Yahoo! Mail and Hotmail, which all had significant declines in usage over the previous period. These trends are expected to continue, along with Android’s ascent as more of their smartphones get into users’ hands.
Longer Trend Period – Note that the time span between reported values is 2 years, whereas in most other metrics in this report it is only 3 months.
Sources
StatCounter – StatCounter was chosen as the primary source of data due to its large sample pool (15 billion pageviews per month collected from across the StatCounter network of more than 3 million websites), and its reporting of both U.S. and worldwide data. http://gs.statcounter.com/faq#methodology
Adobe / Millward Brown – The Millward Brown report on Flash install-base data is the most widely cited source, but that may be due to its availability on Adobe’s website. Adobe owns the technology behind Flash, which leads to questions about the methodology used, and whether it’s the best way of collecting Flash browser install-base data. adobe.com/products/player_census/methodology/
StatOwl – StatCounter was chosen for Flash data due to its large sample pool (an average of 28 million unique visitors per month to the network of web sites that they collect data from), and as an impartial source for Flash install-base data. http://statowl.com/about_our_data.php
NetApplications – NetApplications was chosen for its unique availability of mobile screen resolution data, and due to its large sample pool (over 40,000 websites, with approximately 160 million unique visits per month). http://marketshare.hitslink.com/faq.aspx
Chitika – Chitika was chosen for its unique availability of iPhone & iPad iOS version breakdowns. Chitika for Advertisers provides the ability to show ads within the Chitika Network of 80,000+ publisher sites that receive over 201+ Million unique users per month. http://chitika.com/advertisers
Android – Android.com provides data about the relative number of active devices running a given version of the Android platform. http://developer.android.com/resources/dashboard/platform-versions.html
BlackBerry – Blackberry.com provides data about the relative number of active devices running a given version of the BlackBerry platform. http://us.blackberry.com/developers/choosingtargetos.jsp
CampaignMonitor – CampaignMonitor reports 50,000 customers in 100 countries using their software. They don’t share more specific information on number of emails sent, but their customers include Facebook, Twitter, eBay, and Apple to name a few, hinting at a very large recipient base. http://www.campaignmonitor.com/customers