Welcome!

Login Register
< > iPhone: The great game development ... MIDEM 2010: Nokia warns music ...

Mobile analytics: Could do better

Mobile analytics: Could do better

Don't expect web-style depth from most mobile analytics, says Bango's Ray Anderson.

Millions of users have discovered the ability to browse from a continuously available, personal and compact screen – the mobile phone.
 
As a consequence, brands and content developers are directing more resources to creating on-line
content suitable for mobile phone use, in turn, the companies that provide technology to enable online business are prioritising investment in mobile platforms, tools and devices.
 
But al too often, the mobile versions of cornerstone internet technologies in commerce, search, on-line advertising, tracking and analytics are unsatisfactory.
 
It's not hard to see why: billions of phones in varying specs all designed to hit different geographies, budget, usage priorities, networks and so on.
 
With such complexity, solutions successfully crafted for the PC internet do not translate easily into mobile internet. For example:

• Search results from a phone deliver web pages unsuitable for phone browsing and fail to correctly index the corresponding mobile internet pages.

• Payment from a phone requires users to enter their network and phone number, or send text
messages. If the user isn’t identified by the network, they often cannot pay at all.

• Analytics and CRM tools frequently fail on mobile, seem unreliable or, worse, actually break the mobile internet user experience (try Google or Omniture analytics). 
 
Many companies assume that mobile is a smaller, less capable version of the PC web. But this assumption does not recognize the true importance of mobile, the complexities it introduces or the massive benefits it brings.

Far too many analytics solutions today simply add handset detection and basic operator IP address identification, but completely fail to address the real complexities that mobile presents. The most significant shortfall is their inability to identify individual visitors, especially on most of the mass market handsets.

Here are some of the key issues:
 
Firmware: Operators typically specify variations in phone capabilities which can include different browsers and connection settings. For example, some operators will let phones connect to Wi-Fi in preference to using their own network. 

Cache and cookie behaviour: Some phones reset cookies when the browser is shutdown or the phone power is lost. Phone capabilities can vary considerably based on operator settings, for example, some operators prevent media from being saved locally to the handset while the same phone with another operator permits it.

Browser variations: Browsers such as Opera Mobile on today’s phones provide many more standard web capabilities than the low spec WAP browsers found on older handsets. But they still may not provide reliable visitor identity due to many other mobile factors. Even the most modern mobile browsers may have security restrictions in place that impact analytics – for example the iPhone blocks acceptance of third party cookies. 
 
Markup languages: Most new devices support xHTML rather than WML or other older formats, but there are still many legacy handsets in the market that should not be ignored.

Gateways, countries and MVNOs: An operator gateway acts as a proxy to all data and will be the IP address reported for the connection. This means that all customers connecting through a given gateway will report back the same IP address, making it of little use for visitor identification.

Connection type:  WAP gateways are still heavily used and tend to assume the browser is limited. They may prevent cookies reaching the handset or deploy transcoding solutions more actively.
 
Cookies: While most modern mobile browsers support cookies there are still complexities to consider that prevent cookies from being as reliable as they are on a PC. Some handsets still fail to persistently store cookies between sessions or clear cookies when the browser exits or the phone loses power.

Transcoders: These are commonly used by many operators to “improve” the layout of traditional websites for use on mobile phones. Transcoders often remove content, change headers, block cookies, strip out operator information and change the site content, including tracking code.

However, with the right knowledge and partnerships, mobile analytics has the ability to deliver higher levels of accuracy than the best traditional analytics solutions can achieve. 
 
These shortcomings are pretty serious for any brand or content provider seeking a mobile presences. It's why Bango has developed a mobile analytics solution that takes many common mobile complexities into consideration to deliver up to eight times higher accuracy than traditional analytics solutions.

You can find out more here.

Advertisement

Tags: This article has no tags