Mobile Musings

Revenue Leakage and Breakage Bonusses

I recently read an interesting article about Revenue leakage over at MocoNews: http://www.moconews.net/entry/419-vodafone-fesses-up-on-faulty-reporting-content-companies-setting-up-own/

There are definitely inaccuracies in reporting, but it goes both ways:

A major carrier has outages in its billing system from time to time - usually no more than a few minutes.  When this happens, off-deck connectors into the billing system can't collect money, so they don't allow users to get the content (In fact Bango automatically switches to PSMS to keep revenues flowing) . 

We discovered that during such outages, the "on-deck" content is still sold (to provide good customer service) but ther fail over to "free mode" so all content becomes free during the outage. In this case however, it turned out that the carrier was simply not reporting the sales to the content provider, who was therefore subsidising the "good sevice"....  EEEEK!

Two ways we have seen content providers or aggregators benefit:

(1) Content providers sell content using a single or double opt-in Premium SMS system.  Unfortunately the user does not have WAP connectivity or the content does not work on their handset.  Provided the user does not claim a refund, the CP can detect that content has not been delivered, pocket the money (potentially for many months on a subscription service) and not pay any royalty.  For some CP's and aggregators, this "breakage" accounts for their entire profit margin - everything else is break even!

(2) We discovered that a "top five" music publisher was allowing users to purchase from their WAP site by sending them a premium SMS and then routing them to a URL where the content was downloaded.  Between 10-15% of all the PSMS's  failed to deliver revenue (user had run out of credit, barred by user, corporate account etc.). So, at the end of the month the company simply reconciled things and did not pay the artists any royalty on these sales.  Unfortunately, the end users had already been given teh URL for the content - so it had been supplied.  A back of the envelope calculation indicated that they potentially owed artists royalties to the tune of $450,000.  This was a strong motivator for them to sign up to Bango - where such leakage was eliminated!   There must be many other such situations.

Over in the UK this week the big "scandal" is about  the alleged $80million phone in fraud: http://www.theregister.co.uk/2007/04/24/gmtv_phone_scam_allegations/

Sobering!

Bar Codes and other Camera stuff

There is some interesting discussion about bar code scanning from mobile phones over at http://tech.groups.yahoo.com/group/momolondon

Bango worked with Interactive Wireless on phone to web stuff in 2002,
and again with another company (that sold out to OD2) in 2004. We
also did a big trial with Vodafone in 2005.

http://www.3g.co.uk/PR/July2000/3721.htm
http://bango.com/assets/data/pressreleases/57_bangospot.pdf
http://theponderingprimate.blogspot.com/2005/07/mobile-marketing-its-
biggest.html

We discovered the "two gotchas" about this technology:

(1) Walled garden think. For many years, mobile operators have been
beating phone makers and preventing progress in Java standards to
prevent the microphone and the camera on phones being accessible to
downloaded software. This was because they were petrified by people
writing software that curcumvented the MMS sending and voice features
of phones - which they believed would generate a lot of money.
Allowing MMS "bypass" and VOIP clients would be a bad thing. Even
though by 2004-5 the walls were coming down, the change of course in
phone design means a 3-4 year delay in open platforms coming
through. SO, expect progress on downloadable software that can use
phone features like camera with widespread use in 2009-10

(2) Patent fear. Neomedia is a small company that has a patent on
linking barcodes to internet content.
( http://www.eff.org/patent/wanted/patent.php?p=neomedia ) Mobile
phone companies like Vodafone, T-Mobile and Orange were paralysed
from acting by fear of a patent infringement case. As a result,
the "safe route" of letting QR codes "covertly" appear in phones (as
a "japanese feature") seems to have been the strategy adopted. As a
result, hardly anybody knows that their phoen can read QR codes - so
they are not used outside Japan. This is a further "business
depressor" for innovative barcode companies with more robust codes
than QR.

The Great Phone Number Harvesting Scam

There has been a great deal of interest in UK mobile phone number harvesting and it is increasingly being referred to in the mobile industry as "the next subscription scam".

I thought it would be useful to outline what the scam is, the problems being caused and the steps Bango thinks need to be taken quickly, before it gets out of control.

The Scam
When someone visits an internet site, there is no way that the site owner can find out their email address or the phone number of their broadband line without being given the information by the user. In the same way, users browsing the mobile web do not expect their identity to be revealed to the sites they visit.

But mobile operators connect phone users to the internet.  Whilst they know the subscriber's phone number they generally don?t provide it to third parties, except under strict confidentiality requirements, on a case by case basis. However, in recent months, the UK mobile operators have started to provide phone numbers (MSISDN) to accredited intermediaries, to enable them to send premium text messages to users and facilitate payment for content. 

A set of  "scheme rules" known as Payforit that mandates that the MSISDN is only provided to content providers that have user permission is currently being established by the UK operators. Unfortunately, citing technical difficulties and other reasons, some accredited intermediaries have started to simply pass on phone numbers in breach of the guidelines. 

This has in turn led to the problems that were foreseen by the operators when they devised the Payforit scheme rules.  Many thousands of users have already visited mobile web sites that have harvested their phone number.  And just like valid email addresses, these phone numbers have considerable value for marketing purposes.

The classic scenario is that a content provider buys popular search terms in Google Mobile Search. When users click on a link, the content provider harvests the user's phone number, which they could then sell on to third parties. 

A typical top slot in Google Mobile search for a popular term like "ringtone", "porn" or "free" might cost around £1.00 a click - because the content provider can make a return by selling content for more than that value. A harvester will actually sell the content and then use the phone number for further marketing to that user, or sell on the phone number to others, with an indication of what that user is looking for. 
The return on this is obviously much higher than the simple content sale, so the harvesters are able to out-bid the clean suppliers.

The Benefits of the Scam
Interestingly, the short term gains can be significant for all the parties - except possibly the end user:

- The content provider sells more
- The intermediary gets more traffic by bending / breaking the   scheme rules
- The operator participates in more billing and gets more data traffic and more message traffic
- Google gets higher sales values for its keywords

This makes the motivation to stamp out the abuse or comply with the rules much lower.

The scam appears at the moment to be most popular for users seeking out adult content, perhaps because the likelihood of complaints is much less than with ringtones etc. However, it is possible that many of the phone numbers that are being harvested are those of children ? this has the potential to open up a legal minefield.

The Solution
The Payforit Scheme rules state that:  "Merchants wishing to use an MSISDN (phone number) for marketing purposes must only do so based on the consents obtained from the Payment and Subscription Confirmation pages presented and accepted by the consumer" and "Accredited Payment Intermediaries must log all consumer consents and ensure that the information is auditable."

Unfortunately, since the phone numbers are being passed to content providers without consent of the consumer, once those numbers have passed from one content provider to another, it is impossible to track down the original source of the number.

Vodafone took steps to stop the scam when it surfaced last year:
http://www.mobile-weblog.com/50226711/vodafone_blocks_ipx_in_uk.php
however since that time it appears that the problem has become more widespread - across many mobile operators and with many intermediaries.

For information on Payforit
www.bango.com/payforit

Web Apps vs Phone Resident (Java) Apps

There still seem to be quite a few companies an dindividuals pushing the idea of mobile resident java based shops, portals, magazines, vending machines etc.

I'm really surprised that people see them as having any future, especially now that browsing speeds and flexibility are improving on mobiles.

Web 2.0 is all about building things on the server and enabling them to be used by all on teh device.  Its avbout incremental development an dno need to update the client.

A browser approach eliminates the user install cycle – each day the service can be improved without me having to upgrade any software. The functionality of the archetypal PC app was fixed 5 years ago by whatever was implemented by Microsoft or whoever, and issues with installed base, file compatibility, simultaneous migration of multiple clients etc mean that the core functionality can only now evolve very slowly, if at all.

Of course there are also all of the counter-arguments in favour of resident apps. By my main point here is that these benefits are incidental.  A great mobile service benefits from mobility AND the network.

The “long tail” application world (on mobile at least) is made possible on the mobile web by virtue of the small incremental cost of development and teh power of teh browser model.

Paypal promoting use from Mobiles

Not many people remember that X.com turned into PayPal. Until recently you could actually log in to x.com on your mobile and log in to paypal.

Back last year, Bango started working with PayPal to enable mobile internet sites to use PayPal for micropayments. The collaboration is working very well. See http://www.bango.com/investor/paypal_120905.pdf for more info. Users can navigate round WAP sites and make one click payments, because Bango is tightly integrated with Paypal's systems. Nevertheless, connecting a PayPal account with a mobile phone to enable this process requires (at least once) a user to go to a PC.

From this week, PayPal have started educating users about mobile from the PC end. Paypal mobile! uses a mobile phone as an extension of the PC to instruct payments to be sent to paypal merchants or other users. Once a phone is connected to a paypal account, the authentication of the payment is done by IVR (basically Paypal initiates a voice call to the user).

This is great news. Using a text message or a voice call to transfer money to people or merchants is a huge step forward from requiring a "logged in PC" and will open up PayPal to a much wider audience.

While the new system is more complex for mobile internet sites, where the existing PayPal/Bango system works very well, the new system has application where the user is not on the mobile internet when they make a buying decision. Bango will be able to leverage both the text driven and the voice driven stuff to enable users to pay or pre-pay money to mobile content providers before the user visits their site to collect the content.

With thousands of content providers using Bango to get to market, Bango is obviously very keen to exploit - on their behalf - any new payment methods that achieve mass market traction. Lets hope Paypal by text and voice is one such method.

Trends seen in 2006

2005 has seen a migration from content through SMS to WAP portals (just like 1995 saw an evolution from email services to web services)

Mainstream brands have entered to market directly, rather than solely through operator portals. Examples include SKY TV, THE SUN, Bands like OASIS, WESTLIFE, EMINEM Music Labels like EMI (The Raft)

Moble Network Operators (Carriers) are moving to support the D2C content channel. They see D2C on the mobile internet as the technical solution to SMS billing headaches, providing better consumer transparency and a better audit trail.

Bango's slashing of the costs of going "off-portal" provides lower costs of entry, meaning more content players can get into the market. We are seeing independent music labels getting into mobile, smaller content providers etc.

In the US, the content market is gradually opening-up giving consumers greater choice and driving better quality through competition. While Carriers like Verizon are still opposed to giving their customers choice of content supplier, the leading network - Cingular - is on a clear path to supporting its users in getting what they want.

In the PC-centricUSA, linking mobile services off web sites is a key traffic generating tool. Sun Microsystems uses its Java.com web site to promote the benefits of the Java user environment and allows the user to experience Java content by literally ordering the download off the web page.

Motionbridge bought by Google - or was in Yahoo!

Heard a rumour today from a credible source that Motionbridge (www.motionbridge.com) may be getting bought by Google for about £10million. No announcement from Google or Motionbridge so perhaps work still in progress. Siemens Mobile Accelerator Fund ( the biggest investor) is trying to sell on its investments at the moment so that makes disposal at an early stage somewhat sensible.

Motionbridge provides mobile focussed search for several mobile operator portals (O2 UK and Orange FR for example) and has some quite neat technology. They run the "paid for keywords" system used by O2's search advertisers.

Google or another search company like Yahoo or MSN might like to buy Motionbridge to get a short-cut to operator portal presence, and to provide the "metatag" type search that works beter than crawling for mobile sites.

Access to Sprint http://www.sprintpcsinfo.com/modules.php?name=News&file=article&sid=1008
might also make sense.

Why MMS is dead for mobile content delivery

I was on a panel about mobile content at the mobile youth conference recently in Barcelona, and happened to ask the audience (mostly mobile operators) if anybody believed that MMS had a future in delivering mobile content (rather than peer-to-peer usage). Surprisingly for me, I could find nobody to support MMS for content delivery - whereas a year ago, many mobile operators would have said it did have potential or were investing in MMS for premium content delivery.

This accords with my view. Message based delivery content is less successful for Content Providers for three fundamental reasons:

1. There’s no easy way to know the capabilities of teh target device when preparing a message. The provider has to take a lowest common denominator approach to the content they deliver. In the WAP model, where they send a link to teh content page and the user's device accesses the web site, the CP can optimize content delivery to the device type, giving the more data hungry customer with a better device a very good experience.

2. Marketing is expensive with messaging services and it’s hard to build customer loyalty. In the browse-and-buy model, we see a 46% user return rate versus only 18% for messaging services. User spend per visit is up to 50% higher for 3rd party content delivered through a mobile internet site versus the one-shot messaging delivery approach. Content providers prefer the browsing model (which I assume is better for carrier data revenue….) because it’s a better brand experience and users can bookmark favorite sites.

3. Users prefer to pay for services through a WAP site presentation. The charges are much more transparently presented to the user at the time of purchase and there isn’t the over-reliance on “small print” in the initial advertising. In short, there are fewer customer purchase problems to deal with.

Content providers in Europe dropped MMS as a viable content model in favor of WAP browsing. MMS is used in certain contexts for alerts and some outbound marketing, but almost all 3rd party providers now use the mobile internet for their core content services.

.MOBI has no clothes

I wrote a letter to the editor of New Media Age today just to make clear my views on the .mobi "top level domain". Here it is:

.MOBI HAS NO CLOTHES

The "approval" of the .MOBI top level domain by ICANN is interesting but pointless. .MOBI is a red herring that will be consigned to the dustbin of irrelevant initiatives - like PL/1, Ada, Algol 68 and OSF/1. Here's why:

Content Providers want to promote ONE "address" in written, electronic or verbal forms and want it to work through any medium - mobile, fixed or word of mouth. The technology exists to support this today. For example http://www.vodafone.com/ works well a PC or from a mobile phone.

The WWW Consortium advised ICANN to reject .MOBI early in 2004. Browser protocols provide an HTTP_ACCEPT header to a single URL to be used across all device types - hiding complexity from users. World Wide Web creator Tim Berners-Lee warned that a new .MOBI domain could be harmful, potentially breaking the web's device independence and putting mobile users in a walled garden.

The sponsors of the Irish company behind the "scheme" claim to want to provide a good user experience. It's therefore amazing that after several years, most of their own web sites give an error message to hundreds of millions of mobile users. gsmworld.com, microsoft.com, sun.com, and even nokia.com all give an error message to most users because they only work on PC browsers.

If the .MOBI initiative gained any traction, it could complicate and confuse the Internet experience for mobile subscribers, and slow down the delivery of mobile services. Fortunately, enough people in the Mobile Content industry have pointed out the obvious flaws in the proposal, so few people will be embarrassed by praising the clothes of a naked emperor.

However, just in case its not absolutely clear, let me be like the little boy who spoke the truth: ".MOBI has no clothes! " If any of the sponsors need help in developing their web sites to enable them to better support mobile devices, as Vodafone has already done, we are certainly more than willing to help provide them with the appropriate information.

.MOBI top level rejection notice

After some days studying the proposal to ICANN for a ".mobi" top level domain, I decided that the idea was not only ill thought out, but (like Tim Berners Lee) could actually harm the progress starting to be made on the Mobile Internet.

I therefore filed my objections to ICANN here:
http://forum.icann.org/lists/stld-rfp-mobi/msg00017.html

It turns out I'm in good company - almost everybody who has commented seems to realise the problems with this crazy idea.

However, given past experience with ICANN, plus the propensity for large organizations like the sponsors of .MOBI to not really have a big enough view of what is going on to really understand what they are backing, I expect that in a year or so the .MOBI idea will be approved and we will see some PR effors and stuff in 2005 to try and get it going.

I doubt it will really get anywhere - but who knows! :-)