A weak signal is a sign which is slight in present dimensions but huge in terms of its virtual consequences (Pierre Masse 1965)
Author: Ian Hay
Team: VST&I Strategic Advisors
Contact: firstname.lastname@example.org +44 7976 355070
Date: Sunday, 14 May 2006
This paper outlines a potential area that FT/Orange could pioneer and lead the market in as the web evolves and the patterns of user behaviour changes through generational differences.
The longer term vision outlined here can be built towards in small steps, the nature of a personomy is that it exists as soon as some information is there but gets richer and more useful as additional amounts and types of data are added, this also leads to more services being possible as the data source grows and diversifies
An initial step towards this vision could be an SMS/MMS vault where every message sent or received by a user is stored permanently, they would be able to access this through the web or it could be exposed through the API to applications such as IM clients or email so that when writing to a contact you would be able to see recent messages relevant to the specific contact
The next step could be call details, again exposed in a similar way to enrich the contextual information around any communication by providing a history, or equally it could be ‘clickstreams’ the data for each page visited in a browser, how long etc
Personomy is a recently coined term and is defined as the Personal Information Cloud that is formed by but independent of any platform or system. Effectively it is the digital footprints or traces left by a person in a digital services world, like footprints in the real world the majority of these traces are left behind with no further interaction or use made of them, and eventually they fade away.
To take one example relevant to Orange we can look at SMS, as a user sends and receives 100s or 1000s of SMS they create a communication trail that is rich in contextual and social metadata, however apart from the few saved to a handset the majority of these rich pieces of data pass through rating, tariffing and billing systems and are then either stored and archived/forgotten or simply purged/deleted.
The mobile phone is an excellent data gathering device if all the relevant info is stored centrally and can provide a great deal of semantic information relevant to building a personomy, once this information is harvested and stored in a central repository it can then be made of use of in a multitude of ways, however the ideal tool for doing this is not the mobile but the PC through web services and a feature rich API
For a personomy to be most effective the amount of raw individual feeds to it must be completely representative of the activities of the user in the digital world, Orange already has an advantage in this area due to the existing streams it has the possibility of harvesting:
SMS, MMS, Call Details, Billing Records, Location, Mobile Data, Cameraphone pictures
With further development of a personomy aggregation platform and the required API it would also be possible to capture further info streams such as PC clickstreams (the history of websites visited and when) IM conversations, email traffic, VoIP calls for example. By having a flexible API available to allow 3rd parties to write for almost any application using IP protocol could be made to submit to a user’s personomy
There is a lot of focus on the Web2.0 meme lately with an emphasis being placed on open APIs to allow the user or developer to mix up or mash different data sets together, at present this is focused on vertical silos of information gathered by specific applications such as a Flickr photostream with a Google Maps API. However the silos of information are held in exactly that, silos within applications with lots of different providers and sources.
Sites like Technorati allow forms of aggregation of varied sources through Tagging but this relies on querying several data sets for the same tag (keyword) and aggregating the results on a web page.
A growing meme is also Attention Economy as described by organisations like AttentionTrust.org which state that a user should always own their own data and choose what to do with it, be it keep it secret or to publish it for some form of gain, the portability and ownership of user data is a key aspect of the burgeoning Web2.0 world.
Orange and FT are in a highly advantageous position to become a leader and pioneer in this area due to the very nature of being at the core of a user’s communication world and by virtue of having sufficient data sources at hand already to begin the process of building a comprehensive personomy.
As web2.0 evolves there are problems emerging too, mainly in the area of ownership of data, portability of data and usage of data, initiatives like StructuredBlogging, RedirectThis are starting to address this but they are working without the inherent benefits that Orange has as stated above. The richness of communication between services through open APIs is billed as the future of the web and if this is true. Which I believe it is, then the data that flows through the APIs is the currency and a Personomy can be viewed as a bank account. It follows that the organisation that can build ‘the bank’ to hold all these accounts will be in the ultimate position to leverage the wealth stored within. This is not to suggest that a
Knox is required where the wealth is kept stored behind maximum security protocols though, it must be addressable through an equally rich API that will allow 3rd party development to occur.
The analogy holds true in terms of privacy and usage, the user that stores his information there is always in complete control and will decide what can be done, i.e. what services his information can be invested in, and of course can withdraw his info at any time should he choose to leave, however once enough information has been built up and services are exposed through it the level of stickiness becomes such that there would be little benefit in closing an account.
In simple terms
Orange should create a Personomy Vault which can capture information sources through an API, in the beginning by storing and indexing those data sources that already exist such as SMS and MMS, add in Call Details and location the available personomy is already rich enough to start offering some services.
Further extension by use of plugins for the mainstream browsers the users’ clickstreams would enrich it further and proxying IM traffic could allow for an extremely rich communication personomy to become a reality
Orange could initially offer personomy data out through Widgets to be either simply displayed on a user desktop or published through a blog should the user choose to do so.
Extensions to IM clients could see semantic data displayed while a conversation takes place such as when a user last sent an SMS or MMS to the person they are talking to, or the last call made, email sent, time they were together etc
By making the platform completely open in terms of data sources and in republication of that data (strictly according to authentication rules and privacy naturally) the value of a personomy will grow as more data is added, there will be a growth period required while raw data is collected to achieve its full potential but even in the its earliest form useful services around SMS storage and retrieval and communication history in general can be created very simply.
With the growth in open API structure in the web it is also possible to import existing data streams from certain services into a users personomy thereby further enriching it with semantic data that Orange has had no part in creating but that it can still help a user leverage.
The key is providing the service and allowing the user to feel in charge of something they haven’t been before, empowering them to make choices about how their own data is used and giving them tools with which to expose that data as they see fit to allow them to create more contextually rich services than are currently possible.
Additional examples could be the seamless building and collecting of personomy data which would lead to an address book that is contextually enriched by usage info, much like Google Gmail contacts, all contact points would be stored meaning that even if a mail is only received once the contact details would be available to any other service accessing this part of the personomy. One final example could be that presence information could be enriched with contextual communication history for each contact in the list, when last called, SMS etc
Once the data within the personomy grows not only per user but in total database size the possibilities are endless, and certainly not predictable at this point which is why openness to the developer community is key.
With the explosion of new web services we are seeing, the multitude of vertical silos of information, the growth in the Millennial Generation and perhaps most importantly the full on adoption of web2.0 principles by the big 5 GYMAA (Google, Yahoo!, Microsoft, Amazon, AOL) the time is certainly now to be developing this play.
The market is nascent at best in this particular area with the strategy of GYMAA and other web2.0 successes like Flickr and Delicious so far being to offer API access admittedly but only to the vertical silo of user information they have gathered so far.
Due to its position Orange should be able to implement an early entry to this market and in fact help shape it through the assets it has at its disposal today, the key being to start with what we have and then allow user feedback to shape what gets added, additionally the developer market will foster innovation in this area and hacks (in the positive sense of creating something unexpected and new from something existing) will enrich it in ways it is not possible to predict at this point.
The provision of a personomy vault and the associated services could give Orange a competitive advantage and differentiation in the market place and play a key role in supporting its claim to be the first fully converged operator by showcasing what is made possible when true convergence happens in a communication company, especially one that disproves traditional Telco preconceptions (as perceived by those on the outside that complain about the walled gardens and lack of access) and not only embraces the evolution of the web and mobile past the current 2.0 memes but leads the way in empowering and putting the user truly in the centre of their digital world so that they can create significantly more personalised services than they could with other providers.
A recent article by a prominent blogger in the telco space pointed out the following:
You would have thought that learning from user behaviour and anticipating need would be a high priority for mobile operators looking to reduce churn. (Don’t expect handset vendors to help – they want handset churn, not stickyness). Yet I suspect that precisely zero of the operators are making the move from vertically integrated telcos to horizontal plays as their core strategic driver. Without breaking any commercial confidence, I would say that the criteria used at Sprint all related to secondary (or lower) phenomena, or confused strategic objectives with tactical metrics, or were too generic to be of real value*.
* related to an initiative he consulted on at Sprint to make a platform that kept user data hostage and attempted to make it too hard to leave