NB A full PDF copy of this briefing can be downloaded here.
This special Executive Briefing report summarises the brainstorming output from the Technical Architecture 2.0 section of the 6th Telco 2.0 Executive Brainstorm, held on 6-7 May in Nice, France, with over 200 senior participants from across the Telecoms, Media and Technology sectors. See: www.telco2.net/event/may2009.
It forms part of our effort to stimulate a structured, ongoing debate within the context of our ‘Telco 2.0′ business model framework (see www.telco2research.com).
Each section of the Executive Brainstorm involved short stimulus presentations from leading figures in the industry, group brainstorming using our ‘Mindshare’ interactive technology and method, a panel discussion, and a vote on the best industry strategy for moving forward.
There are 5 other reports in this post-event series, covering the other sections of the event: Retail Services 2.0, Content Distribution 2.0, Enterprise Services 2.0, Piloting 2.0, Open APIs 2.0, and Devices 2.0. In addition there will be an overall ‘Executive Summary’ report highlighting the overall messages from the event.
Each report contains:
- Our independent summary of some of the key points from the stimulus presentations
- An analysis of the brainstorming output, including a large selection of verbatim comments
- The ‘next steps’ vote by the participants
- Our conclusions of the key lessons learnt and our suggestions for industry next steps.
The brainstorm method generated many questions in real-time. Some were covered at the event itself and others we have responded to in each report. In addition we have asked the presenters and other experts to respond to some more specific points.
The implementation of new ‘Two-Sided’ Telecoms Business Models has major consequences on telco network architecture. Perhaps most importantly, data from separate internal silos needs to be aggregated and synthesised to provide valuable information on a real-time basis. Key process interfaces that enable new services must be made available to external parties securely and on-demand. Network and IT functions must start collaborating and function as a single entity. Operators need to migrate to a workable architecture quickly and efficiently; vendors have to support this direction with relevant new product offerings and strategies.
- What are the implications of adopting 2-sided business models on telco technical architecture?
- What does the roadmap to a Telco 2.0 architecture look like?
- As the network becomes more intelligent to support smart phones and App Stores, what are the most important investments for telcos?
- What are the priority areas for transformation to enable new services?
- Why are user profiles so important for telcos?
Stimulus Presenters and Panellists
- Werner Vogels, CTO, Amazon.com
- Richard Mishra, Director, Strategy and Standards, Amdocs
- Alireza Mahmoodshahi, CTO, COLT
- Paul Magelli, Head, Subscriber Data Management, Nokia Siemens Networks
- Michel Burger, Head, Service Architecture, Vodafone Group
- Thomas Rambold, CEO, DESS, Associate, Telco 2.0
- Chris Barraclough, Managing Director, Telco 2.0 Initiative
- Dean Bubley, Senior Associate, Telco 2.0 Initiative
- Alex Harrowell, Analyst, Telco 2.0 Initiative
Stimulus Presentation Summaries
Thomas Rambold, Associate, Telco 2.0, presented on the end of clearly defined services – we now face thousands of segments, and no distinction between ”just voice” and ”just data”. Broadband means that I can simultaneously be an Amazon user, a father, and many other things. Distinctions between enterprise and consumer services, private and public, have changed.
This implies much greater complexity in the customer relationships; the Over-The-Top (OTT) players have struggled to get their arms around the total relationships. Carriers have identified Fort Knox on the diagram (customer and network data management) as somewhere they can excel.
The walled garden is no longer sustainable. OTT offerings are already growing fast and, when they start using Telco APIs in earnest, this will accelerate. We need to link these services with customers – the carriers are the only actor capable of acting as a broker between the API users on the one hand, and the full suite of customer data (residing in ‘Fort Knox’) on the other. It is vital to be on-demand, if not necessarily truly real time. You can’t make people fill in forms to start a service – authentication, sign-on and billing and payment need to be automated.
Paul Magelli, Head, Subscriber Data Management, NSN: ”Would you be happy letting a Telco do data mining on you that you wouldn’t tolerate from the Government?”
Currently we are struggling to bridge the gap between the network and the OSS/BSS systems. We simply can’t get the on-demand response we need. So now we’re merging telco and IT organisations together, analysing the telco and IT environments, trying to get them to work together, in concert, and react faster.
The service delivery environment; a year ago, this would have been a big SDP from a traditional telco vendor. But now the services have moved out to the network – into the cloud, into mash-up environments, on to native applications on devices. It’s increasingly challenging to make these services useful and timely and to guarantee security and privacy across all these domains.
Service Value Management, Amdocs’ vision for next-generation services
Richard Mishra, Director, Strategy and Standards, Amdocs, said we are living in interesting times: We face sophisticated customers, exotic delivery platforms, and a recession! Perhaps we need to revisit core values, get back in the box and think more deeply. Revisit the core strengths and disciplines of operating a carrier-grade telecoms network (but not retain the bad attributes).
There is constant talk of service, even by the TM Forum back-office people, who are never seen from outside the Telco. But the next step after resource deployment is, inescapably, fulfilment. And with all this talk of service, what about the shareholders? Finally, having created the service, spent the capex, and deployed, you need assurance – monitoring service performance against the performance you offered the customers.
We’re working with ever-increasing interdependence between infrastructure, service, consumer and enterprise applications, and devices. This needs diagnostics and monitoring for all these levels, and careful management of the consumer experience.
Our Full Service Provider Integration Framework provides tools for continuous improvement as defined in the left half of eTOM; covered by contracts designed to match. In our deployment in Atlanta, we used this and special tools we developed for the carrier Ethernet network. We didn’t try to recreate the special capabilities there; but we did subject it to the traditional disciplines of managing a carrier network.
As a rule, Telcos no longer customise stuff; we can’t make our own SDH management tools any more, notably for reasons of intellectual property. So it’s now a question of assembling agile value chains from many other vendors, components and sources.
Building Trusted Relationships
Paul Magelli, Head, Subscriber Data Management, Nokia Siemens Networks said that customer data plays a key role in Telco 2.0…but does it exist? Is there enough available in our networks?
Nokia Siemens is working on the following assumptions:
1. A multitude of business models;
2. Broadband connectivity everywhere;
3. 5 billion subscribers;
4. Applications all migrated onto the Internet.
This implies that successful applications and services will be information- and subscriber-centric.
Richard Mishra, Head, OSS Strategy and Standards, Amdocs: ”Data will become a treasured part of the business model.”
We really need rich profile information – profiles are what we say about ourselves. It’s clear that there is enough information, but can we get at it? 76% of respondents in our survey think it’s the most important issue; 86% think it’s important for network development.
Consider a use case in banking. To improve contact with customers, we need to know things like: is the customer available? Interested? Is it a good time to reach me? Is this the best way to reach me? Is this the right language? For this, we need the ability to do real-time subscriber profiling as well as historical data analysis.
But it’s more complicated than that. Current data isn’t enough – time series is really important. It is surprising that operators recognise this but haven’t done very much to solve it. Only 14% have real-time data analysis. And identity is more complicated than that – people have multiple devices, multiple SIMs, and multiple identities. Privacy is another big issue. Permission is frequently abused; there is a huge generation gap in attitudes around what constitutes privacy, and the legislation is very different between different jurisdictions (and usually lags the market).
If we could provide a single point of access for managing your identity…huge opportunities await. But it’s crucial to resolve the privacy issue by giving customers control of their own data.
Building the business-grade cloud
Alireza Mahmoodshahi, CTO, Colt asked how much do our customers know about cloud computing? Not much, he said, we’ve done well in clouding their minds. He gave a short brief on COLT; its origins in the City of London, its European fibre network, its large enterprise customers.
There is a framework for ICT; the Telco at the bottom, providing dark fibre, bulk data and voice. Then, above that, data centres and IT infrastructure like hosting, co-location, network operation. Then a vast range of applications specific to tasks run on top of that.
Not many operators are enamoured with their position at the bottom of the stack…
…on the other hand, it’s hard to find anyone who can replace the things that Telcos can do. Most clouds don’t offer any kind of SLA, so critical transactional services can’t use them. Traditional clouds are on the left of the diagram, similar to IT infrastructure. Operators don’t want to be penned into the low-value bottom right: they need to push up and left to escape. Meanwhile, clouds tend to have no SLA for the sector from the cloud to the end-user, and not necessarily between the enterprise and the cloud: there are too many participants to provide an SLA covering the whole thing. So the opportunity for the Telco cloud is to provide end-to-end SLAs.
The crucial development to enable business-grade clouds is to virtualise all elements of the system. Then, control priority for applications running on them. This is achieved by queuing them through a COLT-patented policy scheduler; this despatches tasks to a pool of virtualised servers, themselves providing a pool of threads.
Then we set up the API/SDK third-party access to the platform to empower the applications developers. If you want to do this you need to control QoS and also application-layer dispatching across the entire system; carriers are probably the only actors who can offer this.
The fragmented and sometimes opposing views of the audience in the feedback (and voting) are not new or surprising. It is a key indication for service providers and operators that changes are overdue and must be taken seriously. It also demonstrates the complexity of current silo’d approaches, and the inability of a single company to change this. Despite this complexity, operators will have to focus investment on strategically important projects, particularly during these difficult financial times. Such an approach will give us the chance to reduce complexity and reduce operating costs. Most importantly it will enable operators to produce agile technical architectures with the required flexibility to meet customer demands.
Feedback: grounds for optimism
The technical architecture session produced some significant levels of optimism from the audience….
· Promise of new future. [#5]
· Some real applications of Telco 2.0 model. [#6]
· Good ideas for large enterprise and government segments. [#12]
· Triggers my thinking and confirms my assumption. [#15]
…albeit tempered with a degree of cynicism about the validity of the examples cited
· We don’t know when we will get there. [#9]
· Still industry jargon driven. [#17]
· All seemed to be a bit ‘defensive’ of legacy models and were more oriented to Telco 1.0. [#18]
o Re 18 good point. 90% of large carrier environments could not move to Telco 2.0 due to restrictive contracting structures with their existing OSS vendors. To add to the pain, they usually don’t have their source code so they don’t really control their destiny (especially if outsourced). [#43]
· Mostly still talk, no examples on the ground yet. [#20]
· Still very much 1.0 and technology driven, to view user and real new business model. [#44]
· Unfortunately human beings are not compliant to marketers Telco use cases ;-). [#46]
o 46; Telco use cases are uninformed by real data. [#49]
· Would it be better to define what to do and where to go, before arguing potential obstacles and potential regulatory matter before the idea has started? [#62]
Feedback: Customer data issues
In particular, the subscriber data management model seems aspirational, if trust issues can be resolved. However, there were a lot of questions wondering whether Telcos’ use of customer data would be either as accomplished – or consumer-friendly – as Internet-based players like Amazon’s. There also appears to be a debate brewing over whether or not Telcos’ customer data is “better” than that of web players’.
· Single customer data ‘vault’ from Nokia/Siemens is a great idea; doubtful any carrier brand has the consumer trust to pull this off-more likely for VeriSign, Symantec or someone else. [#25]
o 25 – I agree. There also needs to be reciprocity. I want to be able to authenticate on an operator network with my Facebook ID or similar – or with another operator’s identity. [#33]
· Very illustrative presentation about profile, profiling and identity. [#39]
· Nokia Siemens idea seems good, but how can they guarantee that the customer data go to a trustworthy person? The bank manager of today could be the disgruntled laid-off employee of tomorrow selling the data illegally to somebody else. For my part, I rather trust my voice mail box for everybody to leave a message. [#48]
o 48 no business has succeeded by arguing that innovation were dangerous to its established business paradigm. [#51]
· Give customers even little incentives and they willingly hand over their data. [#63]
· In France this week a guy spent a night in jail for receiving a SMS whose content was considered ‘security threat’ ….trust????? [#60]
· Any examples of where operators have offered services using the subscriber profile mgt capabilities discussed by NSN? [#10]
· In the NSN subscriber profile example, to which extend are the subscriber data exposed to 3rd parties, and by which standardised approach? [#35]
· For the NSN concept, do we first need a regulatory framework for ‘identity portability’ so that we can churn ID providers? [#37]
· Why as a user can I not hold my data and share it with who I want, if it is so valuable why don’t I charge for it as well? [#47]
o Re: 47 THAT IS THE IDEA! The idea of the 2 sided biz model (IMHO) is that your data is valuable. You should sell it by allowing the operator to monetize it with 3rd parties then ‘pay’ you through reduced service charges and greater service coverage/offerings. [#55]
o 47, in a sense you do. You get better products 0n Amazon by giving product feedback; on Blyk you get free minutes and data in exchange for insight and data. Payment from users is not always monetary (it is sometimes data and information) and, likewise, payment to users for their ‘data’ will not be monetary but through increased value. [#61]
· Further to Werner’s question of who owns the data — who do operators give access to sensitive data? Would customers like it if they know that the carrier may allow their BSS vendors to manage call information or data event records? Do carriers have access to the source code to verify integrity of mission critical apps? [#52]
· Private information is valuable. Why do not think share the value of private date with the consumer? [#56]
· Today – what percentage of Amazon’s data it collects is used to improve user experienced and increased engagement + purchasing? [#59]
· Access to data for Telco 2.0 is regional/legal/societal dependent. Operators need a good PR effort to allow opt in for customers to fuel the two sided revenue model. [#66]
· Maybe there could be a bureau operator that can abstract the customer and all related data on behalf of the operators … where people that are willing to be Telco2.0 involved can opt in. [#68]
· Bravo Werner. Telcos need to think about the customer first. What Telcos facilitate the option for customers to send data to 3rd party for monetization opportunities? [#69]
· Isn’t Telco 3.0 a free connection to the network in exchange for shared consumer behaviour data? [#71]
o 71 could see Google trying this in 5 years time if the cost of Telco network is driven down sufficiently. [#75]
· People are scared by poor data sunk into Telcos that they have no control over. When is data opening up to cleansing going to happen? [#73]
o Re: 73 the data is already there. [#76]
There was plenty of interest in the COLT presentation on the Cloud…. but also plenty of cynicism again
· Colt just showed what Telco 2.0 is all about. That model enables innovation thru cost reduction and state of the art architectures. Why don’t operators run their back office in a similar approach? [#13]
· Is Colt’s cloud real yet and have COLT sold anything yet? [#16]
· Bravo Colt. If it exists it is a model as it should be. They sign up for the SLA and they lay forth a bed of enablement. [#19]
· If Colt has that architecture in operation, they should be running the back office for other Telco operators. [#34]
· Relative to Colt’s view – what about Salesforce.com who seems to have pulled this off on their own. [#29]
· Does COLT use its own cloud services? [#38]
· Where the incentive is for existing OSS vendors who have long terms managed services contracts to innovate and move to Telco 2.0 architectures (e.g. SaaS) like Colt? [#28]
… although whether COLT’s vision (or Amdocs’) is as advanced as that presented by Amazon was doubted
· Amazon ‘eats’ their own cooking (their cloud) and COLT? [#14]
· What is colt’s business model (price structure) and how does this differ from Amazon? [#8]
· How does Werner really feel about legacy telecom operator back office Telco environments? [#30]
· How does Amdocs view Colt’s model? [#23]
· How would Amdocs view Colt and Amazon models? Seems much difference with their approach. [#7]
Feedback: Carrier-grade – what is it worth?
But do end users really think “carrier grade” is important as the operators? Will they pay a premium for it?
· I’ll believe in ‘carrier grade’ when enterprises can get SLAs for mobile coverage. [#45]
· Is carrier grade not shifting now that many users have several alternative means to perform a communication? [#57]
· Seems the industry still have major problems in freeing themselves from network drives. [#22]
· Carrier interoperability and collaboration could solve many network issues. [#24]
· From what I can gather there has been talk about all this for some time. Based on this the Telcos have failed to transform while other companies have. Surely Telcos will change too late to take any true advantage of the possible opportunities that exist today and other companies will take advantage. Why don’t Telcos focus on their core competence – the network? [#36]
o 36, for the simple reason that the value of the network itself is in continuous decline. Got to look elsewhere – make money because they own network not through the network itself. [#42]
· These arguments are the same ‘net head’ vs. ‘bell head’ arguments that occurred in the late 90’s. How do we resolve the issues? Telcos clearly are behind and are insisting on making everything ‘carrier grade’ at extraordinary costs? Doesn’t seem very 2.0, does it? [#41]
There were also various other comments about the enablers of the two-sided platform, and the consequences of personal device proliferation:
· Cascading SLA’s is old stuff, nothing new. [#50]
· XaaS is very valid a concept, and so is cloud computing/storage/network, but these are two different things – XaaS may or may not be using cloud principles and technologies. [#65
· Simplification of capabilities available for third parties. [#27]
· What is the highest priority issue for Telco 2.0 enabling architectures that does not exist today? [#21]
· OPEN standards vs. licensed software? [#32]
· 5 B user…. 5o B devices….customer data from all these!! [#26]
· How many people have more than one mobile device? 5 are extreme. [#31]
o re31 – in Europe, probably about 1.4 mobile devices per person, and >2 in Italy. Verizon in the US has suggested that ultimately 4+ is not implausible. [#40]
Participants’ “Next Steps” Vote
Participants were asked “How well current industry activity around technical architectures supports the development of Telco 2.0 business models?”
- Very well – the industry is shaping up well for delivery and new business models.
- Good start but more needs to be done – major building blocks are missing.
- Lost cause – the industry will never deploy the capabilities for new business models
Since the development of broadband access, the Internet world has recognised that customers can have many, dramatically different roles and attributes, needing specific functionality, preferences, and user profiles. Operators are in a unique position in that they have a fuller picture of customers than any single website or retailer or service provider. Several have already recognised this, and a number of vendors are offering scalable platforms which claim to be in line with the current EU legislation on data protection.
Marc Davis, Chief Scientist, Yahoo! Mobile: ”Data is to the information economy as money is to the economy. But there is a missing infrastructure – because there’s no user interface for this data and what is the equivalent of a bank for this data – who looks after it?”
But as well as user profile data, the 2-sided business model requires on-demand response from the network infrastructure. It will not matter whether it is the network or OSS/BSS/IT element that is breaking down – customers won’t care, they will just find the situation unacceptable. Both the network and IT elements must work together to deliver this. Operators are moving in that direction organisationally and structurally.
Telco 2.0 expects that this will result in new implementations of control & monitoring systems such as Resource & Service Control Systems (RSC). As services are the key business drivers, the opening up of the walled gardens is changing the service delivery platforms quite rapidly, as most new applications are centred around apps stores, mash-up environments, XaaS environments, and smartphone Web browsers, etc. which do not demand a traditional SDP or SDF. In addition, enabling services are becoming an essential element in operators’ core products. These enabling services will, in the future, allow operators to monetize their network assets.
These enabling services need a framework, which is highly flexible, agile and responsive, and integrated with the features defined by NGMN. While not all these points are implemented yet, there is increasing understanding at the operators, upstream service providers, and regulators that this new phase, opened up through the 2-sided business model, represents a historic opportunity for all members of this ecosystem.
Marc Davis, Chief Scientist, Yahoo! Mobile: ”What if we had new, industry standard terms of service under which users owned their data?”
Before the technical details can be finalised, of course, business models need to be scoped. However, the major technical areas discussed above are focal points for technology development. In the short term, Telcos should:
- Build up a logical semantic database as preparation for database integration;
- Include migration from 2G and 3G and backwards compatibility in LTE tenders;
- Prepare a user profile database;
- Reduce the number of OSS/BSS systems;
- Develop real-time responsiveness in OSS/BSS systems;
- Separate the control and data planes, separate services from transport;
- Implement and deploy an RSC system as a multivendor abstraction layer.
In the longer term, operators will need to:
- Integrate the network and IT elements of the on-demand infrastructure;
- Set up a full user profile with privacy protection and more granular information;
- Integrate provisioning, activation, network and bandwidth management, and policy enforcement;
- Recognise that Web-based service environments will overtake the SDP;
- Develop a collaborative approach to multi-vendor app stores.