Life Cycle and Portfolio
Management
A BoF Session on Product management at the TNC2004 9.6.2004
(notes on the session by KH, HK, PL (Funet/CSC, Finland))
1 Topic
The BoF was called to tackle three different aspects of product management:
1) New product development
- how the development processes are being managed, and;
- how the quality of new services is being guaranteed
2) Product life cycle planning
- the time that the service will be offered; or
- the criteria under which the service will be offered; or
- checkpoints for the offering of the service
3) Product portfolio management
- what is the proper set of services for the chosen clientele
These questions have become important in the NRENs for various reasons:
- services have been quite often offered without contracts or any service description
- there is a growing need to define services
- there is a growing need to be flexible to customer needs
- resources don't grow and therefore reallocation is a must
- killing services is a pain.
2 Participants
CARnet (Croatia)
- Zlatko Jelaèić (zlatko.jelacic@carnet.hr)
Funet/CSC (Finland)
- Kaisa Haapala (kaisa.haapala@csc.fi)
- Harri Kuusisto (harri.kuusisto@csc.fi)
- Pekka Linna (pekka.linna@csc.fi)
GRNET (Greece)
- Ζαχαρόπουλος Δημήτριος, Zacharopoulos Dimitrios (jimmy@ccf.auth.gr)
- Δημήτριος Αδάμος, Dimitrios Adamos (dadam@ccf.auth.gr)
HEAnet (Ireland)
- Ann Harding (ann.harding@heanet.ie)
RedIris (Spain)
- Diego López (diego.lopez@rediris.es)
Rhnet (Iceland)
- Jón Ingi Einarsson (jie@rhnet.is)
Surfnet (the Netherlands)
- Maurice v/d Akker (maurice.vandenakker@surfnet.nl)
- Sandra Passchier (sandra.passchier@surfnet.nl)
- Elise Roders (elise.roders@surfnet.nl)
- Jacques Schuurman (jacques.schuurman@surfnet.nl)
- Ton Verschuren (ton.verschuren@surfnet.nl)
Switch (Switzerland)
- Urs Eppenberger (eppenberger@switch.ch)
- Lars Skogan (lars@uninett.no)
3 General pressure: funding problems
Almost all participants were reporting either cuts in the centralised funding or difficulties in covering the costs with the fees collected from the customers/users.
4 Life cycle planning - service retirement problems
In the NRENs there are no systems, nor culture for tackling questions of this sort. The necessary efforts in the beginning of the life of a service are typically recognised but not the ones in the end of the life cycle.
In a few NRENs there is an effort to systematise life cycle planning, in some a decision is made upon the delivery of individual services every year.
Need to shut down a service:
- service no longer supported (and there is a need to maintain good quality in the services because of image reasons)
- service is not scaling well and the maintenance/support demands a lot of resources
Questions relating to shutting down a service:
- what is the limit for people objecting (when do you not need to care any more)
- it takes time to be gentle and this takes resources
- how to find out the actual need for a specific service (what is
the figure that tells the need)
5 Portfolio management - how to meet the needs of the community
Finding out the user/community/customer needs
- individual customers wish for new services especially when there is no direct (‘favours’)
- market pull can be sought out with the help of user panels
- sniffing the air and making educated guesses about the user needs
- the people that pay your expenses, they tell you what to do (but this is mostly only the ministry that gives extra tasks - not the HE community)
The portfolio cannot be defined (only) based on technological
possibilities:
- Switch-model: mission - clients - technologies - political and economical environment -> strategy (every three years) -> service portfolio
- SURFnet observation: well-defined new-product -launch system does not guarantee a consistent portfolio that meets the needs of the community: need to find the balance between technology push and market pull
- there are strategic papers where the tasks/missions of the NRENs are being defined
- some NRENs negotiate with the local ministry about the services given
Roles of different actors within the NREN in relation to the portfolio
- who decides what and when
- not everybody defines an activity to be a service with the same criteria (‘it is only development work / co-operation’)
- there is a move from backbone to middleware and applications
- there is a move from one-way service delivery towards consultation and co-ordination
- more difficult to get a picture of the whole, more difficult to
tackle the needs with the customers (‘would you feel that you need to
be co-ordinated?’)
6 New product development - how to launch well-defined services
Switch: AAA-infrastructure service was well defined and different from others. Steering comes from the top down. Also took a very long time that is unusual. The problem now is to get the HEIs to utilize the service and to make it economically sustainable.
SURFnet: A new fairly well-defined framework for new product development: technology assessment / go or no-go / proof of concept / go or no-go / piloting / go or no-go / customer support etc. -> product launch
- the problems seem to be on the side of portfolio management: what are the criteria for creating a service at all?
UNINETT: Users may think something is a service, when it's not, and in that sense product launch can get out of hand.
CARnet: The problem is pushing the users into using all the services.
7 Contracts and billing
Some of the NRENs do not have service contracts with the HEIs
- users on top and on the bottom: this takes care of the steering and supervision (HEI representatives are in the board of the NREN)
A change in technology may change the nature of the service so that the old contracts are totally out-dated (e.g. the move to optical networking may cause this)
Users may depend on the service although not much has been promised
- in this situation there are growing risks for both parties when responsibilities and their limits have not been properly articulated
- in this situation it might be rational to express the limits of the service: to make a contract in order to avoid conflicts, not just in order to win them
Billing
- NRENs don’t seem to bill accurately from each customer all the small services that they have been using
- not all the NRENs have SLAs/SLSs
- instead of SLAs a redundant infrastructure
- critical services built on the IP-connection, for example phone
systems; higher demands on reliability and overall quality
8 Future co-operation
E-mail list
- Pekka will ask Terena to start a mailing list for this purpose
- the participants will be added to that list
- at least in the next Nordunet meeting in Spitsbergen, maybe also before that
- more discussion on the mailing list