430 likes | 522 Views
Some aspects of ICT-strategy. The lecture is based on my experience as advisor for the top management of Telenor. Jan Audestad. My background Theoretical physics, NTH (NTNU), 1965 Electronics design (radio relay) – Nera Bergen, 1967-71)
E N D
Some aspects of ICT-strategy The lecture is based on my experience as advisor for the top management of Telenor Jan Audestad
My background • Theoretical physics, NTH (NTNU), 1965 • Electronics design (radio relay) – Nera Bergen, 1967-71) • Research Televerket (Telenor) Research Establishment, 1971-1995 (senior researcher, research manager, chief of research) • Satellite communication: North Sea, Svalbard, ships (Inmarsat), land mobile (ESA) • Maritime terrestrial services and data communications to ships • Land mobile: GSM – vice chairman of GSM responsible for network aspects • Intelligent networks – vice chairman of ETSI group • Distributed processing • Information security • Chairman of several specification and research groups in ITU and EURESCOM • Advisor for top management of Telenor, 1995-2009 • Techno-economic strategy • Establishment of Telenor and shaping of new businesses • Advanced education programmes (postgraduate master, PhD, courses in techno-economic studies) • Machine-to-machine communication and advanced platforms • Adjunct professor in distributed processing NTNU, 1993 • Adjunct professor in information security at Gjøvik University College, 2002-2012
Content 1. Stakeholder analysis 2. Organizational analysis 3. Value creation analysis 4. Cash flow analysis 5. Core competencies analysis 6. Scenarios 7. Prospect theory and anchoring 8. Real options
Wikipedia: Strategyis a high level plan to achieve one or more goals under conditions of uncertainty. • Strategy is long-term – tactics is short-term • Generally, strategy is a diffuse phrase – may be anything, and is related to most modern buzzwords used in business management • I will present some methods that may be important to investigate in order to • Understand the business and its environment • Identify uncertainties and what they may signify • Support decision making • Identify what may go wrong • Perform in-depth analyses – you don’t understand the problem befor you have penetrated deeply into it!
ICT stakeholders Support Security Platforms Sourcing Cloud Installation Maintenance Insurance Etc Technology suppliers User terminal Databases Servers Fibers Base stations Satellites Mainframes Platforms Etc ASPs Platforms Resellers Distributors Integrators Content suppliers Web sites Robots Machines Infrastructure Censors ISPs Internet Mobile Satellite Dark fiber Virtual operator (VO) Users People Machines Software Policy makers, regulators, interest groups
Some BIG strategic problems • A user may be a person, a machine or software!! • Categories: • Person to person (e-mail, speech, SMS) • Person to machine (web) • Machine to machine (censor networks) • Internet of Things (IoT) • Size: • People: 7 billion (109) • 1 megabit per day 100 Gbps • Machines: more than 1 trillion (1012) and increasing fast • 1000 bits per day per unit 10 Gbps • Strategic challenges: • Overlooked as new source for communication • Extremely price sensitive (huge systems, small margins) • One order of magnitude better reliability • Autonomous recovery • Extreme scalability • Very sensitive to information attacks Focus of ISP Usually not included in strategic plans
Example: sensor network: this is how it looks like External service providers Application operator Service provider/user Internet/ leased line Internet/ leased line Telcos Installation – most expensive Platform Platform services Platform owner - key player Transport network Internet/leased line Internet/leased line providers Back-end system Access network Internet providers/ Mobile operators/ Fixed radio operators . . . Terminal/gateway/reader T/G T/G . . . LC LC Local communications Front-end system Installation. maintenance . . . S S Sensor, RFID etc. S
Sensor systems • Smart grid > 2 million in Norway; > 100 million in USA • Chips for car identity and toll road payment: > 1 million in Norway • Supervisory cameras: several thousands • Burglar alarms and protection: several millions • Detectors in infrastructure: several millions • Position monitors: several thousands to several millions • Monitoring cattle: several thousands • The market is enormous and complicated! • Challenge number one: scalability and rate of growth
Effect of scalability: Two examples • Case 1 • # of sensors: 0 – 100 million in six months • NOK 100 per censor NOK 10 billion • NOK 1000 per installation NOK 100 billion!! Prior to income – impossible!! • Case 2 • # of censors: 0 – 1 million in six months • NOK 100 per censor NOK 100 million • NOK 1000 per installation NOK 1 billion! Prior to income – possible but difficult • Installation cost is the major cost component! Often last thing to take into account in the economic planning. • If 1% must be reinstalled: • Case 1: NOK 1 billion • Case 2 : NOK 10 million • For many managers, 1% is a small number! It may destroy the company!
Schematic evolution of ICT market in a developed country Sensors Size Mobile traffic Now Fixed internet traffic Fixed telephony – will disappear (2017) Time Revenue Sensors? ? Time
New revenue only from new businesses • Machine to machine (censor networks and the like) • Strategically difficult (Cinclus) • Large number of devices, cheap devices, expensive installation and maintenance, critical communication cost: • With 1 øre per message, the cost per year for an alarm service is kr 10,000, with 10 øre it is kr100,000 – enormous price sensitivity • New thinking in platform design (scalability, speed of growth) • Information security and infrastructure protection • Closed networks for infrastructure • Secure business networks • Distress communication
C4 divides ICT business into 4 industry categories • Any owner of information • Newspapers, publishers, cinemas, opera houses… • Time Warner (TW) • Disney etc Content • Computer manufacturers • Providers of resources • Google, Facebook, cloud operator, server stations… • Outsourcing Computer Communication • ISPs in general • Particular operators • Mobile, internet, satellites, local… • Iridium, Teledesic Consumer electronics Computers, smart phones, mobile phones… Any equipment containing CPU Manufacturers of RFID Software
C4 strategies: • Alliances (1995-2005) between communication operators – disappeared because of internationalization – cooperators became competitors (Bell companies in US, three big alliances among telecom operators) • Buying: • Communication operators buy content (AOL buys TW (00) TMW(05) TW sells AOL (09) TW buys cable systems and TV networks (10) • Content operators enter into communication (Schibsted– internet, TW) • Consumer electronics into communication (Teledesic (IP in the sky) –never realized, Iridium failed (handheld mobiles and low orbit satellites)) • Communication operators buy computer (to support outsourcing) • Problems Content 1 Content 2 Content Less business Binding Binding Communication 1 Communication Communication 2 Less business
3C Time Warner Facebook, CNN Content ISPs Google, American online 1 5 2 6 7 Amazon Connect 4 3 Commerce Another way of looking at it – another consultant
3. Value creation Most businesses in ICT are value networks in the definition of Stabell/Fjeldstad. Explained in TTM4165 ICT, organization and markets Repetition in next two slides.
Bank Newspaper Depositor Borrower Reader • advertizer Mediation Mediation Publisher Telecom Facebook Author Reader User User Mediation Mediation
ISP mediates between users and ASPs • ASP mediates between information providers and users ASP ISP ISP ISP Info ASP Parallel ISP1 ISP2 ISP1 + ISP2 Series Coopetition
Value network business: mediation – connecting users • ISP offers connectivity (network, internet resources, connection management) • ASP offers connectivity with information provider (search engines, web servers, data storage…) • Information provider produces content (newspapers, brokers, Facebook, Amazon…) • Business characteristics • ISP, ASP, information providers are value networks having marginal costs 0 • ASP and information providers may “sell” their product for nothing and get their income from different sources • ISP cannot identify the services provided by the ASP and information provider as shown in next slide
Transport layer schism • Transport layer (TCP/UDP/SCTP) is opaque: • The ISP cannot determine which service is active, the actual data volume, and the duration of the connection • Almost all services on ports 80 (http) and 443 (https) • Streaming connections are set up on ports 80 and 443 : after set-up, conversion to UDP or SCTP or application layer protocol • Hence, the ISP cannot differentiate charges for different services • Ending up with only subscription charge plus addition for huge volumes • Revenues almost independent of traffic • Investment cost increases with traffic • The big volumes are from big web servers and search engines; however, initial request from user • Charging the web sites and search engine will destroy their business – if each access to Google costs money, few will use it • Charging for incoming traffic is nonstandard Revenue Investment cost Capacity of internet
Integrated company ASP ASP Other sources Running cost ISP ISP Connection Subscription Remuneration Task: identify all cash flows and internal expenditure and see if > + condition: Internal costs is satisfied. Strategy: increase in-flows, reduce out-flows; what is the impact of owning an ASP? Will it increase the in-flow? Example: premium rate services offered by ASP owned by ISP (Telenor Link)
Cheap Cash flow enabler • Service competition • Telephone • Mobile (Djuice…) • Data • TV • Premium rate • Reselling Integrated company ASP In-flow Expensive • Network competition • Telephone network • Mobile network • Data network • Cable network • Broadcast ISP Out-flow • Enabler • E.g., subscriber line • Example: Telenor was to be split up into service provider and network operator (1998). Never done because • the company’s revenue was generated/enabled by the network operator • competitive price pressure on service provider, no competitive pressure on network operator • main in-flow via service provider • enabled by network operator
Method: • Identify the company’s competencies and incompetencies (what it can do and what it cannot do but should be able to do) – compare with the branch in general – what makes some companies better than others • Yellow pads and brown paper exercise – get as many ideas as possible on pads , then group, regroup, find patterns… • What is a core competency? • It is not easy for competitors to imitate • It can be re-used widely for many products and markets • It must contribute to the end consumer's experienced benefits
Market value Core competencies High Renewal of product (technological excellence) Market knowledge Low Low High Shaping the future Reshaping the business: Competencies: now future
Benefits: • May sustain superior performance over a long period • Good to know the company’s strengths and weaknesses (together with SWOT) • Fallacies: • Myopia with regard to evolution • Overestimating own skills • Overestimating sustainability of products
Companies with huge core competencies that failed • Norsk data: • Core competency: one of world’s best on tailor made mini computers • Myopia: did not see the coming of the workstation and the PC • Result: bankruptcy (1990) • Facit: • Core competency: number one on mechanical calculators • Myopia: did not see the evolution of the electronic pocket calculator (1971) • Result: out of business (1971) – but survived as a company • Ericsson mobile terminal production: • Core competency: radio system – acquired by buying SvenskaRadioaktiebolaget (1982) • Merger with Sony (smartphones) – new core competency: software platform • Sony has taken over the mobile product line, while Ericsson is no longer in the mobile terminal market (radio technology has become more or less trivial)
One problem is disruptive changes: • Being stuck with obsolete competencies • Building up new competencies (research, education, business understanding) • Anticipate changes and manage change strategies (first mover – early follower) • Examples • Transistor 1948 : vacuum tubes transistors (1960s) – changed electronic industry. Survivors: those who saw that the evolution was irreversible and changed quickly to transistors (e.g., Philips transistor radios in plastic casing, Nera Bergen radio relays). Slow change (more than 10 years) • Microprocessor (mid 1970s): changed computer industry pocket calculators , PC and enabled automatic mobile systems (NMT). Very rapid change. • Web Internet (about 1995): came as an enormous surprise upon telecom operators – did not anticipate implications for the economy of telecommunications. Very rapid change. Chaotic market – dot-coms (2000) – enormous expectations, no business • Smartphones (2008 Android): changed the mobile competencies from radio and channel coding to computer technology and programming. Very rapid change. • Optical fibers replacing cables and satellites. Very slow change. • Important: Rapid changes be first-mover and get initial market shares and recognition! Keep up with evolution in the field (Nokia did not!) (Google and Apple did)
Telenor scenarios 1995: • Gloom and doom (Norwegian recession) • Nightmare (global recession) • Blinding light (10% market increase per year – telecom bonanza) • One arm tied on the back (10% market increase but heavily regulated market) • In 1998 followed up by an enormous system dynamic model for network operation, confirming the results of the scenarios • Method: • Analyze drivers and problems in the market (yellow pads and brown paper) • Determine plausible evolutionary paths and write plausible scenarios • Evaluate economic and other results for each scenario • Find a strategy that is good for all cases • Fallacy: does not take into account uncertainty!
Examples where scenarios go wrong • Belgium: • In all scenarios, nuclear reactors cheaper than power stations run on gas • With uncertainty concerning the energy requirements, gas is cheaper since nuclear reactors cannot be regulated to produce variable power outlet (overproduction of power or insufficient supply) – not visible from scenarios • Telenor: fixed vs mobile (analysis done around 1998) • In all scenarios fixed is cheaper • With uncertainty, mobile is cheaper. Reason: Scenarios are developed for a anticipated number of users and does not take into account that the anticipation may be wrong: • Remedy: use stochastic optimization: much more difficult! Fixed Cost Mobile Number of users Scenario 2 Scenario 1
7. Kahneman and Tversky (KT): Prospect theory and anchoring See: Kahneman, D. (2011). Thinking Fast and Slow, Allen Lane 2011
Prospect theory: • people are risk-averse when it comes to gain: we go for a certain gain rather than an uncertain even if the expected gain is larger • People are risk-seekers when it come to losses: we will not opt for a certain loss if we may choose an alternative where there is a significant probability that we may loose nothing even if the loss is statistically larger • The two alternatives may be identical but only phrased differently! • Example (KT): If nothing is done 600 people will die. Two treatments A and B. Two identical formulations with different outcome: • With treatment A, 200 will be saved, and with treatment B, everyone will be saved with probability 33% and everyone will die with probability 67% - 75% opted for treatment A • With treatment A, 400 will die, and with treatment B, everyone will be saved with probability 33% and everyone will die with probability 67% - 75% opted for treatment B • Economic utility theory (everyone is rational) is not always correct! – resulted in Nobel prize in economy for Khaneman (Tversky was dead)
If a survey in a company with 10.000 employees showed that 90% of the employees said that they followed the new security policy, what would you then, as responsible for the IT department tell the management • 90% of the employees follow the policy, - this is good! • 10% does not follow the policy, - find a better policy! • 1000 of the employees does not follow the policy? – find them and fire them! • Big numbers are better than small numbers: In Moss, the politicians had to decide whether plastic waist should be returned based on the two identical claims: • This is very important because this corresponds to the CO2 pollution of 30.000 cars • This is insignificant because plastic only amounts to 0.3% of the total CO2 pollution in the region • Of course, the first claim won!
Other results: • People are overestimating own skills • 70% of students regarded themselves to be above average in leadership abilities, only 2% regarded themselves below average (KT) • 80% of all drivers regarded themselves to be better than average, only 5% below average (Norsk Gallup) • Anchoring: • The first proposal tends to survive - it is always hard to convince anyone that this proposal is poor or even wrong • Prediction of the future is usually too much biased on the past • Law of small numbers: • The uncertainty of small statistical samples is larger than that of large samples • In a class of (say) 30 students any distribution of the grade is “normal” – just use the chi-squared test to see that any distribution over30 items is a normal distribution (or any other distribution) with 95% probability
Financial options 1. Financial call option: You have the right but no obligation to buy an asset for a given price (the call price) before or at a certain time. This right costs you a premium paid to the seller. Ideally, the premium should statistically equal the loss of the seller if the price of the asset is higher that the call price. The balance is determined based on the volatility of the price of the asset. The volatility is the same as the standard deviation. This is reasonable except that the statistics of the price of the asset depends on more than volatility, for example, the fourth order moment called kurtosis. The kurtosis is related to the tail thickness of the distribution. Because of this kurtosis risk, much trade n options have failed. 2. A financial put option is an option where you have the right to sell an asset for a given price.
Real options • Real options are assets such as projects, factories, firms, machines etc • Real options has to do with flexibility – what can be done to increase profit or reduce loss? • Common options: • Expand or contract or status quo? • Initiate or postpone? • Abandon? • Sequence products/operations? • Product line flexibility (flexible market adaption) • Process flexibility (exploiting different production methods) • Intensity (flexible output)
Decision parameters • Current value of project • Uncertainty (volatility, sometimes also kurtosis) • Strike price (investments required) • Option term (the time during which a decision must be taken)
Decision tree . . Possible outcome . . Probabilities . Now . Time when decision is taken