Technology Management

Vasos Panagiotopoulos

Objectives

u How to manage technology in business

u Will learn all aspects of management as applied to technology

u For scientists/engineers as well as managers/financiers to learn how to deal with technology in business.

– Team up with someone who complements not emulates your background

Technology Management

u Change the rules

– Sometimes unexpected results

u Risk & Uncertainty

– Need for integrated coordination

– Need for combinatorial diversity

u Complexity (ie RFPs, matrix management..)

– disempowering

– arbitraged, not obeyed

Technology
Management

IS

u Minimising steps for robotic assembly then making it even cheaper by hand.

u Designing drug molecules then making them.

u Using mammalian cultures instead of just fermentation.

ISN’T

u Replacing typewriter with a PC, without changing work habits.

u Putting the same forms on a computer screen.

u Cross-testing every possible resulting molecule the way you looked for new antibiotics

Groupthink
doesn’t question dominant function
[Arthur Rock HBR Nov 1987]

u Big Firm : Finance dominates

– late PLC mentality (cost cutting)

– ignores paradigm shifts

u Mid Firm: Marketing Dominates

– Emphasis on incr. sales

– Ignores scale-up, inventory, dim’g returns

u Small Firm: Tek Dominates

– Thinks biz skills unintellectual

– Ignores timing of payables and inventory

– Emphasis on discovery

Industry Structure

Product vs Process
Research & Design

u Freeze design as late as possible

u Include order-winning goals early in process & rewards

u Develop concepts, processes, platforms and product families before developing products (ie Intel 80x86).

u Speeding a strategic product to market will only accelerate commoditisation

u Scale-up capability before demand builds

u Improving processes keeps your competitive advantage.

u Knowing when to stop & redesign process model all over again

Order Winner/Qualifier
Decision Modeling
(Interorganisational)

u Tek Sales or Licensing -- same method

– Tune perfrmc measures to customers & strategy

u Search the decision tree & the org chart

u Relationship or Attribute selling?

– Relationship might be heuristic abbrevn of attributes

u Who REALLY decides/influences/approves/uses?

u Is it price, quality, time, or some hidden agenda?

u Central buying or tek decentalisation

u Stage of budget/procure/design cycle, gating, benchmarking

cf HBS-9-582-117 ,9-489-084

Chosing Financing Vehicles

u M&A: get quickly, might lose people

u Licence: internal usability/fit

u JV: share, learn but conflicts

u VC: access but entrepereneurial

u Option: (warrant): unproven, modest risk

u Grant: expand R&D, poor incentive

The Licensing Deal

u Research Presentation...........3 Months

u Deal Structuring..................1 Week

u Deal Pricing.........................1 Month

u Due Dilligence.....................3 Months

u Termsheet/Ltr Intent............2 Weeks

u Development Visit...............1 Month

u Corporate Approval..............2 Weeks

u Legal Negotiation..................2 Months

Typical Royalties & Strategies

u 25% of Profits or 5% of Sales

u 1-4% of Sales: Semiconductors, Chemicals (Chem per kg, too)

u 5-15% of Sales: Spec Chem, Drugs, Med Devices

u If you publish, you perish (novelty, prior art)

u Maybe the guy in Sweden wants the Korean to make it but would get billed double royalties. Allow collab betw your licensees.

u All JVs to learn.. are all temporary

u License the demo in case a rejected vendor takes it to your competitors.

u Cash flow valuatn:(mkt shar change,royalty relief, resid income,cost savngs)

u DEFINE: Strategic product identity criteria, royalty base, non cash consideration, income exemptions (tax, ship, commiss), min perf rqmt for licensee, triggering & terminating events, derivative products tol modfcns, field of use, need to know, inspxn of records, tol currency risk, tol sales-rel disclosures, indemnification for breach

Hi Tek Finance
(for example)

uBeta = 3 software, 7 biotek, not 1

uP:E= 40 not 12

ur= 20% not 5%

beta(x)=covar(x,market)/var(market)

Venture Capital Method
And Overvaluation in the “New” Economy

u INTEREST= REAL (4%=marg prodxvty capital)+RISK+INFLATION

u = RISKFREE (V LONG T 4%)+BETA*RISKPREM (V LONG T 4-6%)

u = ([1,1.5]+beta)*4%

(Cash flow, Value Added, Liquidity and risk make capital more costly in early ventures)

Finding the right variables or proxies is rarely done right. Which rate? Which growth? (Mystical obfuscation of no dividend hi-tek and tax effects: Miller Modigliani JB 1961; Farrar Selvin NTJ 1967; Brennan NJT 1970; Miller Scholes JFE 1978. High future expected growth rates might create bubbles unanchored by actual dividends.)

u P:E=(1/(r-g)) =NPV([ONES],-4%(1+beta), neg=> forward looking

u =(1-growth/(ROI(1+beta))/(WACC-growth))

u =(ROE-DivddGro)/(DisctRt-DivddGro)/ROE

HBS VC Method 9-288-006 p12 footnote 5; HBS Valn Models

9-281-067 p5; Copeland, Valuation, Wiley, 1990 p79

Options Pricing
(Contingent Claims Analysis)

u Hi r diminishes NPV input from distant earnings

u Discounted Binomial ProbabilisticDecision Tree

– matches milestone structure common in tek xfr

u Merck & J&J use Options for R&D budgets

(Lewent HBR 2/94; HBS Conting Clms 9-286-114, Trigeorgis, Real Options, MIT.

Merck worked but not J&J because of trust – managers didn’t arbitrage inputs)

u Crudely with Black-Scholes

– S=PV(proj fin val); X=Launch Cost; t=time to launc

– r=WACC for proj; Dividend Yeild = 0

– sigma=est from pul-trad compar proje (subtract for attrbn)

u Russ Parr of AUS says can’t measure Excercise price

u Volatility also hard to Guage

u Complexity is disempowering

u Look good if suggest but not if implement

– Hire an expert to do it; for strategic initiatives only

Hi Tek Proj Management
Intellectual-Process PERT/CPM

u Teks are lousy estimators of cost and time

– they simplify in order to think thru

F good for tek

F lousy for biz

– rough fudge is 2.3 times their estimate

u Gurus do 80% of work for team

– gurus overestimate not underestimate

– never hesitate to grant gurus overtime

– go out of your way to keep them for life

Tek Life Cycle
Sales=a*tanh(b*t)+c

u Idea, plan, identify customers

u Prove concept, feasibility, id enabling tek

u Design, protoype, patent

u Validation, test

u Scaleup, commercialise

u Build up market share

u Maintenance, Creative destructn/substitn

Pr Nelson Fraiman’s
17 Step Tek Change
(Columbia B9811x95,8827x96)

INFORMATION TECHNOLOGY

u Info is dear, data is cheap

u Longterm firm-wide intercompatibility

u Vendor reliability or source code in escrow

u INVARIANT:1 man-hr per line of code (fully debugged, documented,tested, maintained)

u Dominant design (microchannel vs AT)

u Other issues: data-flow, work-flow, requirements, prototype, recovery, distrib contention/integrity..

Regulatory Effects
eg Drug Discovery & Approval $2.5 BN
Dollars below are minimal prepartnering budgets, valuation reward
Difference is because failures aren’t counted and attributed

u Discovery (5yrs) (10,000 compounds; Screening leads, rational synthesis)

u Preclin (lab & animal, pharmacology, toxicology) (6.5 yrs) (250 lead candidates) $.6MM

u IND (Investig New Drug) applxn; Corp IND vs Physician’s IND

u GMP (Good Mfgg Proc) pilot plant built & subseq scale-up

u Phase I (25 non-ill voltrs for safety & dosage) (1.5yr) (80% of 5 candidates pass) $.7MM, $25MM

u Phase II (200 vol patients for efficacy & side-effects) (2yrs) (30% pass) 1.4 MM,$75MM

u Phase III (2500 vol patients to mon long-t rxn) (3.5 yrs) (80% pass) $7MM,$250MM

u NDA (New Drug Application, 100,ooo+ pages!)

u FDA Review & Approval $.1MM

u Post-Marketing & Phase IV (1.5yr)

u Orphan Drug - fewer than 200k potl pt 7yr exclusivity

u Dx is right only on third try, fouls up trials, cf openclinical.org

u Adaptive Max Likelihood trials

Where to get more information

u These Slides at

biostrategist.com/BzTekMgt.pdf

u Innovation Management Seminar

u IRI/RTM 1550 M NW #100 WDC 20005-1708

u Samani Marions Panyaught Consultancy