Stanford GSB Sloan Study Notes, Week 3, Autumn quarter
The highlight of this week was of course the Friday our Sloan class owned the GSB courtyard with our little dance act. It took weeks of preparation, late nights and sweaty trainings for many to make Project Chai happen. It would have been too easy to be and remain sceptical of the entire venture from start, given how YouTube is overflowing with Gangnam style flash mobs, but the sensation of the entire class an many of our partners going through this was just amazing. (Mind you, this is not a high school, but perceptionally quite high competition and serious workload business school we’re talking about). Thank you, Herbert, Hans, Cherie, Tracee, Jonathan, Gitanjali and everyone else for pulling people through this.
(As a side remark, keeping this video online is a terrible experience – Youtube is blocking some mobile viewers, Vimeo did a full takedown for a while of our mobile friendly version of this video… We’re trying to talk directly to music rights owners now, but how on Earth are flash mob videos usually distributed?)
On academic side, covered in this issue:
- Why not drink Diet Coke in Las Vegas?
- How some penny-pinching retail operations can grow bigger (and more profitable) than most countries.
- How to fire a CEO.
- How GAAP accounting discretion in used and abused in public company financial reporting.
- How charisma can be broken into components and trained.
- Why reward people with stuff as opposed to money?
- Why would you still want to correlate the size of codebase to the value of software IP, despite of obvious pitfalls?
And here on to the full notes:
GSBGEN378 – Decisions About the Future (Hardisty)
- Economic theory of self-control (Thaler and Shefrin, 1981): Self(a) is a planner and Self(b) a doer. Doer is deploying the resources to get the planner stay in line.
- (Pre)commitment must come from Self(a) because Self(b) prefers to take the wicked thing, unless constrained.
- Commitment must prevent Self(b) from acting on bad impulses, or from coming into existence at all.
- Baumeister & Vohs (2003): willpower is like a muscle: limited strength, gets tired with use and gets stronger with repeated use.
- Discount rate vs interest rate differentiation [DOC on CW]
- “You can manipulate… sorry, structure the situation”
- Defaults change discount rates:
- people more impatient for something expected tomorrow getting delayed
- more patient for an event from 2 months away expedited to happen sooner
- Sign effect: it is generally rational to take$100 now and give it as late as possible
- yet half of people want to give now to “get it over with” (reaping debt)
- for small losses discount can be even negative (prefer giving $10 now to $9 later)
- sign effects when portraying the future
- positive: accumulating pensions
- negative: massive debt piling up, global warming -> some so hard to explain that people just sign off (= super-high discount rate)
- Irrational desire for (receiving) increasing sequences: preferring to get $100, $120, $140 monthly payments
- Spread-out preference: “If you can get 3 great meals over next 1 month, when?”
- Most answer now, 2, 4 weeks – rather than “all ASAP”
- Hyperbolic discount model captures both
- Delay: now VS in a week
- Interval: one salary in a month VS four equal weekly payments
- Acts of self-control reduce blood sugar (The Physiology of Willpower: Linking Blood Glucose to Self-Control, Gailliot 2007)
- will-power literally (physically) spends energy
- flip-side: sugary drinks decrease impulsivity (Wang 2010)
- never have a Coke Light in Vegas! 🙂
- negotiate during / after lunch, never before.
- Kahneman‘s Prospect theory (Nobel-prized) introduced
- probability distortion (certainty effect): what is the value of reducing cancer risk 48% -> 47% VS 1% -> 0?
- Hedonic framing
- distribute wins (2 x $5 win > 1 x $10)
- consolidate losses (get them over with)
- mix big gains & small losses (to ease the pain)
- Construal level theory
- distant (physically, socially, psychologically) events are abstract
- the reason to sell vacation homes based on 3D pictures before building starts is not just to build up pre-payment cash-flow…
STRAMGT279 – Global Strategic Management (Roberts)
Some more details & guidelines given for our group project (in our case, build a strategy for Nordstrom’s expanding to Turkey). The aim is to learn how to learn about an unknown business and country. Examples of research items: how welcoming to FDI, GDP, income distribution, middle-class size, difficulty of doing business, labour market, competition, import needs & hurdles, cultural differences. Sourced on anything from CIA World Factbook, UNDP, Freedom House to interviews, any commercial information source in Stanford Library (Bloomberg and the likes).
- Guest speaker: James Liang, co-founder & chairman of Ctrip.
- now a Stanford GSB PhD, thesis on how FDI in emerging markets draws down salaries for educated work force
- usually mis-interpreted by governments as overproduction of university students
- actually just a timing issue
- multinationals want a degree plus experience
- in fast growth periods a seeming overproduction of those with degree, but still waiting to gain experience
- now a Stanford GSB PhD, thesis on how FDI in emerging markets draws down salaries for educated work force
- Case: Ctrip (largest OTA – online travel agency in China)
- For a culture of open discussions
- arguments need to be fact based
- for facts, everyone needs access and training to data
- (and in Chinese culture of obeying rank special encouragement / hiring filters)
- Industries more suitable for data-centric strategy
- large volume
- production of data a side-effect of service/good delivery anyway
- repeat customers (there is something do with the analysis output next time)
- For a culture of open discussions
- Specialization is a function of the size of the market (Adam Smith)
- a call center with 10 calls a day will always have a single agent serving all of them
- Further reading: Division of labour
- Walmart cases (2002, 2007)
- Of Top10 discounters operating since 1962 (Wal*Mart’s start) none survived by 1993
- “Good-size stores into little one-horse towns which everybody else was ignoring”
- Pattern of expansion: inside-out. Never jump & then backfill geographically.
- Walton: buying trip expenses should not exceed 1% of purchases.
- “Management by walking and flying around” can always look on the other side as “… by wearing you down” or “… by looking over your shoulder”
- Shrinkage: a retailing euphemism for pilferage & shoplifting
- Slow-moving merchandise => overstocking => need for deep discounting.
- Cross-docking: transfer products directly from in-bound to store-bound vehicles (zero time in inventory). 10% for WM in ’94
- Procter & Gamble as largest customer: 70 ppl stationed in Bentonville to manage products for WM alone, $3B in revenue, 10% of P&G globally
- “No superstars, WM is a company of ordinary people overachieving”
- Percentage of salary to profit sharing account, employee can choose upon leaving if to cash out or convert to WM stock.
- No regional HQs to save cost (15 planes instead). Saturday morning meetings at 7am for entire mgmt team and their friends and families combining informal entertainment with non-nonsense business information sharing.
- Cashing in on “economics of density”
- pizza delivery to dorms (versus suburbs)
- taxis in a city (versus countryside)
- Barrier of entry in retail competition = barrier of exit for those already in
- You can’t rent 100,000 sq ft space -> need to build to your needs -> high capes -> sunk cost / no-one to exit to profitably
- India had a protectionist ban on foreign ownership of retailers selling >1 brand
- GAP or Gucci OK, Wal*Mart not
- lifted this month
- When there is cost in the system it is pointless to argue who picks it up
- Always the one who can do it cheaper, charging (some of the cost) to the less efficient party
- Huge variation in retail supply models still today
- Zara’s just in time, quick cycle
- GAP’s 2 purchases a year
- If Wal*Mart raised their prices by 1%, every single person in US would need to spend $10 a year more
GSBGEN566 – Real-Life Ethics (Mark Leslie, Peter Levine)
- Saying that option-holders have nothing to loose (only upside) is wrong: TIME
- you can walk away, yes, but you never get the time in the trenches back
- Potential negative effect of options: recepients go in a holding pattern, waiting for vesting
- Guest speaker: Robert Joss (former Dean of GSB, ex-CEO of Westbank covered in case)
- In an ethical dilemma:
- Talk to someone (doesn’t matter who: just air it)
- Secure (confidential) banking in tiny places (Cook, Virgin islands, etc) so big part of local economy, real issue when stopped
- Cook Islands: 15k inhabitants, 200k tourists. 2/3 of GDP is tourism, rest is banking
- Blacklisted by OECD (after the case), now back with improved regulations
- Lawyers warning: “if banks peel the onion too far with hidden accounts, they can be held more liable in tax evasions or anything else they find”
- Whistle-blowers risk: a hero or a troublemaker, depending who is assessing
- WIIFM: deeply human “what’s in it for me?”
- A guy called Allen Stanford sentenced 110 years in $7B Ponzi schemes
- no connection to the Leland Stanford family he claimed
- university decided to turn down his tempting generous donations years before due to background checks
- Solving an ethical questions for yourself:
- Awareness: is there an issue?
- Analysis: how big are the good and bad for me, society? who and how many people hurt?
- Act: the courage to actually do something
- if you already feel uncomfortable, you’ll likely never forgive yourself if you do not act at all
- Practice and role-play complex situations in real life too to prepare, don’t stop this at school!
- Escalation of issues is often not enough
- without action & closure, you’re (mentally, if not practically) not off the hook
- Guest lecturer Robert Chess
- “As a chairman, when you get an email from a VP not via the CEO – nothing good can come of it”
- “Biotech industry sometimes moves at geological time”
- Assessing skip-level escalations:
- Explicit questions
- Implicit questions (asking directly would have unintended side-effects)
- What impression do you want to leave? (fair, balanced, listening)
- What not to do (take sides, give promises you can’t keep)
- In a situation where you don’t want to take sides, stay quiet and let others talk
- Fair asks from the escalators, to help stabilise the org
- I need time to assess and decide
- Don’t talk to others
- Don’t leave or even start looking for jobs before letting me know first so we can discuss
- CEO the most difficult person to get information on in the company
- Any probe is really a sign of distrust
- A legit reason for a board member to to skip levels over CEO: his performance reviews
- This cycle could be too long (yearly?) for any crisis management
- % of senior execs who come out well on the other side when put into a performance management program: zero
- on senior levels there is an expectation of character and competence
- hard / impossible to change those with a 3-month program, from outside
- rather make the cut faster: “Company feeds off the energy of the CEO”
- Typical protocol for CEO reviews in US
- compensation committee -> board -> CEO discussion with board / chairman
ACCT219 – Financial Accounting (Guttman)
- Companies own activities explain (rˆ2 in regression) about 2/3 of its stock price (Dechow 1994)
- on that backdrop, 0.2-0.4 rˆ2 for any influencer of price is a massive significance
- the correlation is highly dependent on the time period
- quarterly -> annual -> 4-year basis reporting
- EPS explains: 3% -> 16% -> 40%
- CFO (cash flow from operations): 0% -> 3% -> 11%
- “Special items” in income statement are unusual or infrequent, but not both
- Asset write-offs, gains & losses from asset sales, restructuring…
- Examples of both: natural disasters, irregular lawsuits
- SEC Chairman Arthur Levitt
- Famous 1998 speech The Numbers Game
- covered ways which companies are managing their earnings: cookie-jar reserves, revenue recognition, etc
**MS&E 472 – **Entrepreneurial Thought Leaders
- Olivia Fox Cabane on The Charisma Myth
- charisma = presence, power & warmth
- Presence is the opposite of trying to keep a conversation going while thinking about something else
- Little tricks to restore presence in a few seconds
- Focus on the physical sensation in your toes
- Focus on the colour of the other’s eyes (creates a deep, “soul-searching” look in the eyes as a side-effect)
- Stanford research on alpha humans – they claim space exactly as gorillas do
- Sitting on one chair, hand on another, legs stretched far out, etc
- This triggers a measurable cycle of increase in hormones that re-enforce the sense of confidence and power
- Imposter syndrom – lack of confidence that you’re true, that your fraud (being something less that you’re presenting yourself) will be found out
- 2/3 of top biz school fresh admits think they were the one mistake admissions committee let slip
- How to address
- de-stigmatize
- detach – that the thought is in your head is no proof that it is valid
- brain filters out what a) we conciously tell it to, b) what instincts tell it is important for our survival
- see it as a graffiti on the wall
- re-script
- body language is just physically too many elements to control consciously
- method acting focuses on the input (becoming the character) rather than output (how the character will behave)
- cognitive reframing annoyances to relieve your physical stress (an idiot driving in traffic -> what if it was a mother trying to save her baby choking on the back seat?)
- Training: short for improv (stage confidence), jujitsu/yoga (to own the space around you), tango (interaction of power dynamics)
- Two charismatic people conversing: everyone else around them can get a high from it
- Charisma is a similar tool in the toolkit as language
- Harvard Negotiation Project
OB278 – Organizational Behaviour (Flynn)
- On the folly of rewarding A, hoping for B – a short & powerful article on incentive systems gone wrong (PDF here)
- In US, whistleblowers collect up to 30% of fraud they report. Thus, prospective whistleblowers are encouraged to delay reporting, or even actively help its continuance before reporting, to run up the damages (and their cut)!
- Common reasons for rewarding something different than we actually hope for (in business, public policy, …)
- Fascination with an “objective” criterion. Likely to cause goal displacement outside highly predictable areas
- Overemphasis on highly visible behaviours. (Team building, creativity; results on wrongly pronouncing a sick person well more visible than treating a well person as sick)
- Hypocrisy.
- Emphasis on morality or equity, rather than efficiency. (Governmental intervention usually creates market inefficiencies, presumably for a “higher cause”)
- Core reasons for lack of motivation:
- I can’t do it (getting to the required performance after all the effort put in)
- I won’t get it (does hitting the desired performance actually result in the intended outcome for me)
- it isn’t worth it (the outcome, even if achieved, doesn’t matter to me)
- Goals _focus attention _& set direction – people love to be in that state of mind
- On a rainy day, there are only 30-40% of NYC cabs out
- high demand, but supply drops? wage elasticity? – economists get annoyed
- reason: cabbies set a “double-up the fee” goal
- …and deliver on it faster on a rainy day and go home
- Increase of desired behaviour is not the same as reduction of undesired behaviour
- Threats, even when stopping bad, don’t motivate good
- Research of spot rewards (small, frequent, cash or stuff awards in businesses)
- plus: less delay, frequent reinforcement of exemplary behaviour
- giving stuff is far more effective than cash!
- ex: $3 chocolate doubled the response compared to $10 cash award
- mental accounting is based on comparisons
- “$10 is little compared to all my existing money,” but “1 chocolate is a HUGE increase from all of my current chocolate: zero”
- economists get mad: how would you choose less value over more value + discretion?!
- at large levels will flip: cash is preferred over a car
- Peer bonuses more effective than manager-granted
- first is perceived as “they wanted to give it to me”, latter: “he just has to give it to someone”
- When designing perks, weigh your commitment to each before implementing them
- what would need to happen / what could you resist in order not to remove them again?
- 10% drop in revenue?
- missing a quarterly profit target?
- loss aversion makes the removal much worse than the initial good created
- Ex: an investment bank cutting everything to plastic spoons and paper plates after 2008
- what would need to happen / what could you resist in order not to remove them again?
- Entitlement is a neurochemical process
- Unexpected rewards release dopamine, predicted ones not
- Intrinsic motivations:
- autonomy – “own job redesign” a growing Valley trend, can make 3-5 changes to own job spec at reviews
- mastery (competence) – maximizers vs satisfiers
- relatedness – relationships themselves, but also social consequence to your actions: responsibility & recognition
- purpose – distance to your contribution or impact
- doing exactly the same research for the same pay in an university vs company could be less rewarding because you are further away from your theoretical work materialising, impacting other people
- When you as a manager need to divide a task list of 10 in team, you actually have 2 simple ways to increase motivation
- obvious: ask people to volunteer and pick the tasks they want and like to do (are comfortable doing)
- less obvious: don’t try to spread around about the remainder that nobody wants, but assign them to yourself
- Meta-analysis of motivation research papers shows there is material impact only in the span of control
- change is inflected by direct manager, not CEO speeches
- “Hippo syndrome” – big mouth, small ears
- Modelling (doing something in the hope of others replicating the behaviour) needs to be
- authentic
- public, observable
- routine – once is not enough
- create a pattern, not frequency
- people adapt to the regular, not the unusual
- often leaders attempt to design one-off, splashy modelling “events”, which are ineffective as don’t meet the above
- Singular communication focus: can you take 4min of any of your speech and have random audience agree what your main message was
- Apple keynotes
- counterintuitive: if you have 16 topics, there is temptation to share to show competence, “add max value” to the audience
- Persuasion needs contrast:
- “this is good/cheap/…” versus “this is better/cheaper than…”
- one of most effective contrasts is simplicity VS complexity (of competition)
- former is hard to resist for humans
- Communication repetition feels awkward
- boring?
- patronizing to the audience?
- use multi modality (say, show, play, demo, …)
- reinforce the idea instead of making redundant statements
- Book recommendation: Mindset: The New Psychology of Success
- Stanford prof; mandatory reading for all incoming managers at GOOG
CS207 – Software Economics (Wiederhold)
- “When you see a fork in the road – take it” – Yogi Berra
- “You would expect an average Stanford student to go through $10M in their lifetime, hopefully saving some of that for retirement”
- Technical parameters for IP valuation
- Life of the product
- Diminution of the IP over life
- Loss of value if you don’t maintain (fix, upgrade) it
- Not amortisation, because we’re not saving for a replacement
- Lag period – from transfer to diminution
- Relative allocation
- Crucial assumption for quantitative valuation: IP content is proportional to SW size
- Easy to argue against: not all code is equal, early code defines the product = worth more
- It is the best metric we can obtain?
- Codebase growth
- Hennessey: S(n+1) = 2 to 1.5 x S(n)
- Inspired from his hardware background, Moore’s law, etc
- Microsoft (Bernstein 2003): V(n+1) <= 1.3 x V(n)
- When you grow codebase >30% per version, can’t get it debugged (you loose the baseline)
- David Roux: V(n+1) = V(n) + V(1)
- SW grows per each version by the size of the first version that worked.
- Hennessey: S(n+1) = 2 to 1.5 x S(n)
- Cost of maintenance grows exponentially with size
- number of interactions between code segments grows faster
- deleting code is risky
- If competitive market keeps end user price the same across versions
- The revenue per code unit reduces by 1/size
- Yet you often need to add features (code) to compete
- Normal cure doesn’t apply to sales models in reality (infinite before the product is created)
- Erlang distribution curve: realistic (m=12 common), but more complex to calculate
—
For more posts on the Stanford GSB Sloan life – click here to search by tag “sloan”.