Key Words: Strategic Change, Innovation, Risk, Uncertainty, Ambiguity, Prediction Logic, Creation Logic, Planning vs. Testing, Project Management, Agile SCRUM vs. Waterfall, Lean Startup, healthcare.gov, JC Penney, Lululemon, Georgia Tech, Coursera.
How Executives Get Fired
On Oct. 1, 2013 the much anticipated healthcare.gov went live. And, almost immediately, it crashed. Unanticipated surge in web traffic was blamed for most of the problems. Even those who were able to get through faced a multitude of issues and errors – confusing instructions, missing drop-down tools, unexpected hang-ups and puzzling design. Those who gave up and called the customer service reps didn’t fare any better. The reps couldn’t access the online market place either.
On Nov. 29th, 2013 JC Penny (JCP)—an original member of the S&P 500 since 1957—was kicked off the list for its sharp decline in market value. While JCP still has more than a 1000 stores and 2012 revenues stood at $17B, the historic 100+ year old U.S. mid-range department store has fallen on hard times.
In Oct. 2004, Myron Ullman, a former executive at Macy’s and LVMH was named CEO. Ullman bought in brands like Liz Claiborne and introduced mini-shops within the department store. By Feb. 2007, JCP’s shares had doubled to nearly $80; a 10-year high. The economic downturn hit JCP hard and by March 2009 the stock was trading at $14. In Jan. 2011 William Ackman, a hedge-fund manager who had built up a sizable position in JCP stock, was appointed to the board. Amid Ackman’s push for new leadership, in June that year, former Apple retail-star Ron Johnson was named JCP’s new CEO, replacing Ullman. Johnson arrived at JCP in Nov. 2011.
In Dec. 2011, JCP acquired 16% of Martha Stewart Living Omnimedia stock and planned to put “mini-Martha Stewart shops” in many of its stores by 2013. In Jan. 2012, Johnson introduced a strategy involving in-store boutiques and a pricing plan that eliminated the popular JCP coupons. Instead, it would have “Every Day,” “Monthly Value,” and “Best Price” strategies. Prices would also not end in 9 or 7, but on whole numbers instead. In May 2012, JCP announced a 20% drop in sales and a $163M loss in Q1. Suddenly in June, the Head of Marketing, Michael Francis, who had come 8 months earlier from Target, said “he was leaving.” He was blamed for the marketing messages that were not resonating with customers.
In Aug. JCP started rolling out the “Shops” strategy in stores. Simultaneously, an overhaul of the home department in 500 stores was also started. In Nov 2013, JCP reported a Q3 loss of $123M as sales fell by another 27%. However, CEO Johnson said the firm won’t diverge from the strategy he laid out. By the end of the first year of Johnson’s turnaround strategy, JCP had amassed nearly a billion dollars in losses and a 25% drop in its revenues. In April 2013, Johnson was fired and Ullman rejoined the firm as CEO.
In Feb. 2013, an online internet course offered by Georgia Tech and hosted by the leading online learning firm Coursera promised to teach 40,000 students how to create their own massively open online course. The online platform asked participants to sign up using Google Docs. When the crush of students tried to sign up the system crashed. According to Google, apparently Google Docs only allows 50 people to edit a document simultaneously. A small detail, that seemed to have been overlooked by the planners.[i]
In March 2013, the high-flying Canadian yoga apparel maker and retailer Lululemon had to recall more than $60 million worth of a women yoga pants for being too see-through. Within a month there was an announcement stating “product chief to exit.” The following month the CEO announced that she was “stepping down.”
As we all know, expressions like “to exit,” “stepping down,” and “spending more time with family,” are just euphemisms for getting fired. Why do CEOs and executives get fired? The #1 reason is Mismanaging Change.[ii]
Analytical vs. Emergent Strategies for Growth, Innovation and Change
In a press conference after Johnson was let go from JC Penney, Bill Ackman—who had pushed for Johnson’s hiring, said that Johnson deserved criticism for unleashing a series of pricing and merchandising changes without first testing consumer views. Other Penney insiders criticized Johnson for eliminating the company’s sales and coupons last year without a broad market test, a move that led to a sales slump. The key words to focus on, in the above two statements, are “testing consumer views” and “broad market test.”
In March 2013, six months before the healthcare.gov website went live, McKinsey was asked to do a risk analysis and to develop mitigating strategies. McKinsey submitted a 14-slide presentation to the White House by early April. I have pasted two key slides from that slide deck. The first slide is about “complexity” and the second is about how to manage “complex projects.”
A website like healthcare.gov is a massive and complex undertaking – too many variables and too many unknowns. When you are dealing with unknowns, one is talking about uncertainty and/or ambiguity. On the other hand, risk is about the “known” world – known variables with data from the past. You can calculate and estimate risk using analytical tools. When you know the variables and you have data from the past, one can analyze, predict, plan and then take action. Specifically, we can go into the future in primarily two ways – (1) Analysis before Action and (2) Analysis after Action.
Traditional approaches to minimize and manage risk in innovation and change management projects is by doing a lot of analysis before taking action. BHAGs (Big Hairy Audacious Goals) are announced with much fan-fare. Then, the future is approached by performing an environmental scanning (SWOT, STEP, Value Chain Analysis) and followed by the setting of a project plan to execute strategy. Trend lines are predicted based on IRRs and WACC or projected cost benefits; KPIs and milestones are set and budgets are allocated. When project performance does not meet projections, money and energy is spent to get the project back on to the predicted trend line. Unfortunately, heads roll when the predicted future fails to materialize after a couple of tries.
Below: McKinsey’s exhibit demonstrating the magnitude & complexity of the healthcare.gov website project.
This approach to change and project management makes a bunch of assumptions: (1) all process and outcome variables are known and can be accounted for ex-ante, (2) existing data from past projects can be used to predict the process and outcome of this project, (3) some variation to projections can be accommodated along the way using managerial judgment, and (4) failure is not an option.
This concept of going into the future is called “predictive logic” and the method is called “analytical strategy.” I just call this the BIG BANG approach to change. Most large firms, governments and institutions predominantly still prefer this mode of going into the future. I call the firms, organizations and individuals who principally use this strategy of going into the future as “PLANNERS.”
On the other hand, all innovative and complex change management projects have a number of unknowns. Specifically, there are two types of unknowns – known unknowns and unknown unknowns. Uncertainty is about known unknowns. In these situations, you know which variables may impact the process and outcome of the project but there is no data from the past to assign probabilistic numbers. Ambiguity is a second order uncertainty. One cannot surmise as to what variables may be lurking in the background. They only appear once the project is underway. Unfortunately, analytical strategies do not account for these unknowns ex-ante. So, when there are a number of unknown variables, most analysis and hence prediction of outcomes a priori becomes a futile exercise.
In the presence of unknowns, the way to manage projects is drastically different. Seasoned entrepreneurs, innovators and VCs test their ideas for potential opportunities predominantly through Analysis after Action. They Think Big, but Start Small. They start several small projects to test their hypotheses. They prototype rapidly and try to establish proof of concept by quick feedback from the market – voice of customer, voice of technology, voice of supply and voice of demand. They try to fail fast, fail cheap and fail smart. In doing so, they learn quickly by uncovering hitherto unknown variables and/or create data where there is none. With this new knowledge they refine their hypotheses and business models. They iterate this process of prototyping, failing, uncovering unknowns and establishing a viable business model. They pour in more resources only after a positive proof of concept has been established and the successful business model is replicated and scaled slowly. I call this approach to going into the future as START SMALL; as against the BIG BANG approach described previously. I call the firms that employ this technique as “TESTERS.”
In 2009, this method of going into the future was termed “discovery driven growth” by Rita McGrath. Decades earlier in 1978, Henry Mintzberg termed this as “emergent strategy” as against “intended or deliberate strategy” (analytical). In the mid-1990s software developers started using Agile Scrum (iterative emergent techniques) vs. the traditional Waterfall methodology (sequential analytical techniques) based on the work of Takeuchi and Nonaka in the mid-1980s. Most of today’s “Lean Startup” (Steve Blank 2012, Eric Reis 2011 and Ash Maurya 2012) concepts, principles and frameworks profess this very same emergent strategy. At Babson, our entrepreneurship and innovation faculty has been teaching this stuff for decades.
To summarize, PLANNERS usually follow the traditional BIG BANG approach that is characterized by the following sequence: Analyze > Predict > Plan > Act > Full Scale Launch. TESTERS on the other hand follow the START SMALL approach that is characterized by the following sequence: Design > Build > Test > Learn > Redesign > Scale Slow Launch.
Below: McKinsey’s slide that contrasts, the Start Small “emergent strategy” technique predominantly used by “Testers” (on the left) vs the Big Bang “analytical strategy” technique usually used by “Planners” (on the right).
I am not suggesting that one approach is good and the other is bad. The right question to ask is: when do you use analytical strategies and when do you use emergent strategies. The Big Bang approach to change or project management works very well for version 2 or 3 of a product. For incremental innovations and for the known world – known technology, known products, known customers, known business models etc. – when we have lot of data and prior experience, the Big Bang approach still works very well. Unfortunately, in the unknown world – unproved technologies, unidentified customers, untested business models – and for radical innovation or major organizational change the Big Bang approach fails miserably. The Start Small approach works much better.
Unfortunately, healthcare.gov chose the Big Bang approach. At least 5 months prior to launch, McKinsey’s warnings were quite clear, “…there was scant time to test the system before launch;….there wasn’t enough testing and revision;….create a Version 1.o before full launch….”
I have elaborated on these the Start Small concept in a previous blog as well:
Have a great holiday season!