• Running

    You can learn a lot about how people approach their careers by looking at how they approach their hobbies. Running is such an important part of my life that I have created a separate blog for it, Predawn Runner. Whether you are recreational or competitive, I welcome you to join me there in discussing how we fit running into an already-full life.

Developing a Tolerance for Ambiguity

By Greg | Published: January 14, 2010
Ambiguity - What happens in Vagueness stays in Vagueness
Celebrate your appreciation for ambiguity with a T-shirt like this from the Mental Floss store

Early in my career at GE, one of the leadership values the company mentioned, but never really emphasized, was a “tolerance for ambiguity”.  There was never any coaching on how to develop such a capability. Even now, I can’t really find a reference to it on the web (though current GE CEO Jeff Immelt recently emphasized the need for leaders to be “comfortable with ambiguity” in a speech at West Point).  Over the years, however, I have come to appreciate how important this skill can be for a leader and learned how to develop it as a strength.

Wikipedia defines “ambiguity tolerance” as “the ability to perceive ambiguity in information and behavior in a neutral and open way.”  I prefer a more active definition, so I consider a tolerance for ambiguity to means“planning and executing appropriate actions in light of limited information.”  The emphasis is on being able to move forward in spite of limited or conflicting information, as opposed to just “neutrally” recognizing that such a situation exists.

Ambiguity is all around us, whether it is in the form of uncertain business or economic conditions, unclear job descriptions or expectations, or vague corporate strategies. Today’s economic environment, which some economists have taken to calling The Great Ambiguity, presents an extreme case of an uncertain outlook.  Your nature and upbringing can shape your natural tolerance for ambiguity.  Education and early career experiences also play a major role. For example, accountants tend to have little tolerance for ambiguity as their education and experiences are based in clear rules. Marketers may have too much tolerance, and often resist moving towards metrics-driven processes.  Engineers (author’s note: my education is in engineering) may surprisingly have the best background for developing this capability, as more complex problems require making assumptions to deal with limited information.

Give me Ambiguity or...
Sometimes ambiguity is the only choice… (photo courtesy of VatorNews)

Whatever your background, you can improve your ambiguity tolerance.  I have found the following steps to be helpful in doing so:

First, make sure all potential data sources are exhausted. Tolerating ambiguity is not an excuse for lacking diligence in data-gathering; well informed decisions are still better than the best assumptions.  And I don’t mean just surfing the Internet.  Get out and talk to customers about business conditions, or colleagues about roles and responsibilities.  You will know you are reaching a limit when the incremental effort to acquire more data outweighs the value of that data.

Brainstorm assumptions you can make to close the gaps in your data. Depending on the scope of your problem, this may or may not involve others.  Assumptions should be based on whatever known data, historical precedent, or “common knowledge” may exist (but be particularly careful on the latter).

Document your assumptions transparently. This serves two purposes. First, you will have a reference to go back and review and test as results occur (which may help to confirm or refute your assumptions).  Second, this helps you CYA in the event outcomes are different than planned.

For important situations, consider alternate assumptions. This may take the form of “worst case” analysis, or at least analyzing a scenario where assumptions do not play out as favorably as in the “base case”.  This can help you prepare contingency plans to react quickly to new data or unexpected results.

Define a process and rhythm for testing your assumptions. This process should define the frequency which you will review results, the metrics used to test results, and adjustments you will make as performance differs from targets.

Execute your plan. This step is obvious, but a reminder not to get over-involved in planning to the point where execution is forgotten is always helpful.

Let’s look at an example. Today, with many workers asked to do more with less time, money, and resources available, ambiguity about job definitions and descriptions is rampant.  My own job description is vague, and frankly given the environment we work in I’m not sure it can be made much better.  Here is how I have dealt with this ambiguity.

Gather data. I started by speaking with my manager about his expectations and measurements he would use to evaluate my performance. I also spoke with my predecessor about what priorities he placed in the role, and where he felt his performance was strong and weak.  As a marketing and product management leader, I also spent time discussing expectations with other significant stakeholders in my performance, such as sales and engineering.  This still resulted in incomplete or conflicting data (and usually will) but at least established some “boundary conditions” on the role.

Brainstorm assumptions. I then dedicated 30 minutes to summarizing what I knew and listing possible assumptions to fill in the gaps, based on needs I knew my employer had, past responsibilities I’ve had, objectives that my manager had in his role, and other related topics.  For example, I knew that growing the business into new market segments was a key objective for my manager, so I listed developing and implementing marketing plans for new markets as one of my own objectives.

Document assumptions transparently. Once I had narrowed down the list of assumptions to those that seemed reasonable and achievable (allowing for a bit of stretch of my capabilities), I documented my understanding of roles and responsibilities and scheduled a meeting to review them with my manager.  If he asked for any changes, I made those changes, and then emailed him the final document.

Define a process for reviewing assumptions. In our fast changing world, waiting for an annual review (if you’re even lucky enough to get that) isn’t enough.  Therefore, I planned quarterly re-calibrations of priorities, where I would sit down with my manager and review accomplishments, comparing them to the responsibilities document, and discussing what changes were necessary in light of evolving business and organizational conditions.

Execute. Of course, everything else is worthless if I didn’t go out and deliver on my commitments.  The responsibilities document is one of the few items I allow to clutter up my corkboard.

Ambiguity is a fact of life.  Learning to develop coping mechanisms that enable you to perform not only helps position you for growth, it is almost a matter of survival in the corporate world.

Bookmark and Share

You may also find these interesting:

Posted in Leadership | Tagged , , | 9 Comments

A Marathon Training Schedule for Advanced Runners

By Greg | Published: December 22, 2009

Several of my running friends have asked me what system I use for marathon training.  I always respond with praise for Hal Higdon’s marathon training system, which has something to offer for runners of all levels.  Having started with Novice for my first marathon (and essentially meeting my goal of a near-3:30 time), I have progressed through his Intermediate I and Intermediate II stages (and was able to qualify for Boston using just the Intermediate I program with some speed enhancements) and am now going to make a stab at Advanced I.

Hal Higdon provides an outstanding range of marathon training schedules to bulid from

One reality of being a father and an executive with a busy spouse living in snow-prone Northeast Ohio is that any running schedule requires some flexibility.  Thus it is important to understand the principals of your training to know where you can make adjustments.  The key elements of any marathon training program targeting a specific finishing time are as follows:

  • A build-up of distance, both weekly total and the longest run each week
  • Sufficient focus on pace, including workouts focused on outright speed.

As you make adjustments to your schedule to fit your realities, you need to protect your ability to include these elements, with appropriate rest before and after to make sure you are able to get the most out of the workouts.

What I’ve always liked about Hal Higdon’s 18 week training programs is the consistency of the weekly schedule.  The ramp rate is logical and easy to adjust.  For example, I will be heading into training for the 2010 Cleveland Marathon in better shape than usual, having been putting in around 30 miles a week of late.  However, there is no need to add mileage at the later weeks in the schedule, so I’ll add a few miles per week at the beginning and just have a slower ramp rate.  Additionally, since I’m an early morning runner but do appreciate the occasional ability to sleep in (meaning to 5:30 on weekdays), I’m going to replace one of the easy runs with a rest day, or perhaps cross-training at lunch. Should the urge strike, it’s easy enough to add another run occasionally.

The modified schedule is attached.  Note that I also need to work around a minor surgery in January, which should lay me up for a week.  Also, since I don’t live near any hills and don’t want to drive somewhere to run, I’ve replaced hill training with more tempo and interval runs. After two weeks of ramping up the weekend length, there is an easier step-back weekend (which is always a relief).

A Marathon Training Schedule (Cleveland 2010)

View more documents from Greg Strosaker.

Here are a few definitions:

Intervals (4 x 1 mile and similar): these are the core speed workouts, consisting of a hard run for the interval with repeats, allowing a 3 minute jog in between.  The original program calls for 800m intervals but full miles seem more appropriate for marathon-level speed work. The goal is to improve the pace each mile by about 5 seconds over the previous one.  The average pace should be just a bit slower than 2x your target marathon time; for me (3:15 target time) that comes to around a 6:40 interval time.

Tempo:  these runs start at an easy pace, build to 10K race pace near the middle (for about 30-50% of the run time), then finish easy.  To calculate a 10K race pace, use this calculator and your targeted marathon time, assuming such a target is realistic.

Pace:  These are medium-long runs at your targeted race pace, and such runs are a great opportunity to practice running progressive splits to really get a feel for pacing.  They constitute about two of every three medium-length weekend runs.  When combining such a run with the long run the next day (peaking at a 10-mile pace run followed by a 20-mile long run) you build your core endurance for the marathon. Note that it is important to keep the medium and long weekend runs on back-to-back days when possible to get the most out of the long run workout.

Long runs:  Hal advocates doing these at 45 – 90 seconds slower than race pace, and making 1 of every 3 a “3/1 training run”, where you push the pace to something just a bit slower than race pace.  It can be hard to slow yourself down to such a pace, so again trying to achieve progressive splits can help.

I like to note actual distances and pace as I train, so I can go back and see what might need improved for the next race.  This is the most aggressive training schedule I have put together so far, but not by much, so I’m hoping it is achievable without burnout.  Hopefully I’ll see some of you at the Cleveland Marathon in May.

Addendum:

I’m happy to report that I did, in fact, qualify for Boston at the Cleveland Marathon with a personal best of 3:08:48.  Thus, this training program proved a success. I have published the next step with further improvements to an advanced marathon training program targeting a 3:00 marathon on the Predawn Runner blog.

Bookmark and Share

You may also find these interesting:

Posted in Running | Tagged , , , | 6 Comments

Market Segmentation for Electric Vehicles

By Greg | Published: December 17, 2009

Nick Hodson and John Newman of McKinsey recently published a brief article proposing a segmentation approach for electric vehicles.  This study was driven by a concern about the expense of electric vehicles (driven by the cost of batteries) and the ambiguity around the market potential for such vehicles.  There are two aspects of this article worth discussing: the approach taken to segmenting the market, and the product development recommendations that this analysis yields.

The authors’ approach does follow some of the principles I laid out in a previous post on effective market segmentation.  To some extent, they characterize potential segments based on behaviors (driving patterns) as opposed to characteristics (such as age, area of residence, income, or even a specific average drive length).  Their focus on how vehicles are used (commute, around-town errands, delivery, sales calls) yields insights into specific customer needs, specifically the range that batteries must cover.  The differences between the segments they define are significant enough to be meaningful for product development and marketing purposes.

The use of these segments to propose differentiated product development is even better than the segmentation itself.  As I discuss in a prior post on using market segmentation to define pricing strategies, proper analysis can point out design approaches that yield a selling advantage into specific segments.  In this case, the “optimal” battery range for each segment provides the potential to design lower-cost options for the “driving around town” segment and more robust and flexible (and correspondingly pricier) products for commuters.

This graph from McKinsey shows the significant difference in optimal battery sizes for plug-in hybrid electric vehicles by market segment.
This graph from McKinsey shows the significant difference in optimal battery sizes for plug-in hybrid electric vehicles by market segment.

For automakers, deeper segmentation will be necessary.  The authors do not use actual customer interviews to determine additional behaviors that will impact preferences for electric vehicles such as level of concern for the environment, ability to understand and appreciate return on investment, or the desire to be seen as owning “the latest trend” (or aversion to new technology).  Such differences will be important in defining whether any reasonable premium for an electric vehicle over traditional IC engines is possible. It is going to be some time until the economics of electric vehicles provide a clear incentive for buyers, so appealing to non-economic criteria will be important to establishing a lead in the industry.

One-size-fits-all has never worked for the automotive industry, and this will become even more true as the various flavors of electric vehicles (hybrid, plug-in-hybrid, and plug-in) are introduced to the market.  Add to that competition from diesel engines, advanced internal combustion engines, and, potentially, hydrogen fuel cells, and the choices (and opportunity for mistakes) in a product roadmap will quickly become overwhelming without a clear line of site to customers based on thorough segmentation of the market. McKinsey has also published an outstanding article on the outlook for the automotive industry as they plan strategies for electric vehicles.

The authors have done a fine job of researching available information and wouldn’t be expected to carry the segmentation further.  Based on the history of US automakers (where a “Chevrolet” is simultaneously a $10,000 Cavalier and a $60,000 Corvette), I doubt this will be done effectively.  Smart management of the shift to electric vehicles will be the key determinant of winners and losers in the automotive industry over the next 20 to 30 years. Or, they can just wait for the current generation of children to be subject to propaganda pieces like They Might Be Giants’ song Electric Car (though I have to admit its a catchy tune, and my boys love it).

Bookmark and Share

You may also find these interesting:

Posted in Marketing | Tagged , , , , | Leave a comment

The Running Manifesto

By Greg | Published: December 10, 2009

The Running Manifesto has moved to a new home on the Predawn Runner blog; please be sure to stop by to view the manifesto and join the conversation on fitting running into an already-full life.

Should you find these thoughts motivating enough (as some apparently have), I’m happy to announce The Running Manifesto Shop on CafePress, offering shirts and prints of the manifesto.

Bookmark and Share

You may also find these interesting:

Posted in Running | Tagged | 7 Comments

A Market Segmentation Example

By Greg | Published: December 3, 2009
Results of our market segmentation analysis for the TIM market
Results of our market segmentation analysis for the TIM market

In the previous post, I laid out the principals of market segmentation, specifically differentiating it from classification.  I’d like to provide an example, building from work done in launching a new premium filler material into the thermal interface material (TIM) market.  These materials are used to extract heat from microprocessors and other electronic devices to maintain their optimal performance. We provided the thermally conductive materials that, when combined with a polymer, dictated the performance of these materials.

The cost of this new product necessitated a significantly higher selling price than our existing materials, and we knew that commercializing it successfully within our core business would be a challenge.  Thus, we performed a market segmentation process on the broader TIM universe to see if other segments existed and might be more attractive.  To do this, we spent a lot of time meeting with existing and potential TIM customers to try and answer such questions as:

  • What are their decision making processes?
  • What criteria do they use in selecting vendors/solutions?
  • What are their growth strategies (product roadmap, key applications, etc.)
  • Who are their customers, and what is expected of them as vendors (what is their value proposition)?
  • How do they work with vendors (arms-length, etc.)?

Often, our past experiences with customers yielded insights into these questions as well.  We also engaged an external market research firm who had done work in the industry to understand which of our customers supplied which end users, which applications were likely to grow at above-market rates, etc.  We then analyzed our sales to each company in the market and assigned some “potential” size to each account, and finally worked through several iterations of grouping accounts based on answers to the above questions.

At the conclusion, we ended up with 5 segments:

Leading Edge:  These companies were driven by a desire to retain their position as the specified solution for advanced applications such as Intel or IBM microprocessors.  They often used very expensive filler material such as silver, and experimented with new materials constantly, often driving product development by filler vendors.  Their engineers were true experts in the industry, and purchasing took a back seat in vendor discussions.

Jump to Respond:  These companies were always developing new materials based on demands by high-volume key customers. They typically served “second tier” applications such as heat sink attachment or heat spreading in laptops.  Constantly under pressure to deliver to new customer demands, they liked to have a range of solutions to use to quickly design new products.  Engineers still drove the discussions with vendors.

Grease Monkeys:  These companies were similar to “Jump to Respond” in offering customer-specific solutions but tended to focus on lower volume niche applications.  They also relied more on general chemistry expertise and were broadly involved in advanced adhesive or related applications in a range of industries.  They tended to have more limited resources than the “Jump to Respond” players so needed additional application support from their vendors.

Wannabe but Can’t:  This was the most deceptive group as they had some characteristics of the “Leading Edge” (significant resources, driven to try and serve challenging applications), but at the end of the day they were too adverse to trying more expensive solutions and always wandered back to trying to maximize performance from low-cost filters. They had a small share of the market but grand visions, but it was clear that realizing those visions was always going to be a challenge.

Pump it Out: These companies provided high volumes of standardized materials, and while they used quite a bit of our product, they were more focused on operational expertise and trying to squeeze cost out of existing designs (negotiating prices, reducing filler usage, exploring cheaper fillers, etc.).  They were the highest volume accounts in the market so could not be ignored, but the solutions that worked for them were not the most advanced but the most mature ones.

Notice that the segmentation is based on customer values and behaviors, not on such details as their size, location, types of products, or other classifications.  Each segment contained a range of different sized companies, but these companies were similar in how they valued new fillers.  We could therefore plot them by the criteria that were most important to our new product launch: appetite for new technology and willingness to pay a price premium (see the graphic, plus page 4 of the attached presentation). We identified the segments which placed the most value in our product (Jump to Respond and Leading Edge) and tailored our messages and even our pricing to meet these segments needs price (more details in a later post on how pricing was set using value-equivalence lines).  It also instructed us to avoid spending time on unappealing segments (Wannabe but Can’t, Pump it Out) to focus our resources more specifically.

09 1203 Market Segmentation – An Example

Our earliest success was with the “Jump to Respond” segment, launching the product at a higher-than-anticipated price. We found that we did not quite have the right product yet for the “Leading Edge” or “Grease Monkey” segments, but were able to identify derivatives of the new product for these segments with greater likelihood of success.  Without this exercise, we would have wasted significant effort trying to commercialize the product to our core “Pump it Out” segment without success; at best we would have missed significant opportunities to price the product at a premium for those customers to whom it provided the best value proposition.

I’m happy to hear your thoughts on how we could have done this better, or experiences you have had in market segmentation.

Bookmark and Share

You may also find these interesting:

Posted in Marketing, Strategy | Tagged , , | 10 Comments