Subscribe to the OVP Blog RSS Feed: Enter your email address:

Delivered by FeedBurner


Technology Innovation Along the Education to Employment Pathway - Part I

Recently, McKinsey released a report that, depending on your perspective as man or machine, paints a potentially worrisome view of the future.  Based on a mind-numbingly detailed analysis  of all of the tasks conducted in the global economy, McKinsey estimates that nearly 50% of the activities done today have the potential to be automated based on currently demonstrated technologies in robotics, machine learning, and artificial intelligence.  This represents $8 trillion of wages collected by 1.1 billion people globally, and while the study suggested that less than 5% of today’s jobs could be fully automated such that humans are entirely replaced by robots, more than 60% of jobs could see at least 30% of their underlying tasks done more efficiently by machines. 

We are already seeing examples that cut across all areas of the economy.  Smart warehouses and factories where robots supplement humans, driverless cars threatening legions of Uber drivers – who didn’t even exist themselves five years ago, intelligent chatbots augmenting customer service representatives, a life insurance company in Japan replacing claims adjusters with the algorithms of IBM Watson, and even robot baristas.  As investors and entrepreneurs, the technologies enabling this age of automation – machine learning, artificial intelligence, cloud computing, robotics – offer tremendous potential.  McKinsey estimates that automation could add more than 1% to global GDP growth as a result of productivity gains.  At the same time, a reallocation of how work gets done will create winners and losers in the short term; estimates by McKinsey and the World Economic Forum estimate that somewhere between 5 million and 7 million jobs have already been lost to labor market changes resulting from automation.  Managing through the transition imagined by the McKinsey study and the WEF has profound social and political implications that I will actively avoid addressing – those are important debates for other forums.

Instead, I want to highlight some of the ways that technology has already begun to influence what Innovate+Educate and Whiteboard Advisors have described as the education to employment pathway, which requires re-imagination to better fit the unprecedented pace of change and innovation associated with this age of automation.  The World Economic Forum estimates that 65% of children entering primary school today will end up working in completely new job types that don’t currently exist, while one third of the skill sets required for jobs today could be completely reinvented in the next five years.  The current educational system, with a primary emphasis on a massive up-front investment into a bachelors degree, is poorly equipped to help learners constantly refresh their skill base, which is reflected in surveys such as a 2015 Gallup study that suggested only 11% of business leaders feel that college graduates have the skills necessary to succeed in the workplace or just 44% of students report that their post-secondary education prepared them for the workforce.  At the same time, corporations have dramatically reduced their investment in formal training programs (despite spending more than $130B on corporate learning in 2013, according to Bersin by Deloitte), with one study in the UK finding that the average hours of training per week fell by 50% between 1997 and 2009, as reported by the Economist, while another by the Bureau of Labor Statistics reported that the number of hours of training received by young employees fell from 100 hours in 1979 to 11 in 1995.  These factors have combined to create a perceived “skills gap,” with 40% of employers in the US reporting that they have positions they are unable to fill while there were more than 4 million open job postings during the height of the great recession when more than 11 million people were unemployed.  While several researchers have questioned the existence of the skills gap based on macro level data, very real issues exist at the micro level in certain pockets of the economy.  To take one example, approximately 50,000 open job specs in the US asked for a specific cyber security certificate, yet just 65,000 people in the US hold such a certificate, and it take at least five years of experience for someone to earn one. 

Innovators on both sides of the education to employment pathway have begun to recognize this disconnect between the skills employers require and the capabilities of the workforce.  Innovate+Educate, in their report Shift Happens, highlights a need for solutions across four primary areas.  First, how skills and competencies are developed by both educational institutions and employers.  Second, how this new content is most effectively delivered.  Third, how employers can use data and people analytics to better understand the skills required for the future and create feedback loops to the educational system.  And finally, how to create a more efficient process for matching an individual’s skills and competencies with job requirements that move beyond previous experience and a college degree.  Technology sits at the core of this re-imagination of the education to employment pathway, and in the remainder of this series of blog posts, I’ll highlight a few interesting companies and approaches across each of the dimensions.


Understanding Decision Fatigue and the Implications on a Start-up

The New York Times recently republished a very interesting 2011 article “Do You Suffer From Decision Fatigue”,

which describes academic research into the impact of decision fatigue on the quality of decisions.  Click on the link and read it because it may change how and when you make critical decisions.  Some of the takeaways include;

- As people make decisions, they become tired from the act of making the decision.  While assessing options and implementing a decision after it is made takes energy, the act of decision making itself is by far more taxing. 

- As people are asked to make more decisions, they will increasingly lean toward status quo or no decision versus continuing to make decisions.  Or they will go along with recommendations of others versus making their own decision.

- If you are already tired – at the end of a long day – the energy you can commit to decisions is much lower and your speed to decision fatigue is much higher

- Decision making burns glucose at unexpected levels.  Taking breaks and replenishing glucose can take away much of one’s decision fatigue and can reset the clock for a period of decision effectiveness.

The article talks about members of an Israeli parole board who were assessed on their decisions to parole similar convicts having been convicted of similar crimes.  It turned out that convicts in similar situations had a 70% likelihood of parole early in the morning and just 10% near the end of the day.  Percentages after breaks and after lunch also spiked but rates prior to lunch were quite low.  Across a broad number of cases, these behaviors became predictable and were attributable only to time of day, length of time since the last break, and time since the last nourishment. 

At Osage we think a great deal about decision bias and trying to remove as much bias as possible from the investment decisions that we make.  The writings of Kahneman and Tversky have become widely read and respected with regard to decision bias and we integrate these lessons in what we do.  But decision fatigue may also play a role in many of our critical decisions in ways that we may never be fully aware.  This made us also think about the thousands of decisions the leaders of our portfolio companies make and that their customers make and how decision fatigue might affect business success.

How does decision fatigue relate to start-ups:

Think about sales meetings: We don’t control what has happened during the day to the potential buyer of our software solution but unfortunately for us, this could be as critical as to what she thinks about your product.  So, when you can, schedule meetings or critical sales calls in the morning.  If the meeting is in person, bring a universally liked snack or two – dark chocolate squares, a five pound bag of M&Ms for a big group, vegetables and hummus – for afternoon meetings.  So many sales processes end with no decision.  How many of those fell victim to decision fatigue versus other corporate priorities?

Think about lead conversion:  A lot of analysis has been done about how lead conversion improves with speed of response.  Call in the first few minutes of someone filling out a form online and their conversion rate can be 4x higher than average.  Reaching out to people who fill out a form on weekends, early mornings, or late evenings have proven to be even higher.  There is a reason why a person who expresses interest in a product while surfing the web on a Sunday afternoon has a meaningfully higher conversion rate – their decision fatigue is low and the interest is high.  Getting back to them while they are in this state can accelerate sales cycles.

Think about management meetings at your company: Scheduling meetings in the morning versus afternoon may make a lot of sense.  Not making too many decisions in one meeting may be critical.  If you want to get something approved with little discussion, put it at the end of a long meeting agenda and make a strong recommendation.  Haven’t you noticed that people spend a lot of time discussing things early on an agenda but rarely take as much time to discuss things late in a meeting?  And remember, food for the meeting room could also be critical. 

Think about when you make important choices as a leader: As a decision maker, be careful of quickly agreeing with someone else’s recommendation or simply putting off the decision.  Are you really supporting this person and their preferred path, or are you tired?  Take a look at your calendar and think about the day – are you more taxed than you think? 

Think about hiring: How many interviews should you do at a time?  When do you stop listening?  I know that I have behaved like that parole board when I have been forced to do a ten session slate of interviews in a day with few breaks.  Sometimes it’s hard to listen or engage, and making no decision by simply passing on the candidate becomes the default as fatigue increases.  How many great candidates get rejected because the interviewer is simply fried from the process?  I was reminded of this recently when I was screening several hundred resumes from an MBA program resume book.  With decision fatigue in mind, best practice likely would have a person do this in thirty minute intervals and have a limited number of review sessions each day – and even better, to get several people to score each resume while putting each reviewer’s resumes in different order to make sure the same people were not given early preferential views by each reviewer.  No wonder people whose names are at the beginning of the alphabet seem to see more success.

Decision fatigue – be aware of it.  When you are in a position to make important decisions, ask yourself if you are mentally in the right place to be taking such an action.  If not – take a break, have a snack, maybe even sleep on it.   You won’t regret it.


Common CEO Mistake: The Buddy Hire

It is tempting.  You are a new CEO and you need to make an impact quickly.  You have been successful in the past and there have been people who were on your team previously whom you would love to have with you on this new journey.  You don’t know the current team that well and have not fully assessed everyone’s talent, yet some new blood at the top would definitely help.  It is tempting.  Tempting to hire quickly, to bring in a number of people you know, and to make sure you have people around you who you trust and who can have your back.  Tempting but in so many cases this is the wrong move.

When I first arrived at Verticalnet, where I ultimately became CEO, in mid-2000, we had a new leader who had been hired from a different industry.  This CEO, Joe, practiced this kind of hiring broadly and brought in many senior executives who knew little about the business or industry.  Then his hires acted the same way.  Soon business development was full of former supply chain consultants and marketing was all ex-Lotus people, while product management seemed to be from two camps with completely opposite views of the world, and marketing was run by someone used to selling to big box retailers.  It didn’t end well as 1,500 jobs were shed from that point until about the time I took over and most, if not all, of these buddy hires did not survive the cuts. 

Think about the potential mistakes of this “buddy hiring” behavior if you were the CEO:

-          In crowd / out crowd:  Quickly there are the new hires who have a history with you and the old hires who don’t.  A natural gap is created, especially if the people you hire know each other and not just you.  The motivation of the people who have been there is quick to decline, despite their initial optimism about a new leader and a new approach to the business.  Often, the new hires are not local and will need to be relocated over time or do the weekly commute.  Thus these new people are in the office for only three full days a week, amplifying the challenge.


-          The wrong hiring tone has been set: You hired these new team members without having the existing team interview them or you said to the existing team; “I want you to sit down with Meg and tell me if you think she is a good fit for CTO.  She and I worked together for five years and I think she is awesome.”  Now really, what can your team say?   “She’s awesome” is the usual reaction.  Your hiring behavior just set the tone for the company.  Now others will do the same thing and will hire people they knew before and quickly every function potentially becomes a little clique of previous co-workers. 


-          To a hammer everything is a nail:   Your people might have been good at the last job, but will they be good at this one?  Different business, different customer, different solution.  Some executives are great at jumping from situation to situation but many are not.  Pattern recognition is a negative when the business is different.  Hiring people who say “we did it this way at the last place where I worked with our current CEO” is not only alienating but it often could be just the wrong thing to do and not what you, as CEO, want them to do.


Do yourself a favor and institute a formal hiring process, or if the company has one, abide by it.  Create job descriptions and interview scoring templates.  Have hiring teams who are tasked with interviewing a minimum number of candidates before making a hire.  It is great if you have a strong network and access to talent and you should always be excited about bringing people from your network into the process – as long as you tell people to hire the best person and not the person you know.  Maybe you should interview the top three candidates and form your own opinion of who is truly the best individual for the situation.  Require relocation of new hires if the job is an on-site job.  You will still end up hiring some people with whom you worked before and that’s ok, because everyone will agree that they are great and were part of the process.  Even then, team building exercises are critical, as are creating new stories from shared experiences at the current company while leaving most of the old stories at the door.  

One caveat.  When building out a sales team and looking to move quickly to hire sales resources, leveraging a network seems to create much less resistance.  This is likely because the results are tangible and those that fail will not survive, no matter the past relationships.  Still drive a structured process where possible, but in technology sales, past success is often a predictor of future success.

The behaviors described above are very real.  We see these actions inside of our portfolio and we look to change these behaviors before the culture gets permanently altered.  When we are interviewing CEO candidates, we ask in detail about hiring and HR practices as well as how people leverage their networks.  A CEO with a strong personal network is a great asset, but only when combined with a strong process-based hiring discipline.


Watson Food for Thought

I am a big fan of Watson and the business model vision that IBM has for allowing other businesses and start-ups to leverage the power of Watson for their own solutions.  While Google, Facebook, and other massive data stores seem intent on leveraging their own data for their own use, IBM appears to be taking a more open approach, leveraging its massive base of computing knowledge to accelerate other companies’ strategies and collect recurring tolls along the way. 

As investors this creates what is likely to be a huge next wave in start-up innovation driven by artificial intelligence, and as early stage tech investors, our team here at Osage Venture Partners will do our best to invest to seize this opportunity.  Corporations see the huge economic benefit of replacing people with technology solutions, especially as we shift from replacing bank tellers, toll collectors, customer service reps, and gas station attendants to replacing much more expensive knowledge workers.   The social implications of this potential disruption deserve much debate and discussion but that is for another time and another post.

Perceptions on the viability of AI to displace white collar jobs are changing fast.  Industries that never saw automation replacing key roles are shifting their views of the future.  Take law as an example.  I found this following data in an interesting blog:




There is a lot to remark on in this data.  First, look at how perceptions have changed in four years.  A 50% increase in percent of people who think first year associates can be replaced in 5 to 10 years by a Watson equivalent– up to 35% of those surveyed.  Almost half believe paralegals will be able to be eliminated in this time frame.  As this was a survey completed by firm leadership, of course the percent who believe partners will be replaced is lower – but is this reality or some form of polling bias, as Daniel Kahneman would point out? 

Of course, this is a poll trying to predict the future, so it could be wrong in both magnitude and timing.  Yet recently with technology penetration, the timelines have been shorter and the impact has been higher than earlier estimates had predicted.  Remember how everyone laughed when Jeff Bezos discussed drone delivery in December of 2013?  Remember how recently we laughed about autonomous vehicles?  Four years ago, 46% of law firm leaders polled said computers would never replace human practitioners.  In 2015, it was down to 20.3%.  Time will tell.

Similar trends are emerging in other knowledge worker functions.  In Japan,   Fukoku Mutual Life Insurance is reportedly replacing 34 human insurance claim workers with “IBM Watson Explorer,” starting this month  Today actuaries – tomorrow lawyers?

The good news is that we should be seeing legal fees coming down steeply in the future, although the computer which replaces me may not need to hire its own lawyer.


Salary, Value, and Poaching

We have seen a number of high quality people getting poached or almost poached from different portfolio companies over the last couple months.  The labor market for skilled technical and analytic people is tightening and the war for talent seems to be only getting more intense.

One area where we have been seeing forward-thinking company executives take action is in the attempt to market-price employees, especially less experienced, fast-rising team members.  A mistake that is often made is to hire a very inexperienced person (often just out of college) at a low, but fair starting salary and then to increase that salary at the same rate as the overall average employee base raises.  As an example, many companies will hire a new marketing analytics team member at $45k and give them two years of 5% raises, but that means they are barely making $50k.  If they go to the market after two years of working at your hot start-up, with the great experience you gave them, what could they make?  $60k?  $75k?  The key is to ask yourself what you would pay to hire someone of a similar profile and make sure you are paying near enough to that level to make the employees’ decision to shift jobs to not really be about money.  Then if your culture is good, the upside is high, and the options have potential, these people are most likely to stick with you.

This is mostly an issue with younger hires, as the value of a person going from no experience to three years of experience rises at a very steep slope.  From three years to six years, the slope flattens.  Beyond six years, it is really a person-by-person and market-by-market decision, as at that point you pay for talent because most potential has been realized.  Average raises across your workforce just don’t work for the rise in value of a person in the early years of a successful growth curve. 

My advice for CEOs and functional leaders – make a print out of the cash compensation of everyone on your team on a quarterly or semi-annual basis and look at it over the last 8-12 quarters.  By the way – always look at total cash compensation in making this decision, as bonuses and commissions certainly do count.  Calculate the growth in compensation over time.  Ask yourself how valuable and how ambitious each person is.  Ask if you would have been happy with this level of salary acceleration at the same stage of your career.  Then make the appropriate adjustments.  Cash is king, salaries need to be held down as much as possible, and frugality of CEOs is a great quality – yet software companies have one real asset and that is people.  Grow, nurture, and protect your team.  Match salary to experience and value, not time, and keep the poachers at bay.