Thursday, July 31, 2014

Inability Of Organizations To Manage "The Flow" Of Talent Management


The flow, a concept developed by one of my favorite psychologists, Mihaly Csikszentmihalyi, matches the popular performance versus potential matrix that many managers use to evaluate and calibrate their employees. For people to be in the flow they need to be somewhere in the middle moving diagonally up. Ideally, this is how employees should progress in their careers but that always doesn't happen. To keep employees in the flow you want to challenge them enough so that they are not bored but you don't want to put them in a situation where they can't perform and are set up for a failure.

Despite of this framework being used for a long period of time I see many organizations and managers continue to make these three mistakes:

Mistaking potential for performance

Performance, at the minimum, is about given skills and experience how effectively person accomplishes his or her goals. Whereas potential is about what person could do if the person could a) acquire skills b) gain access to more opportunities c) get mentoring. We all have seen under-performers who have more potential. In my experience, most of these people don't opt to underperform but they are put in a difficult situation they can't get out of. We routinely see managers not identifying this as a systemic organizational problem but instead shift blame to employees confusing potential for performance suggesting to them, "you could have done so much but you didn't; you're a slacker." A similar employee with equal performance but less potential would not receive the same remarks on his/her performance.

Treating potential as an innate fixed attribute

One of the biggest misconceptions I come across is managers looking at potential as innate fixed attribute. Potential is a not a fixed attribute; it is something that you help people develop.

These out-performers who are not labelled as "high potential" are mostly rewarded with economic incentives but they don't necessarily get access to opportunities and mentoring to rise above their work and a chance to demonstrate their potential and make a meaningful impact.

Fixating on hi-potential out-performers

Not only managers fixate on hi-potential out-performers but they are also afraid that these employees might leave the organization one day if they have no more room to grow and if they run out of challenges. As counterintuitive as it may sound this is not necessarily a bad thing.

We all live in such a complex ecosystem where retaining talent is not a guarantee. The best you can do is develop your employees, empower them, and give them access to opportunities so that they are in a flow. As a company, create a culture of loyalty and develop your unique brand where employees recognize why working for you is a good thing. If they decide to leave you wish them all the best and invest in them: fund their start-up or make them your partners. This way your ecosystem will have fresh talent, place for them to grow, and the people who leave you will have high level of appreciation for your organization. But, under no circumstances, ignore the vast majority of other employees who could out-perform at high potential if you invest into them.

Monday, June 30, 2014

Chasing Qualitative Signal In Quantitative Big Data Noise


Joey Votto is one of the best hitters in the MLB who plays for Cincinnati Reds. Lately he has received a lot of criticism for not swinging on strikes when there are runners on base. Five Thirty Eight decided to analyze this criticism with the help of data. They found this criticism to be true; his swings at strike zone pitches, especially fastballs, have significantly declined. But, they all agree that Votto is still a great player. This is how I see many Big Data stories go; you can explain "what" but you can't explain "why." In this story, no one actually went (that I know) and asked Votto, "hey, why are you not swinging at all those fastballs in the strike zone?"

This is not just about sports. I see that everyday in my work in enterprise software while working with customers to help them with their Big Data scenarios such as optimizing promotion forecast in retail, predicting customer churn in telco, or managing risk exposure in banks.

What I find is as you add more data it creates a lot more noise in these quantitative analysis as opposed to getting closer to a signal. On top of this noise people expect there shall be a perfect model to optimize and predict. Quantitative analysis alone doesn't help finding a needle in haystack but it does help identify which part of haystack the needle could be hiding in.
"In many walks of life, expressions of uncertainty are mistaken for admissions of weakness." - Nate Silver
I subscribe to and strongly advocate Nate Silver's philosophy to think of "predictions" as a series of scenarios with probability attached to it as opposed to a deterministic model. If you are looking for a precise binary prediction you're most likely not going to get one. Fixating on a model and perfecting it makes you focus on over-fitting your model on the past data. In other words, you are spending too much time on signal or knowledge that already exists as opposed to using it as a starting point (Bayesian) and be open to run as many experiments as you can to refine your models as you go. The context that turns your (quantitative) information into knowledge (signal) is your qualitative aptitude and attitude towards that analysis. If you are willing to ask a lot of "why"s once your model tells you "what" you are more likely to get closer to that signal you're chasing.

Not all quantitative analyses have to follow a qualitative exercise to look for a signal. Validating an existing hypothesis is one of the biggest Big Data weapons developers use since SaaS has made it relatively easy for developers to not only instrument their applications to gather and  analyze all kinds of usage data but trigger a change to influence users' behaviors. Facebook's recent psychology experiment to test whether emotions are contagious has attracted a lot of criticism. Keeping ethical and legal issues, accusing Facebook of manipulating 689,003 users' emotions for science, aside this quantitative analysis is a validation of an existing phenomenon in a different world. Priming is a well-understood and proven concept in psychology but we didn't know of a published test proving the same in a large online social network. The objective here was not to chase a specific signal but to validate a hypothesis— a "what"—for which the "why" has been well-understood in a different domain.

About the photo: Laplace Transforms is one of my favorite mathematical equations since these equations create a simple form of complex problems (exponential equations) that is relatively easy to solve. They help reframe problems in your endeavor to get to the signal.

Saturday, May 31, 2014

Optimizing Data Centers Through Machine Learning

Google has published a paper outlining their approach on using machine learning, a neural network to be specific, to reduce energy consumption in their data centers. Joe Kava, VP, Data Centers at Google also has a blog post explaining the backfround and their approach. Google has one of the best data center designs in the industry and takes their PUE (power usage effectiveness) numbers quite seriously. I blogged about Google's approach to optimize PUE almost five years back! Google has come a long way and I hope they continue to publish such valuable information in public domain.



There are a couple of key takeaways.

In his presentation at Data Centers Europe 2014 Joe said:  
As for hardware, the machine learning doesn’t require unusual computing horsepower, according to Kava, who says it runs on a single server and could even work on a high-end desktop.
This is a great example of a small data Big Data problem. This neural network is a supervised learning approach where you create a model with certain attributes to assess and fine tune the collective impact of these attributes to achieve a desired outcome. Unlike an expert system which emphasizes an upfront logic-driven approach neural networks continuously learn from underlying data and are tested for their predicted outcome. The outcome has no dependency on how large your data set is as long as it is large enough to include relevant data points with a good history. The "Big" part of Big Data misleads people in believing they need a fairly large data set to get started. This optimization debunks that myth.

The other fascinating part about Google's approach is not only they are using machine learning to optimize PUE of current data centers but they are also planning to use it to effectively design future data centers.

Like many other physical systems there are certain attributes that you have operational control over and can be changed fairly easily such as cooling systems, server load etc. but there are quite a few attributes that you only have control over during design phase such as physical layout of the data center, climate zone etc. If you decide to build a data center in Oregon you can't simply move it to Colorado. These neural networks can significantly help make those upfront irreversible decisions that are not tunable later on.

One of the challenges with neural networks or for that matter many other supervised learning methods is that it takes too much time and precision to perfect (train) the model. Joe describing it as a "nothing more than series of differential calculus equations " is downplaying the model. Neural networks are useful when you know what you are looking for - in this case to lower the PUE. In many cases you don't even know what you are looking for.

Google mentions identifying 19 attributes that have some impact on PUE. I wonder how they short listed these attributes. In my experience unsupervised machine learning is a good place to short list attributes and then move on to supervised machine learning to fine tune them. Unsupervised machine learning combined with supervised machine learning can yield even better results, if used correctly.

Wednesday, April 30, 2014

Product Vision: Make A Trailer And Not A Movie


I have worked with many product managers on a product vision exercise. In my observation the place where the product managers get hung up the most is when they confuse product vision for product definition. To use an analogy, product vision is a trailer and product definition is a movie. When you're watching a movie trailer it excites you even though you fully don't know how good or bad the movie will be.

Abstract and unfinished

A trailer is a sequence of shots that are abstract enough not to reveal too much details about the movie but clear enough to give you the dots that your imagination could start connecting. Some of the best visions are also abstract and unfinished that leave plenty of opportunities for imagination. Product visions should focus on "why" and "what" and not on "how" and most importantly should have a narrative to excite people to buy into it and refine it later on. Vision should inspire the definition of a product and not define it.

I am a big believer of raw or low fidelity prototypes because they allow me to get the best possible feedback from an end user. People don't respond well to a finished or a shiny  prototype. I don't want people to tell me, "can you change the color of that button?" I would rather prefer they say, "your scenario seems out of whack but let me tell you this is what would make sense."

Non-linear narratives

Movie trailers are also the best examples of non-linear thinking. They don't follow the same sequence as a movie - they don't have to. Most people, product managers or otherwise, find non-linear thinking a little difficult to practice and comprehend. Good visions are non-linear because they focus on complete narrative organized as non-linear scenarios or journeys to evoke emotion and not to convey how the product will actually work. Clever commercials, such as iPad commercials by Apple, follow the same design principles. They don't describe what an iPad can do feature by feature but instead will show a narrative that help people imagine what it would feel like to use an iPad.

Means to an end

The least understood benefit of a product vision is the ability of using the vision as a tool to drive, define, and refine product requirements. Vision is a living artifact that you can pull out anytime during your product lifecycle and use it to ask questions, gather feedback, and more importantly help people imagine. I encourage product managers not to chase the perfection when it comes to vision and focus on the abstract and non-linear journey because a vision is a means to an end and not an end itself.

Photo courtesy: Flickr 

Monday, March 31, 2014

Amazon's Cloud Price Reduction, A Desire To Compete Hard And Move Up The Value Chain

Recently Google slashed price for their cloud offering. Amazon, as expected, also announced their 42nd price reduction on their cloud offerings since its inception. Today, Microsoft also announced price reduction for their Azure offerings.

Unlike many other people I don't necessarily see the price reduction by Amazon as waging a price war against the competition.

Infrastructure as true commodity: IaaS is a very well understood category and Amazon, as a vendor, has strong desires to move up in the value chain. This can only happen if storage and computing become true commodity and customers value vendors based on what they can do on top of this commodity storage and computing. They become means to an end and not an end itself.

Amazon is introducing many PaaS like services on top of EC2. For example, RedShift is the fastest growing service on EC2. These services create stickiness for customers to come back and try out and perhaps buy other services. These services also create a bigger demand for the underlying cloud platform. Retaining existing customers and acquiring new customers with as little barrier as possible are key components of this strategy.

Reducing hardware cost: The hardware cost associated with computing and storage have gradually gone down. Speaking purely from financial perspective existing assets depreciate before they are taken out from service. Also, new hardware is going be cheaper than the old hardware (at the original cost). If you do pass on the cost advantage to your customers it should help you reduce price and compete at the same or a little less margin. However, hardware cost is a fraction of overall operations cost. In the short term, Amazon being a growth company will actually spend a lot more on CapEx and not just OpEx to invest and secure the future.

Economies of scale: The cost to serve two computing units is not the sum of cost to serve two one computing units. There are many economies of scales in play such as increasing data-center utilization, investment in automation, and better instance management software. Confidence in predicting minimum base volume and reducing fluctuations also gives Amazon better predictability to manage elasticity. As the overall volume goes up the elasticity or the fluctuations as percentage of overall volume go down. On top of that offerings such as Reserved Instances also are a good predictor of future demand. Amazon views Reserved Instances as how banks view CDs but many customers are looking for a "re-finance" feature for these Reserved Instances when price drops. These economic and pricing implications are great to watch.

To offer competitive pricing to win against  incumbents and make it almost impossible for new entrants to compete on the same terms is absolutely important but it would be foolish to assume it is the sole intent behind the price reduction.

Photo courtesy: Flickr

Wednesday, March 12, 2014

Why And How Should You Hire A Chief Customer Success Officer?


For an ISV (Independent Software Vendor) it is everyone's job to ensure customer success but it is no one person's job. This is changing. I see more and more companies realizing this challenge and want to do something about it.

Sales is interested in maintaining relationship with customers for revenue purposes and support works with customers in case of product issues and escalations. Product teams behave more like silos when they approach their customers because of their restricted scope and vision. Most chief technology officers are fairly technical and internal facing. Most of them also lack the business context—empathy for true business challenges—of their customers. They are quite passionate about what they do but they invariably end up spending a lot of time in making key product and technical decisions for the company losing sight of much bigger issues that customers might be facing. Most chief strategy officers focus on company's vision as well as strategy across lines of businesses but while they have strong business acumen they are not customer-centric and lack technical as well as product leadership to understand deep underlying systemic issues.

Traditional ways to measure customer success is through product adoption, customer churn, and customer acquisition but the role of a Chief Customer Success Officer (CCO) extends way beyond that. One of the best ways to watch early signs of market shift is to very closely watch your progressive customers. Working with these customers and watching them will also help you find ways to improve existing product portfolio and add new products, organically or through acquisitions. Participating in sales cycles will help you better understand the competition, pricing points, and most importantly readiness of your field to execute on your sales strategy.

I often get reached out by folks asking what kind of people they should be looking for when they plan to hire a CCO. I tell them to look for the following:

T-shaped: Customer don't neatly fall into your one line of business and so is your CCO. You are looking for someone who has broad exposure and experince across different functions through his or her previous roles and deep expertise in one domain. The CCO would work across LoBs to ensure customers are getting what they want and help you build a sustainable business. Most T-shaped people I have worked with are fast-learners. They very quickly understand continuously changing business, frame their point of view, and execute by collaborating with people across the organization (the horizontal part of T) due to their past experience and exposure in having worked with/for other functions.

Most likely, someone who has had a spectacular but unusual career path and makes you think, "what role does this person really fit in?" would be the the right person. Another clue: many "general managers" are on this career track.

Business-centric: Customers don't want technology. They don't even want products. They want solutions for the business problems they have. This is where a CCO would start with sheer focus on customers' problems—the true business needs—and use technology as an enabler as opposed to a product. Technology is a means to an end typically referred to as "the business."

Your CCO should have a business-first mindset with deep expertise in technology to balance what's viable with what's feasible. You can start anywhere but I would recommend to focus your search on people who have product as well as strategy background. I believe unless you have managed a product—development, management, or strategy—you can't really have empathy for what it makes to build something and have customers to use it and complain about it when it doesn't work for them.

Global: Turns out the world is not flat. Each geographic region is quite different with regards to aptitude and ability of customers to take risk and adopt innovation. Region-specific localization—product, go-to-market, and sales—strategy that factors in local competition and economic climate is crucial for global success of an ISV. The CCO absolutely has to understand intricacies associated with these regions: how they move at different speed, cultural aspects of embracing and adopting innovation, and local competition. The person needs to have exposure and experience across regions and across industries. You do have regional experts and local management but looking across regions to identify trends, opportunities, and pace of innovation by working with customers and help inform overall product, go-to-market, and sales strategy is quite an important role that a CCO will play.

Outsider: Last but not least, I would suggest you to look outside instead of finding someone internally. Hiring someone with a fresh outside-in perspective is crucuial for this role. Thrive for hiring someone who understands the broader market - players, competition, and ecosystem. This is a trait typically found in some leading industry analysts but you are looking for a product person with that level of thought leadership and background without an analyst title.

About the photo: This is a picture of an Everest base camp in Tibet, taken by Joseph Younis. I think of success as a progressive realization of a worthwhile goal.

Friday, February 28, 2014

Recruiting End Users For Enterprise Software Applications

As I work with a few enterprise software start-ups I often get asked about how to find early customers to validate and refine early design prototypes. The answer is surprisingly not that complicated. The following is my response to a recent question on Quora, "How do we get a target audience for enterprise applications, when you dont have an enterprise customer yet for rapid prototyping?"

Finding a customer and finding end users are quite different. In enterprise software end users are not the buyers and the buyer (customer) may or may not use your software at all. To recruit end users, there are three options:

Friends and families: Use your personal connections through email and social media channels and ask for their time (no more than 30 minutes) to conduct contextual inquiries and get validation on your prototypes. Most people won't say no. Do thank them by giving them a small gift or a gift card.

Find paid end users: This does seem odd but it works. I know of a few start-ups that have used this method effectively. Use Craigslist and other means to recruit people that match your end user role and pay them to participate in feedback sessions. It is worth it.

Guerrilla style: Go to a convention or a conference where you could find enough end users that fit your profile. Camp out at the convention with swag and run guerrilla style recruiting to validate and prototype. Iterate quickly, preferably in front of them, and validate again.