Thursday, December 5, 2013

Don't Waste Money on Analytics

Image courtesy of reynermedia
Today I'm pleased to present a guest post by Gregory Yankelovich.

Analysis is an instrument of learning, defined as "a process of acquiring modifications in existing knowledge, skills, habits, or tendencies." There are substantial volumes of academic research produced over the years on a subject of relationship between learning and beliefs. You can search for them with the keywords "deep learning," "deep belief networks," etc. The gist of these academic inquiries points to an observation that deeply-held beliefs impede learning process, and network- (think groups or organizations) shared beliefs have a tendency to suppress learning process aggressively.

As long as you and/or your executives are not ready to question your current beliefs, no amount of evidence will make you or them to act.

The power of analytics is in its ability to expose patterns of data that can help us to learn. When new knowledge is rejected/ignored by the organizational belief system, all the cost of learning is wasted. If you think the last sentence does not apply to you because you use "free" tools, think again. The time, effort, and political capital you have to invest in the use of "free" tools for learning can be substantial. The probability of acceptance by your "network" of new knowledge, discovered with the use of "free" tools, is even lower.  That is because you bypassed an opportunity to (a) socialize the idea that you may discover something your organization does not know yet and (b) gain conceptual adoption of such result.
 

Use of "free" tools rarely require any approval process within organization. Therefore nobody knows what you are doing and as a result are not prepared to consider any findings, unless they support and re-enforce existing beliefs. Presenting new findings that challenge status quo as a surprise is a very bad idea. The process of selection and acquisition of a tool prepares your audience to consider the findings, as participation exposes them to a potential value.

People and organizations are most likely to consider a challenge to their beliefs at the times of extreme "pain." At such times, leaders open their minds and examine their beliefs to learn how they need to act to improve their lot; the rest are looking for excuses and complain about circumstances beyond their control. Here is an example describing such a moment at Best Buy in 2012:

The one critical thing we offer the world is choice,” said the Best Buy chief executive officer Brian Dunn in a March 2012 phone interview. He was trumpeting in particular his company’s role in guiding customers through the expanding smartphone universe.

“We provide the latest and greatest choice of all technology gear, from Apple products to Google products, and that brings more opportunity to help people put technology to use. That is a great place for us to be.” A week later, reality intruded. The consumer electronics retailer posted a $1.7 billion quarterly loss and announced it would close 50 stores nationwide. On Tuesday, Dunn resigned.

The belief of the Best Buy CEO (at the time) - "The only critical thing we offer the world is choice" - was challenged by customer intelligence that exposed the evidence of "most critical" things from "the world" perspective are in-store service quality and products reliability. The evidence was ignored and a new CEO had to come in.

Here are a few ideas on how to deal with this challenge:

1. If you focus on intelligence that can help to improve the probability that the enterprise will increase its market share, your challenge to status quo is more likely to be tolerated. Business executives are motivated by two desires:
  • An increase in revenue or market share and
  • A reduction of expenses, i.e., increase of profit margin
and two fears:
  • Decrease in revenue or market share and
  • Decrease of profit margin.
Intelligence that improves probability of realizing their desires, and/or forewarn that they are on the path of realizing their fears, is aligned with their system of values and therefore deserves their attention.

2. New knowledge that does not conform our beliefs is a natural suspect.  We credit our beliefs with helping us to achieve our past successes, while new intelligence has no "resume." Applying the new intelligence to historic data can overcome the trust challenge if that application successfully exposes patterns that correlate with actual results in the past.

Gregory Yankelovich is a Technologist who is agnostic to a technology, but "religious" about Customer Experience and ROI. He has solid experience delivering high ROI projects with focus on both Profitability AND Customer Experience improvements, as one without another does not support long-term business growth. Gregory currently serves as the Customer Experience (CX) Whisperer at Amplified Analytics, provider of software as a service for Customer Experience Measurement to support Strategic Marketing and Brand/Category/Product Management applications.

9 comments:

  1. You make a good point that challenging strongly held beliefs can be difficult. Sometimes, it takes a equally strong emotional experience.

    For example, a client of mine was rapidly outgrowing their office space. Their building contained an empty suite that could easily be leased, but the CEO wouldn't make the move even though she had all the data she should have needed to make the decision. One day, the VP of HR invited the CEO to stop by new employee orientation. It was literally standing room only since they didn't have enough space to comfortably accomodate the new hires. Immediately after the meeting, the CEO decided to lease the extra office space. The experience of seeing new employees have a less than stellar welcome was what was needed to convince her.

    ReplyDelete
    Replies
    1. Jeff, you are right, personal emotional response can be very powerful. However, not every business executive has an easy way to share their customers' experiences. That is why the importance of exposing authentic experiences of customers behind the numbers, cannot be overestimated.

      Delete
  2. You make a great point Gregory, as a man who has "Analyst" in his job title the most frustrating part of my job is trying to get people to debunk their beliefs and look at what the data is telling them.

    "There are non so blind as those who will not see"

    ReplyDelete
    Replies
    1. James, I share your pain. As frustrating as it may be, we should not underestimate value of beliefs either. Core beliefs anchor us in the seas of changes and allow us to make consistent decisions. Perhaps not right ones, but consistent :). So your choice is to be frustrated with this reality or master the skill of persuasion with quality evidence that resonates with your audience.

      Delete
  3. Well said! This is a key observation that I make in my book "Business unIntelligence-Insight and Innovation Beyond Analytics and Big Data". We, as humans, make decisions based on many factors, and the information that we receive from our BI, and now analytics, tools is but one (small) part of the input to deciding - and one that can be easily ignored. See also my recent blogs for further thoughts: http://bit.ly/1c4JXY8 and http://bit.ly/189jf1i ...

    ReplyDelete
    Replies
    1. Barry, in my opinion the most disturbing of all "bubbles" in decision making process is the one that imply that what we don't find on a surface, does not exist. Amazingly high % of people think that if they cannot find something by searching Google, that means this something does not exist. My concern is that proliferation of BI/Analytics tools propagates unlearning of inquiry skills by replacing quality of insights with enormous quantity of meaningless data points. The velocity of data is so overwhelming, that most people don't even consider to challenge it's authenticity or meaning.

      Delete
  4. Hi Gregory,
    I'm a big fan of analysis and using data to unearth truths and change perspectives.

    But, whilst many executives will want to increase revenues and profits etc, I think these motivations may be secondary to other, more personal, motivations like:
    - Make me look good/Don't make me look bad
    - Get me noticed
    - Don't get me fired
    - Help me achieve my targets
    - Don't make me look stupid etc etc

    Call this cynical, if you like. But, I believe, that if you think about this you'll agree that there is some truth, to varying degrees, in what I say. So, if we want to use analysis and analytics more widely to help drive higher revenues, profitability and a better customer experience, we probably need to figure out what an executives personal motivation is first before we can figure out how we can help them become more receptive to data and analysis.

    I think Jeff's CEO example falls into the 'make me look/feel bad' category.

    What do you think?

    Adrian

    ReplyDelete
    Replies
    1. Adrian - I don't think your observation is cynical at all. Rather, it's a pragmatic look at how many executives (and people) naturally make decisions.

      In his post, Gregory makes a good point that confirmation bias is an obstacle can prevent executives from using data wisely. You've added to that list of obstacles nicely. My takeaway is good data alone isn't always enough to make a convincing argument.

      Delete
    2. Adrian, I agree with Jeff. Additionally, most of personal motivations you pointed to are completely aligned with revenue/profits motivations. Consider that most senior executives don't look good and get fired when they do not live up to their board's expectations of revenue/profits delivery.

      Delete