In Data Analytics We Trust?

One of the recurring themes in this community is how data and analytics hold great potential to uncover insight to give us direction. For businesses to adopt the means of achieving more data-driven decisions would seem to be an obvious choice. So why don’t they all do it? Some just don’t trust the results. That’s the finding of KPMG’s report Building trust in analytics: Breaking the cycle of mistrust in D & A.

Credit: Pixabay
Credit: Pixabay

To discover the extent of trust in data and analytics, KPMG International commissioned Forrester Consulting to assess it based on responses to a survey that included over 2,000 organizations around the globe. The report took into account the state of four key anchors of trust: quality, effectiveness, integrity, and resilience.

With regard to the four anchors of trust, numbers tended to be on the low side. For resilience, 20 percent of respondents expressed confidence in data and analytics security, and 18 percent in governance. There was more divergence for rankings on integrity. Just 13 percent expressed faith in their organization’s privacy and ethical use of data and analytics, but 44 percent were confident about regulatory compliance.

The numbers were lower for trust in quality. A mere 10 percent of organizations expressed faith in the quality of their organization’s data, tools, and methodologies, and 22 percent expressed confidence in the of capabilities of D & A, Only 16 percent expressed confidence in the accuracy of their analytics models, and 20 percent said they trusted the utility of models and processes.

The majority -- 60 percent -- indicated they are not very confident in the D & A insight of their organization. The report attributes that lack of confidence to a lack of understanding. Decision makers who don’t understand what goes into D & A are apt “to miss the full value of their investments and assume that a large proportion of their D & A projects don’t work.” Consequently, it is essential for organizations to set up ways “to assess and validate the effectiveness of their analytics in supporting decision-making” to foster understanding and appreciation for what they gain from D & A.

But the report is not only concerned with the decision makers’ trust. Because data analytics are expected to become central to more processes in future, it is necessary to gain trust from all who are touched by it. Accordingly, Sander Klous, a partner with KPMG in the Netherlands, is quoted as saying, “We need to find ways to establish societal trust in how organizations operate in the emerging data-driven society."

He added that it is necessary to win over trust to data and analytics, though it won’t be easy. He references a saying of his native country, “Trust comes on foot and leaves on horseback." So there is no easy fix, but the report does hold out hope, offering seven key recommendations for enhancing trust in D & A.

They are:

  • Start with the basics: assess your trust gaps.
  • Create purpose: clarify and align goals.
  • Raise awareness: increase internal engagement.
  • Build expertise: develop an internal D&A culture and capabilities as your first guardian of trust.
  • Encourage transparency: open the "black box" to a second set of eyes; and a third.
  • Take a 360-degree view: build your ecosystems, portfolios, and communities.
  • Be innovative: enable experimentation.

The last point relates to the interview with Dr. Mark Kennedy of Imperial College, London, which was cited in the report. He offers this insight:

    I often tell executives that if they aren’t setting aside enough people and resources to try new things, they’re going to be permanently stuck in yesterday. You need to accept a level of risk and conduct the kinds of trials that start to increase learning and build trust.

What can your organization do to ensure that decision makers and the others who touch data have trust in it?

Ariella Brown,

Ariella Brown is a social media consultant, editor, and freelance writer who frequently writes about the application of technology to business. She holds a PhD in English from the City University of New York. Her Twitter handle is @AriellaBrown.

Emojis Train AI to Recognize Sarcasm

It's sometimes difficult for humans to understand tone and sarcasm in text messages, so consider how difficult it is to train computers to understand it. But a new project is employing emojis to help AI understand when humans are employing irony.

Hiring Trend: Machine Learning Reduces Bias, Increases Applicant Pool

Unilever is using machine learning tools to increase the pool of entry level applicants and then test and evaluates candidates in the pool to cull those who are the best fits. Adding machine learning tools reduces bias, increases diversity, and makes the whole process more efficient, according to Unilever.

Risk and innovation
  • 11/17/2016 4:22:13 PM

Your last point about experimentation/innovation and developing more trust and risk tolerance is a good one, Ariella. Unfortunately, it's at odds with what CXOs and directors are charged with doing, which is to reduce risk or build a safety net that allows them to assume greater risk in their lines of business. Most executives will push away that challenge... "too risky."

Re: Trust and consequenxe
  • 11/16/2016 9:17:32 PM

@kq4ym Indded, but the people managing the campaigns should know better than to trust polls. That's something I already mentioned in the comments on Tricia Aanderud's blog, "The Election Aftermath: Is Data Dead?"

Re: Trust and consequenxe
  • 11/16/2016 8:56:22 PM

At least, we can say the data indicates there will be some turkeys that may survive being served at Thanksgiving tables. But, our recent experience surely is going to lead to some untrusting of lots of data and analytic studies. One experience that defies our expectations can really lead to a distrust of outcome predictions for the future.

Re: Trust and consequenxe
  • 11/15/2016 8:54:54 PM

<<sigh> That will teach them to enlarge their sample beyond the vegan community.>

LOL @Terry! Ah, yes, lasagna seems to rank as the official vegetarian alternative to turkey. 

Re: Trust and consequenxe
  • 11/15/2016 8:48:33 PM

<sigh> That will teach them to enlarge their sample beyond the vegan community.

Re: Trust and consequenxe
  • 11/15/2016 8:46:22 PM

Sorry, Maryam... data's actually having a great year. People, on the other hand, are having a bad year trusting data and those who collect, analyze and interpret it!

Re: ID trust gaps ...
  • 11/15/2016 8:40:53 PM


Re: Trust and consequenxe
  • 11/15/2016 8:06:29 PM

@Maryam to speak to that feeling, I just saw this cartoon shared on Google+

Re: ID trust gaps ...
  • 11/15/2016 8:01:53 PM

@tomsg rather like the unknown unknowns that are challenging to take into account in predictive analytics.

Re: ID trust gaps ...
  • 11/15/2016 7:50:02 PM

I agree. The trust gaps that are not stated, but somehow acted upon are clearly the most insidious ones.

Page 1 / 2   >   >>