MIT SMR CONNECTIONS

M A N AG E M E N T B R I E F I N G

Transform Your Business:

Build and Maintain a Data-Driven Culture

ON BEHALF OF:
 MANAGEMENT
MANAGER’S
BRIEFING | TRANSFORM
GUIDE — TRANSFORM
YOURYOUR
BUSINESS:
WORKFORCE
BUILD AND
WITH
MAINTAIN
SKILLS A
FOR
DATA-DRIVEN
MACHINE LEARNING
CULTURE

Transform Your Business:

Build and Maintain a Data-Driven Culture
Leveraging data for decision-making enables agility while driving innovation and business results. But becoming
genuinely data driven requires significant cultural change — and involves an ongoing journey. Here’s where to start.

It’s hard for any organization to sustain success indefinitely. To stay The (Ongoing) Journey to Data-Driven Decision-Making
relevant, organizations must periodically reinvent themselves. The Ultimately, becoming data driven is all journey and no destination,
introduction of cloud computing set off a generation of reinvention. according to Piyanka Jain, president and CEO of Aryng, a data con-
Now the next wave of reinvention is clearly being driven by data. sulting company. In other words: It’s a constant work in progress.
Leaders need to be able to rely on solid, meaningful data to make
decisions now and prepare for what’s ahead. And data maturity can be rated on a continuum ranging from “digital
naïve” up to “digital elite,” she adds. “Naïves are beginning to wake
Doing that requires building a data-driven organization. Essentially, up to the fact that they have such large amounts of data sitting in
such organizations treat data like an organizational asset, no longer silos,” says Jain, whose books include Behind Every Good Decision:
the property of individual departments. They set up systems to col- How Anyone Can Use Business Analytics to Turn Data Into Profit-
lect, store, organize, and process valuable data and make it available able Insight (AMACOM, 2014). “And people have been using it, but in
in secure ways to the people and applications that need it. Then — nooks and crannies. Can it be combined to make some larger deci-
and here’s the key — they use that data to inform those all-import- sions?” On the other end of the spectrum are the elites, those using
ant business decisions. data as their lifeblood. “They live by dashboards. They live by 0.25%
optimization. They live by real-time data,” she says.
But it’s not always easy to shift to (and then maintain) data-driv-
en decision-making, especially at large companies that grew and Making decisions based on data provides a better way to achieve
thrived with data siloed by function or department. In this year’s organizational goals, according to Richard Huntsinger, executive di-
NewVantage Partners survey of executives from Fortune 1000 and rector of the UC Berkeley Data Analytics Group and a faculty mem-
industry-leading companies, fewer than half said they are driving ber in the UC Berkeley Haas School of Business. The first goal, of
innovation with data, and less than one-quarter said their organiza- course, is usually to increase profits. But there are others, includ-
tion is driven by data. Both findings represent decreases from levels ing driving revenue growth, improving brand awareness, boosting
reported in 2019 and 2020. (For more on the survey results, see “Still customer and employee satisfaction, creating a better customer
Striving for Data Mastery.”)

“Leading companies are struggling to become data driven,” says

Randy Bean, CEO of NewVantage Partners. Part of the decline is due Agility is another significant benefit of
to the evolution of a more realistic view of what it means to be data
data-enabled decision-making: “You
driven. Also, data volumes continue to proliferate along with new
sources of data. “More and more information keeps hitting these or-
can change direction quickly because
ganizations. As quickly as they get good enough to handle one par- you know earlier on when you’re wrong,”
ticular thing, five new things come at them,” says Bean, who is also says data consultant Piyanka Jain.
the author of Fail Fast, Learn Faster: Lessons in Data-Driven Leader-
ship in an Age of Disruption, Big Data, and AI (Wiley, 2021).

MIT SMR CONNECTIONS 1

 MANAGEMENT
MANAGER’S
BRIEFING | TRANSFORM
GUIDE — TRANSFORM
YOURYOUR
BUSINESS:
WORKFORCE
BUILD AND
WITH
MAINTAIN
SKILLS A
FOR
DATA-DRIVEN
MACHINE LEARNING
CULTURE

experience, increasing innovation, and reducing costs. Participants

Still Striving for Data Mastery
in the NewVantage survey also noted that they are using big data
and AI investments to support revenue-generating capabilities, Executives at large enterprises report that they aren’t yet
fully tapping the power of their data. Of respondents from
such as customer development.
85 major companies who participated in NewVantage
Partners’ latest survey:
Agility is another significant benefit of data-enabled decision-mak-
ing. “You can change direction quickly because you know earlier on • 48.5% are driving innovation with data.
• 41.2% are competing on analytics.
when you’re wrong,” says Jain. Let’s say you launched a product in a
• 39.3% are managing data as a business asset.
new market and, based on data-driven planning, you thought you’d • 30.0% have a well-articulated data strategy for
see sales increase by a certain percentage within three months. “If their company.
you don’t see that, that means you were wrong — and that’s fine,” she • 29.2% are experiencing transformation
business outcomes.
says. “You can correct your course faster.”
• 24.4% have forged a data culture.
• 24.0% have created a data-driven organization.
It’s Not Intuition vs. Data
SOURCE:
A common misunderstanding is that it’s preferable to choose data NewVantage Partners Big Data and AI Executive Survey, 2021
over human intuition, knowledge, or expertise. That misconception
often generates significant resistance to establishing a data-driven
culture. In general, younger workers tend to understand intuitively
the value of data-driven decision-making, says Huntsinger. Mean-
while, older people tend to have more experience — and they tend
Four Ways Data-Driven
to value that experience more highly.
Organizations Use Data
Richard Huntsinger, a professor at UC Berkeley’s Haas
But being data driven doesn’t pit data against intuition, says Jain: School of Business, sees companies leveraging data for
the following purposes:
“It is data and intuition, because intuition drives hypotheses.” The
people with the most significant experience should be vital sources 1. Creating new products and services and innovating
for hypotheses that can be tested via data analysis. existing offerings.
2. Improving and streamlining internal operations.
3. Improving intelligence about the market and themselves.
Being adept at data enables institutional intuition and experience to 4. Extracting value from the data in its own right —
be captured and used at scale, says Jain. “Now this knowledge can by selling it, for example.
percolate throughout the organization because it’s being learned
also through data. And it’s getting into the organizational fabric, not
just living within one person.”
obstacles as the biggest impediment to successful adoption of data
Data Silos: The ‘Monster Under the Bed’ initiatives. Employees, especially at large legacy companies, are
When shifting to being data driven, culture presents much more often reluctant to share data and stop operating in silos, which
of a challenge than technology. For the fifth consecutive year, re- grew up because they were an efficient way to organize work. Joe
spondents to the NewVantage survey —76% of whom held the title Hilleary, a research analyst with Eckerson Group, calls silos the
chief data officer (CDO) or chief analytics officer — cited cultural “monster under the bed” in enterprise organizations. Huntsinger

Being able to share data across organizations has driven transformation in many
industries. Life sciences companies, for instance, have reaped unforeseen
benefits by sharing data across vastly different product discovery lines that
would traditionally have been siloed.

MIT SMR CONNECTIONS 2

 MANAGEMENT
MANAGER’S
BRIEFING | TRANSFORM
GUIDE — TRANSFORM
YOURYOUR
BUSINESS:
WORKFORCE
BUILD AND
WITH
MAINTAIN
SKILLS A
FOR
DATA-DRIVEN
MACHINE LEARNING
CULTURE

product discovery lines that would traditionally have been siloed.

“In the data-intensive verticals like pharma, finance, health care, you
Data-Driven Decision-Making in Action
can’t compete anymore unless you’re at the cutting edge of data,”
How data helped a retailer beat its competition says Hilleary. While many companies in these industries did not be-
despite the pandemic.
gin life as “data-first” entities, they have begun to take their cues
It used to be that grocery stores did planning entirely by from their most data-forward competitors. Ignoring the strategic
looking in the rearview mirror — stocking their shelves based importance of data is not an option.
on historic sales data: “If it happened last year, it will happen
this year, too.” But the flaws in such thinking became painful-
ly obvious as the COVID-19 pandemic unfolded in early 2020, The advent of the chief data officer title in the 2010s marked an el-
and certain products became all-but-impossible to obtain.
evated recognition of the importance of data, says Bean. “Now most
Rather than simply making assumptions about what shop- businesses know data is an asset that flows across all of an orga-
pers want, today’s data-driven retailers are recording, mea- nization. Companies need to think of how to harness that asset to
suring, and analyzing actual customer shopping patterns
throughout the year. So rather than having a set floor plan manage it most effectively.”
and product mix in all locations, for example, companies
expect both to change frequently based on actual up-to- Increasingly, the CDO is either a peer of the CIO or reports in to the
date traffic and purchasing patterns.
business, says Bean. “That’s meant to create that connection between
Eckerson Group research analyst Joe Hilleary points to the business value and outcomes, as opposed to IT simply creating a ca-
example of a higher-end grocer that was already closely
pability and then waiting for somebody to discover it,” says Bean.
tracking its stores’ sales data before the pandemic start-
ed — which was, of course, also before shoppers began
rushing to stock up on products such as toilet paper, bleach Data-Driven Customer Experience
wipes, and hand sanitizer. As the earliest product shortages
emerged, the company was able to respond quickly because Being a data-driven organization is the foundation to providing ex-
it had systems in place to quickly gather, store, analyze, and cellent customer experience (CX), says Hilleary. Take the classic ex-
report on that data. So in those first few weeks, company ample of calling the customer service number for your bank. “Now
buyers were able to spot what turned out to be the leading
indicators of a likely run on essential goods in their stores. as soon as I call in, they read my phone number, they know who I am,
they pull up my information. And by the time that I get connected
“They were able to put in orders to restock [these hot
to a person, they already have it in front of them. That improves my
commodities] before any of their competitors because their
data and business were aligned in such a way that busi- customer experience and is only possible because of data strategies
ness decisions around purchasing and stocking could be like Customer 360, things that enable them to bring all of my data
made based on a near real-time store data,” he says. Data
effectively allowed the retailer to jump the line ahead of the together,” he says. A major improvement over the days when calling
other retailers that didn’t see the shortages coming. “They a bank meant bouncing from one rep to another, this customer ex-
managed to keep their shelves full in those first few months perience is now table stakes in financial services.
when folks were out of everything,” says Hilleary. The bot-
tom line? Increased sales.
Layering technologies like ML and the internet of things (IoT) on top
of customer data platforms (CDPs) and customer relationship man-
agement (CRM) systems opens exciting new vistas for CX. iRobot is
joked that when he worked at nearly 100-year-old HP, his col-
already using IoT in its Roomba robotic vacuum cleaners. They con-
leagues referred to him as the SEM (silo elimination manager).
stantly report data back to iRobot. Before long, the company will have
the ability to know exactly what the problem is before a customer calls
Dismantling silos and propagating data throughout a company al-
lows the organization to view data as an asset, not just an artifact of
doing business. Successful data-driven organizations view data as
a strategic asset to be managed and leveraged toward the achieve- What Is a Data-Driven Organization?
ment of business objectives. According to Randy Bean, CEO of NewVantage Partners,
data-driven organizations combine timely data with
Being able to share data across organizations has driven transforma- human experience and intuition to improve business
decision-making.
tion in many industries. Life sciences companies, for instance, have
reaped unforeseen benefits by sharing data across vastly different

MIT SMR CONNECTIONS 3

 MANAGEMENT
MANAGER’S
BRIEFING | TRANSFORM
GUIDE — TRANSFORM
YOURYOUR
BUSINESS:
WORKFORCE
BUILD AND
WITH
MAINTAIN
SKILLS A
FOR
DATA-DRIVEN
MACHINE LEARNING
CULTURE

The handwriting is on the wall. Sooner or stakeholders and IT leaders. The council should formulate and com-
municate the organization’s strategic plan for data.
later, you’re going to have to infuse data
into your culture and your decisions. “It’s really a dialog back and forth where the data folks are learning
from business folks about what is relevant to their decision-making
process,” says Hilleary. The next step is figuring out how to deliver
that high-value data to the business in a timely way.

to report an issue. The agent could have an answer ready without

needing to send a technician to look at the unit — scoring convenience 2. Emphasize data literacy. As with teaching children to read, you

for the consumer and lower cost for iRobot. With the problem solved, can’t just tell people they need to be data literate (or, to use a term

the manufacturer can then use the data to enrich its future models. that some experts prefer, data proficient). “It doesn’t happen over-
night,” says Jain. “When you’re teaching reading, you start with pic-

Being able to leverage data in ever more creative ways is key to de- ture books, getting children excited about the colors and pictures.”

livering experiences that continue to delight customers, which is

important as customer expectations grow inexorably. “The first few The same is true of teaching data literacy. Jain recommends piqu-

times I called my airline and asked a question about my ticket and ing employees’ interest with real examples of data-driven success-

they already knew what I was calling about, it was pretty incred- es. “Show them an end-to-end process” of what data-enabled deci-

ible,” says Hilleary. “But humans adapt very quickly, and expecta- sion-making looks like in action, she says.

tions increase. Now, if I call a business and they don’t know who I
am immediately, I start to say, ‘OK, they’re not up to my baseline Get them excited about the big picture, but also introduce them to

expectations anymore’ and I go somewhere else.” That’s just one il- the smaller things they can do. Give them incremental doses of how

lustration of the potential harm of not being data driven. (To learn to think and then act with data, and then lead with insights. It’s a

how one data-driven retailer met demand during the pandemic, see progressive journey from “data skeptic” to “data enthusiast” to “data

“Data-Driven Decision-Making in Action.”) literate” and “data-driven manager.”

The handwriting is on the wall. Sooner or later, you’re going to have 3. Focus on the best data-use cases. The power of data lies in its

to infuse data into your culture and your decisions. Confused about power to help answer your organization’s burning questions.

where to begin? Consider taking these four steps: As a first step, focus on what Bean of NewVantage Partners calls
“high-value business use cases.”

1. Convene a data council. You can’t make data-based decisions until

you know what kind of data your organization has, where it resides, “Start with your most critical metrics, your key questions and deci-

and how people are using it. Hilleary, of Eckerson Group, advises put- sions that need to be made, and then identify the data associated with

ting together a data council whose role it is to coordinate between these,” says Bean. From there, identify and focus on quick wins. Re-

business and IT. Under the auspices of the CDO or another data-flu- sults will help establish credibility within the business. “That starts to

ent executive-level officer, the council should consist of business really build the momentum, which sets the stage for a strong business
and technology partnership around data,” he says.

4. Avoid complacency. Again, there’s no end destination to being a

What Is a Data Catalog? data-driven organization; the target moves all the time. “The best
firms, those that have been successful at developing data-driven
A key aspect of treating data as an asset is being able to
organizations and data cultures, are never satisfied. They’re per-
store it in such a way that it is searchable within the orga-
nization. Data catalogs connect meta descriptions of data petually restless and never think what they’ve been doing is good
with the original sources of data, making the data more enough,” says Bean. On the other hand, there are also organizations
easily accessible for decision-making while restricting that say they have everything under control. “That’s when I worry,”
data access to authorized users.
he says. “The best data-driven firms are never complacent. They’re
always worried. They’re always striving to be better.” l

MIT SMR CONNECTIONS 4

 MANAGEMENT
MANAGER’S
BRIEFING | TRANSFORM
GUIDE — TRANSFORM
YOURYOUR
BUSINESS:
WORKFORCE
BUILD AND
WITH
MAINTAIN
SKILLS A
FOR
DATA-DRIVEN
MACHINE LEARNING
CULTURE

S P O N S O R ’S V I E W P O I N T

Creating a Data-
Driven Culture:
4 Key Actions
Ishit Vachhrajani is an
enterprise strategist with This conversation has been edited for clarity, length, and editorial style.
AWS. In this role, he helps
enterprise executives
As an enterprise strategist at AWS, I have the privilege of working with hundreds of execu-
by sharing strategies
to increase speed and tives from some of the world’s largest companies and helping them with their digital trans-
agility, drive innovation, formation journeys. I often find that executives need no convincing that data is a strategic
and create new operating
asset. Instead, they need help unleashing its value at scale to reinvent their businesses. They
models using the cloud to
transform their business- tell us that many of their challenges are cultural. At Amazon, we care deeply about innovating
es. Before joining AWS, on behalf of our customers, and we believe a data-driven culture is necessary for a successful
he was chief technology
digital transformation.
officer at A+E Networks,
responsible for global
technology across cloud, Organizations can create more data-driven cultures by taking four key actions:
architecture, enter-
prise applications and
products, data analytics, Engage: Culture change must start at the top. While executive sponsorship is necessary, it is
technology operations, not sufficient. The entire senior leadership must remain visibly engaged beyond just approv-
and cybersecurity. He
ing investments. The organization then needs to create mechanisms to scale this engagement
previously held technolo-
gy leadership positions at to support data-driven decision-making at all levels. Executives must analyze how decisions
NBCUniversal and various are made today and ensure that employees throughout the organization use data to guide
consulting organizations.
their decisions, not just to support them after the fact.

Enable: It’s also essential to make sure that front-line workers can use the data. Democratization
of data is not just about providing access to data, but about democratizing decisions and
data and actions as well. It’s not about bringing data to the decision makers, but instead about
analytics decentralizing the decision-making to bring it closer to the data. In a data-driven culture,
employees use data for everyday choices, not just big decisions.

Data is empowering, but it can also evoke strong

emotions. To eliminate resistance, it’s important
to create a strong change management strategy
that focuses on communication and makes the
use of data not only easier but enjoyable.

MIT SMR CONNECTIONS 5

 MANAGEMENT
MANAGER’S
BRIEFING | TRANSFORM
GUIDE — TRANSFORM
YOURYOUR
BUSINESS:
WORKFORCE
BUILD AND
WITH
MAINTAIN
SKILLS A
FOR
DATA-DRIVEN
MACHINE LEARNING
CULTURE

Educate: A data-driven culture reflects the belief that data profi-

ciency is a core skill for everyone in the organization, not just for ABOUT AWS
those in specialized roles such as data analysts or scientists. This
means investing in education across all functions, in new roles, For more than 15 years, Amazon Web Services has been the
and in capabilities such as storytelling, visualization, and the world’s most comprehensive and broadly adopted cloud
translation skills that connect the art of business with the science offering. AWS has been continually expanding its services to
of data. support virtually any cloud workload, and it now has more than
200 fully featured services for compute, storage, databases,
Eliminate: Perhaps the most critical part of creating a data-driv- networking, analytics, machine learning and artificial
en culture is removing the cultural barriers that protect data intelligence, Internet of Things, mobile, security, hybrid, virtual
silos. A data-driven culture treats data as an organizational asset, and augmented reality, media, and application development,
not as a departmental property. This approach can be achieved by deployment, and management from 81 Availability Zones (AZs)
a top-down commitment, establishing governance with a goal to within 25 geographic regions, with announced plans for 24
enable, not restrict, and by enlisting the guardians of these silos more AZs and eight more AWS Regions in Australia, India,
to become champions in educating others on how to use their Indonesia, Israel, New Zealand, Spain, Switzerland, and the
data rather than building walls around it. Data is empowering, United Arab Emirates. Millions of customers — including the
but it can also evoke strong emotions. To eliminate resistance, it’s fastest-growing startups, largest enterprises, and leading

important to create a strong change management strategy that government agencies — trust AWS to power their
infrastructure, become more agile, and lower costs.
focuses on communication and makes the use of data not only
easier but enjoyable.
To learn more, visit aws.amazon.com.

In short: A data-driven culture thrives when the senior leadership

is engaged (and middle management is empowered), front-line
employees can easily access data for everyday decision-making,
everyone is educated in data proficiency, and silos are eliminated.
Such a culture also creates a more positive work environment,
as it brings with it objectivity, transparency, and innovation. By
using a data-driven culture at scale, a company can successfully
use data as a differentiator in the marketplace — and as a unifier
within its own organization.

— Ishit Vachhrajani
Enterprise Strategist, Amazon Web Services

MIT SMR CONNECTIONS 6