4 characteristics of data-driven organizations—and how to get started

Share this post:


Robin Andersson, Information & Governance Lead, Nordic market

Doubtlessly, there are major benefits of transforming into a data-driven organization. Forbes studies show that such organizations experience a 27% year-over-year increase in revenue compared to 7% for other organizations, and 12% reduced operating expenses from the prior year compared to 1% for other organizations. Behind their success there are four common characteristics: data is centralized and organized, data governance policies are in place, and both data and analytics access is integrated into tools. Question is where to start?

According to the IBM report Inside the mind of Generation D seven out of ten respondents say they aren’t lacking data. Instead, the key to driving business value is how enterprises use data and sophisticated analytics. Many still has a long way to walk; a recent IBM Institute of Business Values report indicates that approximately half of all C-level executives make business decisions on inaccurate or inadequate data.

What does it mean to be data-driven?

One of the main challenges facing organizations today is about siloed data, which limits access to data and sources. In addition, many of the tools in place are anything but easy to use, making analytics harder than necessary to perform.

If we instead turn to successful data-driven organizations, what are their common characteristics?

  1. Data is centralized and organized. To ensure data is recent and relevant, data-driven organizations gather data from across the organization. Don’t limit data gathering to internal sources; there are multiple external data sources that can and should be included in analytics (depending on the industry). Be aware of what data and how much you gather—information overload is a real threat. Today, many companies are establishing data reservoirs with a combination of internal and external data. However, it’s of outmost importance that data flowing into the data reservoir is strictly controlled in order to reap the desired benefits of the investment. A key guideline? It’s not about size; data variety is superior volume.
  2. Data governance policies are in place. To safeguard the quality of key master data objects like product, customer and supplier, there should be clear data governance processes in place. According to Gartner, master data quality deteriorates at a rate of 2% per month if not governed. Master data quality is key for high quality analytics.
  3. Data is accessible. In data-driven organizations everyone has access to some data. Almost no one has access to all of it. Data security and privacy requirements limit what data can be accessed. A guiding principle? Everyone should have access to data required to perform the job, and data should be easily accessible via different tools—smart phones, desktops and laptops.
  4. Analytics is integrated into tools. Analytics tools at data-driven organizations tend to be among the most innovative, with functionality to easily define analytics models. They are also embedded into existing tools, making them more likely to be used. Competitive advantages arise from analytics models that allow users to predict and act upon business insights and thereby optimize outcomes. When analyzing data, organizations should always start with a business question.


Where do you start to become data-driven?

Rule no. 1: Don’t undertake massive change! Firstly, start to collect and look at your data; it’s the fundamental building block and you cannot answer any business question without it. Secondly, transform the organizational culture and empower your people to access data and encourage them to use it—even with self-service. And thirdly, don’t overengineer data models.

Who should own your data strategy?

In order to accelerate transformation, you can appoint a Chief Data Officer (CDO). A CDO can take on different responsibilities, but in general the CDO drives the data agenda and owns the group-wide strategy. Roles and responsibilities should ideally be clarified before the role is implemented to avoid treading on anyone’s toes.

In US, we now see an increase in the number of appointed CDOs among IBM clients. How quickly will other countries experience a similar pattern? After all, data is today’s most important enterprise asset, yet not fully utilized. My intention is to continue this interesting discussion in social media—please feel free to contribute!


Följ mig gärna på:  LinkedIn logo LinkedIn

More stories

Två hundraåringar spekulerar – kommer AI att ta våra jobb?

När IBM nyligen blev inbjudna att hålla en traditionsenlig gästföreläsning på Stockholms Handelshögskola och kursen ”Humans vs. Algorithms: Judgment, Prediction and Nudges” var det AI & etik som stod på schemat, med extra fokus på interaktiva assistenter. Som du kanske anar saknades det hundraåringar både framför och bakom katedern denna dag. Dock fanns det gott […]

Läs mer

Hundra år av Vasaloppsdata, del två

Stavningsrättning av OCR-texten På min YouTube-kanal finns en kort video där jag tränar OCH tänker på det här projektet, samtidigt: Efter att jag tagit mig till etappsegern “Hundra år av Vasaloppsdata: Vad skrevs om loppet 1922?” i jakten på insikter från tidningsartiklar från när Vasaloppet var ungt, ville jag se om kvaliteten på texten som […]

Läs mer

Hundra år av Vasaloppsdata: Vad skrevs om loppet 1922?

På ett möte tillsammans med Vasaloppets ledningsgrupp tog en idé form i huvudet på mig: Skulle man kunna använda AI för att vaska fram intressant kuriosa om hundraåringen Vasaloppet? Ur gamla källor, som varit glömda och gömda under många år? Kan det finnas fakta från dessa år som inte ens de mest pålästa känner till […]

Läs mer