Blog

BLOG

Combat data skepticism with these 3 steps

PUBLISHED ON 9 JULY, 2015

Reliability of data is crucial in developing organizational cultures of data-based decision making. When poor quality data has been identified, teams may come to distrust their information and turn away from insight and towards intuition.

Poor data quality can be the result of numerous factors including:

• Faulty initial implementation of data collection solutions which leads to inaccuracies, inconsistencies or incomplete data. Or even worse: unreliable collection tools.
• Compilation errors either due to:
  • human error, where data is manually compiled,
  • or poor mapping where data sources are incompatible.
• Lack of accessibility, or dissatisfactory delivery, in terms of;
  • inappropriate presentation,
  • the serving of irrelevant information,
  • and/or failure to deliver data in a timely manner with a suitable periodicity.

In cases where these factors disrupt data quality, entire reporting structures can be compromised and data-based decision making most likely won’t be adopted, and where it is it may not lead to optimal outcomes.

In order to combat this skepticism, we suggest following these 3 steps:

• Run a quality auditing process
• Define a strong reporting model
• Implement an automated information delivery

Run a Quality Auditing Process

You may receive initial resistance to implementing an auditing system as individuals may fail to see the value in spending time on such a complex and time-consuming process. Investment in this process, however, can have significant long term benefits. Consuming inaccurate data can be even worse than not accessing any at all as decisions are then made under false premise. So if you are using analytics and reporting tools, it’s advisable to ensure your information is completely reliable, and it is in fact the type of information you are looking to collect – for example, that you are measuring the correct metrics, and that you have applied the correct dimensions, filters and segments, etc. As the implementation of measurement systems often relies on the work of multiple individuals, there is a high chance of human error, improper configuration or misunderstanding of the actual use of the solutions involved. Therefore an initial time investment in getting the systems set up is essential to confirming that you are consuming the correct information.

Define a Strong Reporting Model

The foundation for success in forming and maintaining a data-driven culture within your organization is ensuring that everyone has access to the most relevant information for them (and only the most relevant information for them). A good reporting system does just this: properly defines which information should be delivered to whom. While analytics tools make vast quantities of information available, delivering it to individuals without regard for their needs is simply bad practice. Firstly, we are all restricted in the amount of information we can digest, and secondly, we do not all need access to the same strategic indicators. So remove vanity metrics that only serve to distract from key indicators that allow business consumers to make decisions.

The primary characteristic of a well defined reporting model is that each individual receives the information they need to find actionable insights. In order to ensure your solution will do so, you must involve all key stakeholders such as consultants, analysts, and business-minded individuals. They must have a central role in the development of the model to ensure it is completely aligned with their needs.

Implement an Automated Information Delivery Solution

If we are able to combine quality information that has undergone a strong auditing process and a well defined reporting system, we are well on the way to having the information we need to build a data-based decision making culture within our organization. We are still, however, missing a crucial piece of the puzzle – the delivery.

If your organization is collecting and compiling valuable information for various departments or individuals, but this is inaccessible, you might as well not have gone to the effort to begin with. Only where information is served to those who need to consume it in a manner that best suits their style of work, will they be able to act on this data. For example, if I constantly travel, I may need to access my reports as PDFs or on my mobile device, but if they are only available on a web app, it is unlikely that I will adopt an insight-based decision culture.

A highly reliable automated reporting system, in addition to an initial quality data audit and the development of a strong framework, will ensure continued access to quality information. Unlike with manually produced reports, in automated systems, after the initial quality tests, you are able to rest assured that you are receiving the data you need.

Therefore, while initial investment in auditing the quality of your data, the strong definition of your reporting model and the implementation of a powerful automated information delivery system may appear costly, the combination of these three aspects will greatly assist you in not only making strong data-backed decisions, but also in building a culture around this. Although you may be in a hurry to get your reports out, it’s better to calm down and carefully plan the implementation of your reporting system to guarantee long term cultural adoption and effective insight-based decision making. In the long term, the financial benefits of ensuring quality data will be immense as you are able to take actions that optimize business outcomes.