Trust is a vital component of any data strategy. When users trust the data, it can be used accurately and efficiently to provide invaluable BI. However, when users have doubts about the data, it can be detrimental to the insight extraction process because of the lack of consistency and clarity across the organisation.

Nimbus Ninety members gathered online to discuss how to build trust, user engagement and adoption and encourage cultural change using an analytics platform to foster greater collaboration.

WHY IS TRUST SO IMPORTANT?

All organisations are driven by data. Even if data is not a company's main business output, there are numerous important ways data is vital to an organisation. The end goal of data analytics is the insights it creates and every business wants to harness those insights to have that competitive advantage. Simplicity is often considered as a downfall when it comes to data - if it’s too simple, it cannot provide the nuance necessary for valuable insights. This leads to issues around trust within the organisation. But as Ian Macdonald, the Principal Technologist at Pyramid Analytics, stated to the audience of members, “making the sophisticated simple should be a priority for businesses”. 

But what happens to those all-important insights when the data is unclear and comes from multiple analytics platforms? What happens when teams spend more time arguing about the quality and results of data rather than exploring what to do with it?

This is where the trust in data analytics tools is incredibly important. As our members heard from Will Blake, the Director of Technology and Analytics at CRU, these are issues his team faced. Since CRU is a data-focussed company, they are not using data solely internally but also sharing data with clients. As a result, having trustworthy, solid data is a priority for them. Will spoke to our members about how CRU built trust in the and reduced system complexity by using Pyramid to bring all the high-level data together and display it in one place, and make it accessible across the business. This means CRU now spends less time discussing the quality, source, and reliability of the data and more time extracting valuable insights from it.

Insights do not just materialise out of nowhere: insights happen when there are fundamental rules in place. Ian explained the 4 key elements that make up a data analytics platform producing trustworthy, effective data:

• Unified 
• Flexible 
• Accessible 
• Collaborative

Unity is about the importance of having the entire analytics workflow resting on a universal platform that can be accessed on any device or browser by any authorised user, preventing  data siloes that occur when organisations are using multiple platforms. Flexibility is important for keeping an organisations’ existing data infrastructure in place and implementing an analytics strategy that conforms to it, rather than the other way around. Accessibility works to make the adaption to a singular system more seamless and user-friendly in adapting to the different ways organisations work cross-departmentally. Finally, it is important to foster a collaborative environment with datasets and resources that are shared, managed, and secured across a common governance model. 

LESSONS LEARNT: REBUILDING TRUST

The discussion groups focused on several areas, including how to foster trust when the data is dirty,, how to drive user adoption of tools, and why collaboration is so important. 

For many members, the trust aspect is crucial. They expressed the issues they face across teams who do not use the same platform, and so receive different insights from the same data. The discussions revolved around how to prevent these varying levels of trust in data extracted from analytics platforms. There was also discussion on the importance of QA processes to increase this confidence. Across the board, the consensus among members was that it is not just the technology that has to change: the culture surrounding the ingrained uses of specific platforms must go along with it. 

The adoption rate within an organisation is arguably the toughest challenge - but not insurmountable. As Will explained, his organisations’ methods of implementation were to have a grace period where all platforms are used and then completely remove all other options. This prevented any sort of overlap or differing results. There was also discussion around allowing users to build their own dashboards and actually having a say in the ways they interact with the platform through upvoting particular layouts they like. Empowering business users to be autonomous in their use of the platform drives adoption rates up, and thus overall capability in the insight space.

Finally, collaboration was discussed as an important element in two ways. Firstly, in terms of advanced capability beyond data science teams. Members expressed that they wanted these analytics tools to extend the governance of data outside of data science teams and spread that across the organisation. Secondly, there was discussion around how users interact with the data in terms of how important it is to have a common language about the data in order for insights to be extracted effectively.

Data is the lifeblood of any organisation, but without trust and the drive to generate the right insights, its value diminishes.

This event was in partnership with Pyramid Analytics, an enterprise data analytics platform  provider.