For at least a decade, companies have worked with the same approach to business intelligence, keeping it locked up in a small, trusted group of managers. All that is changing as a new, collaborative culture - and the tools to support it - starts to emerge.
If your business relies on knowledge workers extracting the maximum value from your assets, you probably want them to know as much as possible about what they have to work with. So it would seem logical to empower them with business intelligence tools that provision them with information and support sharing of business-critical data.
For a long time, however, that is not what has happened. According to BIScorecard, which until 2014 ran a decade-long benchmark of the adoption of BI tools, only 22 per cent of employees on average have these systems in place - a figure that had stayed virtually the same for ten years. To make the picture worse, just 25 per cent describe their BI deployment as very successful and only 28 per cent believe it has made an impact on their business.
The problem has been twofold: firstly, the amount of resource required to extract the right numbers from operating systems has been a significant constraint; secondly, there has been a “hands-off” culture in many businesses where those who believe they have ownership of a critical report are unwilling to yield their control over it.
Solutions to both of these problems are starting to make themselves felt, however. Technology is developing rapidly so that both the source systems have become easier to read (not least as they increasingly move into big data-style infrastructures) and the intelligence simpler to extract and distribute. Culture in business is also moving away from command-and-control towards collaboration, especially as Generation Xers become managers and bring their connected world views into play.
A great example of the new social BI is the “Superstar DJs” programme operated by RBS which was the winner of the Data-Driven Digital Team prize at this year’s DataIQ Talent Awards. Giles Richardson, head of analytics at RBS Personal and Business Banking, says: “In most enterprises, they generate useful insights quicker than they can action them. The conversion rate for something to change is usually about 10 per cent.”
One of the reasons for this is the bottleneck whereby BI teams identify issues in the data which then have to be acted on by another centralised function downstream. Within RBS PBB, there are fifty “journey managers” each responsible for a specific piece of digital real estate. If each had to wait for BI and web design teams to make changes, the customer experience would be sub-optimal.
To avoid this, the direct analytics team has provided those web content managers with access to their own digital analytics, resulting in a transformation in the culture of its content marketing. This new, structured test-and-learn approach has delivered over £1 million in incremental value in less than a year.
Before this project, an internal culture existed that Richardson describes as like “fireworks” - light something (such as a new web experience), run away (without measuring it), then receive the applause (despite the absence of proof of success). To address this, the new programme was launched in mid-2014.
The five-strong team decided to put the power of web analytics and optimisation tools directly into the hands of managers, rather than keeping them within a select group of analysts. Rules were also introduced to ensure changes were only made where it was clear that customers had been struggling, measurement was taken of the new solution and if it showed no improvement, it was changed back, and only positive customer impacts got celebrated.
As a result, instead of three optimisation tests in a year, RBS has been able to run 70. The journey managers are now the Superstar DJs, competing against each other via charts and playlists, the Christmas number one and gold discs. “We have put the democratisation of data at the heart of our strategy. We have invested a lot of time building the skills of those digital teams, but then the only value from analytics is if you put them to use and achieve better outcomes,” says Richardson.
Shifting BI out of that bunker where only one in five of the workforce get access to information is the driver behind a new wave of tools which underpin the new democratised culture. While they are important enablers of social BI, without that critical shift in culture, even the most dynamic software application will go unused.
“Every company wants to think each department is collaborating, but that only works in a ‘no blame’ culture,” says Peter Baxter, managing director of Yellowfin. “Most companies out there have very high-level goals which they are failing to deliver against. A collaborative culture in the company is really important. That has to start at the top and resonate right down to the bottom.”
In some cases, Yellowfin users have been able to drive cultural change through the adoption of collaborative features within the application, since they make the new way of working practical and visible. Baxter uses the term “the Instagram of the BI world” to explain the difference between what his company offers and how it views other vendors.
While this “instant gratification” helps to win user adoption, he stresses that the solution also allows governance and security to be built into information even as it is being distributed more widely across the enterprise. In many existing BI environments, fear of information leaking or being misused is often a constraining factor. “Traditional BI can not offer collaboration. Everybody wants to move away from that model, but they find it difficult if their BI tool doesn’t allow it,” says Baxter.
Just how congested conventional BI practices can become was highlighted by British Gas in the presentations it made at the DataIQ Summit. Eddie Edwards, head of BI, has been at the centre of the transformation of its management information function. As he noted, “we had 20,000 KPIs, thousands of reports, shadow MI and 1,500 people requesting data.” That is a model of multiple, parallel BI activity, rather than collaboration, despite the numbers involved.
Migrating operational data into a Hadoop data lake and pre-building tables that link key pieces of information has allowed British Gas to distribute insights via Qlikview without the MI function having to respond manually to every request. Instead, it has created APIs so business functions can automate data requests and run their own reports. “Data is now very accessible to the business community and we have been able to shut down shadow MI by providing access. It also means we avoid the problem of different numbers being presented to decision makers,” he says.
Payback is resulting from improvements to core business processes where issues have been spotted in the data. “We have identified a £20 million revenue opportunity just from improving on failed customer meter engineering visits,” notes Edwards.
That is the kind of result which most organisations would welcome from their BI, yet most still struggle to achieve. It requires creating a circle of trust in which business-critical data is centralised to ensure its validity and integrity, but access and reporting is socialised through new tools. Technology is no longer a barrier to this vision - what makes a difference is the willingness of the business to lean back and let go.