Menu
Beyond the data warehouse

Beyond the data warehouse

John Brand of Forrester highlights need to rethink the strategy for business intelligence.

Organisations are starting to rethink the business intelligence (BI) landscape as they feel they are not getting the value out of the existing approach, says John Brand, vice president, principal analyst, CIO Group at Forrester. However, Brand says enterprises have to look at business intelligence as a process. He points out that most organisations are very process-orientated in areas like sales and services, but not in BI.

With the latter, “They go, we have this data warehouse that runs reports and hopefully this will be useful,” says Brand at the recent Microsoft Big Picture Roadshow.

Brand says the process of deploying business intelligence entails looking at related areas; does the organisation use BI to set objectives? Does it measure the right performance? Does the organisation have the right tools in place to quickly identify issues that are impeding performance? Is the organisation doing deep level investigative processes, the true data mining?

The last, he says, is about looking for opportunities and extracting this information, and using this for a change that can benefit the organisation.

Brand says it is very rare to see organisations thinking about business intelligence in this manner. “They are often just using whatever they can find in their systems and the data to give them direction.”

The state of BI

A range of trends, meanwhile, is impacting BI, which continues to be in the top five “very hot topics” among CXOs.

Information complexity is increasing, and it is not just volume overload, he says. Storage is also getting cheaper. “The real problem is complexity of the information that is growing.”

Brand says business demands for BI are increasing. Users feel they should be able to get any piece of information anytime, from any location and any device. They don’t care where the information is located, but how easy it is to access the data.

Social computing and Web 2.0 are leading to fundamental changes in the enterprise. Many organisations feel social media is “something for marketing departments to worry about and keeping the consumers happy”.

The irony is it is creating fundamental change as enterprises move from a tightly controlled mainframe environment to a decentralised and complex environment.

“Everything is visible, who made change in the document,” says Brand. “It is a very different paradigm, it will impact IT generally,” he says. Invitation based access control will change how we develop and deploy IT, says Brand, as he cites his own experience inviting people into a workspace to share information with him securely.

A the same time there is a growing interest in adopting BI outside the traditional financial and non-financial data, for example doing research on audio data for call centres.

Meanwhile, a bigger challenge is in the “big data which is affecting the organisation but which is not under its direct control”.

Businesses have internal data from enterprise applications but are also using a lot of data from outside the organisation. Organisations expect to use this data the way they use internal data and blend the two.

Brand says an example is a company that is outsourcing its online store environment, with the outsourcer taking care of the business data component.

“But you are increasingly not only using that data yourself but you start to blend those environments with your internal environment. This idea of blended BI where it is no longer just internally-focused is starting have a real impact. It is causing a lot of complexity because now you have different BI platforms owned and run by different people with different levels of skills and expertise.”

He also sees a “more blended environment” for BI where users are bringing more context to data driven insights.

While it is early days, he says an interesting trend is using cloud computing processing power to provide analytical workload capability.

Challenges ahead

Data quality, however, has always been the number one challenge for enterprises. He advises avoiding the annual data quality event or ‘Olympics’ if this is held every four years, which is the most common approach used by enterprises.

If data reliability is low and criticality is high, you need to fix it. “But fixing data is a temporary step. It is about fixing business processes.”

“CIOs should really look at their BI and general information management portfolio and start to think how that needs to evolve,” he says. “Some of the questions they need to answer are what are the critical applications, who will be running these and where they will be sitting.”

A layered approach

Brand uses the analogy of a layered onion. “At the core is the data that needs to be really tightly managed, tightly controlled,” he says. “You can trust and rely on them as a true corporate asset.

“It is not [an] all or nothing” approach, he says. “You get a core set of data that is really tightly controlled and you get other data which is less formally controlled but still has some management.”

He cites an organisation that uses colour coding or ‘value stamping’ to indicate the quality of the data in the reports. Brand says this acts like a “fuel gauge” so that the person using the report has an idea of the quality and relevance of the data in it. “It does not mean the report is worthless. It just becomes ‘worth less’. There is that distinction,” he says.

“Some companies know how many reports they have got, but don’t know whether a report that is used only once a year might be the most valuable and the report that is used everyday is the least valuable. It is a constant re-evaluation process of what is being produced, why it has been produced, is it still relevant?”

“Increasingly fitness for purpose of information is something the CIO needs to build as a core capability to help the business understand how it is using data, how it should control data more effectively, how it can improve data quality.”

In some larger organisations, there is a person who is assigned to look at purely data quality issues outside of the data warehouse, to see how they can improve the business use of data.

Smaller organisations may not have the resources for these roles, but have to be both more disciplined but not apply too many controls. “It is a hard balance to strike,” he says.

To comment on this article, please email the editor.

Sign up to receive CIO newsletters.

Follow CIO on

Twitter @cio_nz

Facebook

LinkedIn

Click here to subscribe to CIO.

Send news tips and comments to divina@cio.co.nz

Join the CIO New Zealand group on LinkedIn. The group is open to CIOs, IT Directors, COOs, CTOs and senior IT managers.

Join the newsletter!

Or

Sign up to gain exclusive access to email subscriptions, event invitations, competitions, giveaways, and much more.

Membership is free, and your security and privacy remain protected. View our privacy policy before signing up.

Error: Please check your email address.

Tags Microsoftbusiness intelligenceforresterJohn Brand

Show Comments