Blog Post

Get Business Intelligence Ready for the Real-Time Web

If you’ve been wondering why Enterprise 2.0 still hasn’t taken off like it should, maybe it’s because we’ve forgotten about business intelligence, which essentially hasn’t changed much since the ’90s. While the rest of the world has moved on, BI is still stuck in 1.0. The latest research shows that ‘mo apps, ‘mo problems for enterprises, unless they can fix their underlying business intelligence processes first (which also explains why BI adoption rates may actually be declining since last year).

The massive explosion of data the real-time Web has given us looked like a good thing for enterprises at first. After all, the bigger companies are, the better they should be at drinking from this new firehose of raw data. Unfortunately, however, the real-time Web turned into an endless tsunami of Tweets, RSS feeds, live CRM and ERP data, videos and social media.

Ironically, the best tools for managing the real-time Web are available for free to millions, instead of costing millions. Small businesses and consumers are thriving with TweetDeck, Google Docs, Basecamp, Zoho and real-time feed readers. However, these same tools have only made matters worse for large enterprises, because they tamper with the underlying machinery, ruining the business process.

According to a new study by Ventana Research, most “desktop productivity tools obstruct efficiency and effectiveness when used in enterprise processes.” Basically, these new apps interfere with deeply entrenched, process-based systems. For example, using collaborative spreadsheets for reports generates a 47 percent error rate (think about all the times you emailed a file around or someone overwrote your changes on Google Docs) versus standard ERP reporting. But who can blame workers for using them? Even if the enterprise doesn’t provide adequate tools to manage the real-time Web, the job still needs to be done. So workers turn to consumer apps, leading to all sorts of security concerns, data loss, reporting errors, time wasted on Facebook and other inefficiencies.

This in turn has had a detrimental effect on the use of enterprise systems. According to a 2010 IDC survey of midsize and large businesses, 25-75 percent of enterprise licenses are paid for but unused. When a marketer needs to see what customers think and how they feel about a new product, there isn’t time to wait for the quarterly sales data to come in and be analyzed by their ERP system — they need to get on Twitter now and directly monitor what’s going on.

So, if new apps increase inefficiencies and app fatigue, does adding new Enterprise 2.0 applications make sense? Not yet. But perhaps it could if the underlying business process changed.

The Future of Business Intelligence: Harness Human Power

To date, the discussion around Enterprise 2.0 has focused mostly on CRM, ERP and communications, but the larger topic we should consider is really Business Intelligence, since that’s what draws data from all other enterprise systems and translates them into actual reports and processes.

Business Intelligence had its heyday in the 1990s (SAP BusinessObjects, anyone?), based on taking corporate data silos (databases, intranets), applying business logic layers, and ultimately delivering analytics and tangible business processes. This system works great when you have defined datasets.

However, the real-time Web is not defined nor controlled by the enterprise. It’s chaotic. It’s massive. It’s the entire public Internet — not your intranet — and it’s intangible. How can any BI system analyze the entire Web in real time or create a quantifiable business object out of a product unboxing video?

So we turn to consumer tools. We give up on enterprise systems and reporting. We go back to monitoring directly with our own eyes, like a pilot who realizes his instruments are broken or Morpheus scanning The Matrix. Salesforce Chatter is a good indicator of this new trend towards monitoring, but unfortunately it doesn’t help workers monitor everything and it doesn’t help enterprises create actual business processes. We need more. Much more.

The future of BI 2.0 will not rely on server power, but on the human power of the entire organization to capture everything, convert social media into tangible business objects, and automate real-time business processes. BI 2.0 needs to help enterprises do three things:

1.     Monitor everything. We need to give workers the ability to better monitor a million things at once: news, competitors, visual brand identities, consumer feedback, e-reputation, etc. (10 competitors x 10 major news searches x 5 social media sites = 500 sites each worker needs to check manually everyday just to stay industry aware!)

2.     Self-organize. Flickr and YouTube’s content was considered too big to organize, until tagging came along. Similarly, we need to help enterprises self-organize the billions of clicks, searches, Tweets and reports they produce every year within their company by harnessing the tagging talents of experts within their organization. Private tag clouds, anyone?

3.     Trigger processes. As we uncover trending topics within the organization, we need to enable it to automatically trigger actual business processes in real-time, like purchase orders or ad budgets.

Essentially, many new Enterprise 2.0 apps that try to accomplish this, from corporate wikis to social networks, fail because they are external systems (‘mo apps, ‘mo problems). They do not help users monitor everything (rather they are an additional thing to monitor). They are not self-organizing and they fail to trigger real business processes.

The next generation of Business Intelligence systems will need to be able to integrate both consumer and enterprise systems together to fix these inefficiencies and leverage the human intelligence of the entire organization to capture the real-time Web.

Freddy Mini is the CEO of Netvibes, a real-time dashboard platform for both consumers and corporations. Prior to Netvibes, he co-founded music search service musicMe and was the CEO of Ziff-Davis (France) and the Senior VP and Managing Director of CNET Networks Europe.

Graphics courtesy of Netvibes.

10 Responses to “Get Business Intelligence Ready for the Real-Time Web”

  1. The typical enterprise today views business performance through a 24-hour window. Yet for many companies, minutes or seconds can take on critical importance as customer expectations grow. If service levels are not met, then customer confidence and revenue is lost. By having essential information on operations performance in real-time, decisions can be made instantly to respond before customers notice a problem. Make no mistake, companies of all types have critical business processes they need to manage in real-time.

    Business Intelligence solutions are unable to address the moment-to-moment need for real-time decision-making. BI adds real value, but only provides a historical perspective – too late. In contrast, Operational Intelligence relies on real-time feeds (i.e., Event data).

  2. I would have to disagree. Each Enterprise has real time BI. Of course an Enterprise has very different needs from a small website. Microsoft, SAP or IBM have incredible BI systems.
    This funny stuff with eyes monitoring the web is a marketing from the dashboard guys. You must be kidding me.

  3. The old world of enterprise BI allowed you to analyse past performance and try to predict future performance but didn’t give you an easy way to monitor your business’s vital signs in an easy to consume way.

    We’re tackling this exact problem over at  Being able to monitor all of your company’s key metrics in real-time brings a raft of benefits, not least is being able to spot potential problems before they spiral out of control.

  4. The future of Business Intelligence will never harness the needs of human power, but it will demand more of human power to manage and to control the system. Like you said, Morpheus scanning the Matrix. No matter how advance the Matrix system, it cannot scan itself, because only the Morpheus can scan the system. Machines can never overturn the human intelligence. Hehe…