Skip to content

Advanced Analytics for data-driven organization

    The 2015 Pacific Northwest BI Summit is taking place in Grants Pass, Ore., this weekend. The annual event brings together a small group of consultants and vendors to discuss key trends and issues related to business intelligence, analytics and data management. One of the participants is Claudia Imhoff, president of consultancy Intelligent Solutions Inc. and founder of the Boulder BI Brain Trust. At this year’s conference, Imhoff will lead a discussion on increasing the adoption of BI and analytics applications in companies. That topic is similar to one she spoke about in a video interview with SearchBusinessAnalytics at the 2014 summit: creating a more data-driven organization through the use of higher-level predictive and prescriptive analytics techniques.

    In the interview, Imhoff said that basic descriptive analytics — for example, straightforward reporting on revenue, profits and other key performance indicators — is the most prevalent form of BI. But it’s also “the least valuable of all the analytics that companies can perform,” she noted. The next step up is diagnostic analytics, which addresses why something has happened but is “still reactive,” according to Imhoff.

    On the other hand, she said, companies can use predictive analytics tools to look toward the future — by, say, identifying prospective customers who are likely to be receptive to particular marketing campaigns. And prescriptive analytics software can be applied to answer what-if questions in order to help optimize business strategies and assess whether predicted business outcomes are worth pursuing.

    There are a number of issues that hold companies back from adopting more advanced analytics techniques, Imhoff said. One is a lack of internal education about the potential business benefits of effective analytics processes: “We need to start building this culture in our organizations that understands the need for analytics.” Another issue she cited is a lack of analytical prowess resulting from the ongoing shortage of data scientists and other skilled analytics professionals. And an age-old but still common problem, she said, is “putting the cart before the horse”on technology purchases and ending up with analytics systems and tools that aren’t a good fit for an organization’s business needs.

    Imhoff said BI, analytics and IT managers also need to understand that data warehouses aren’t the only valid repositories of analytics data anymore, especially for storing the massive amounts of data being captured from sensors, social networks and other new data sources. To support big data analytics applications, she espoused an extended data warehouse architecture that combines a traditional enterprise data warehouse with technologies such as Hadoop clusters and NoSQL database systems. She sees well-designed data visualizations as another must for fostering a data-driven organization, especially in big data environments: “We’re talking about massive numbers of data points, and you can’t just ‘blat’ that out on a screen.”