We design and sell a wide variety of gear and clothing for outdoor enthusiasts. To ensure we always have what customers need for their adventures, BI software helps managers determine which products are in demand and make sure they're in stock.
In 2005, soon after I came on board, we created a merchandising dashboard available to most employees throughout the organization. We built the dashboard in about three months using Information Builders' WebFOCUS 7.1.3 platform, and rolled it out across the organization in four weeks.
The architecture is straightforward: We use Information Builders iWay integration technology to access point-of-sale information on an IBM AS/400 computer and load it into a Microsoft SQL Server data mart. From there, WebFOCUS presents the data through a series of dashboards that users can view via their Web browsers.
In this way, decision-makers can quickly access a unified, high-level view of key performance indicators (KPIs) such as sales, inventory, and margin levels, and drill down to granular detail that analyzes specific transactions. They deemed the BI environment an instant success.
Now we're enhancing this dashboard to create a multifunction employee workbench called E-Basecamp. We want it to contain all the information relevant to corporate goals, integrated with productivity tools and role-based content customized to each individual user. Finally, we want to bring it to life by interweaving Web 2.0 technologies for collaboration among internal and external stakeholders. We'll launch this new version early this year.
Our company uses about 20 operational metrics to govern the fundamental health of the business. For example, managers in the merchandising area have to stay on top of inventory positions and stock turns. E-commerce managers monitor hour-by-hour Web traffic and conversion rates. Each area of the business relies on the dashboard to learn when certain KPIs are out of tolerance range. We make this easy with a color-coded system of red, yellow, and green alerts to indicate metrics that are over, under, or at plan.
In response to customer requests, we're also experimenting with embedding RSS feeds into the dashboard to drive more focused inquiries. In addition to providing a structure for decision-making, the idea is to encourage online conversations and information-sharing. If certain items sell better than others, it should be easy to analyze the transaction characteristics and selling behaviors that produce these results, then cascade that knowledge throughout the organization.
|
The challenge now is figuring out how to encourage these collaborative interactions throughout the company and make them almost automatic. The goal is to let any dashboard user pose a hypothesis and invite commentary, almost like a notepad alongside the dashboard. That's the type of capability BI is supposed to provide.
Our dashboard examines Web visitors and sales by hour. We've successfully tied that data to an RSS feed so managers don't have to visit a Web page to view the latest numbersthe information now pops up on their desktops automatically.
We also plan to create a wiki that will let the user community test and refine hypotheses. In particular, we want our associates to share tips and best practices and initiate dialogues. If certain items sell better than others, the associates should be able to analyze the transaction characteristics and selling behaviors that produce the results, then use the collaborative tools to extend that insight throughout the organization.
One way we plan to do this is by creating blogs around a piece of data or a key metric. Blogs are a great way to post information to a Web site on a regular basis and invite comments. Many tools make it easy to archive, search, and categorize blogs for easy reference.
For example, if sales per payroll hour hover at $125 most months and suddenly drop to $75, a store manager might want to post an explanation or inquiry concerning the anomaly. A blog attached to a metric might reveal that payroll hours were higher that week to handle additional back-office work. Keeping comments in a blog lets readers observe patterns they might have overlooked using data analysis alone.