Oops! The input is malformed!
Originally published 28 August 2007
A fundamental challenge in the development of a company’s business intelligence (BI) strategy or the implementation of a business intelligence solution is determining what information will provide the most value to groups of users and how best to provide that information to them. These issues become even more challenging when companies need to get critical information to users who don’t know how to use sophisticated business intelligence tools, or who don’t have access to those tools. Many potential BI users in modern companies work in an environment in which they are often disconnected or “un-tethered” from a corporate network.
For most sophisticated users who work in an office every day and are connected to their corporate networks, performing business analytics activities using sophisticated BI tools is nothing new. Their business intelligence needs are addressed by providing them with the tools for running and manipulating tailored pre-established queries/reports or running ad hoc queries against data marts and data warehouses that are regularly updated with data from transaction processing systems and other sources – whenever they want it or need it on demand. They can execute those queries, slice and dice the information, filter and reformat the information, or even download it to an Excel spreadsheet if they feel the need. We’ll refer to this corporate environment as a “pull” approach to business intelligence. This is the primary scenario for which all business intelligence tools, packages and suites in the marketplace have been built from the beginning, providing a plethora of tools to choose from to support the needs of these users.
However, when companies need to communicate and distribute accurate and consistent information to a population of users that spend a majority of their time “un-tethered” from a corporate network, another approach is needed. Those users don’t need, don’t have, or can’t use those sophisticated tools described earlier for one reason or another. The most common problem is that they might not be able to access centralized information when they need to because they are disconnected from the sources of data behind the tools. Conversely, a company wanting an information-rich environment may be supplying reports with too much data, resulting in information that isn’t relevant to the user. The challenge in this situation is not the complexity of the tools or data, but to determine what content to deliver to those users that is straightforward and yet valuable. Content must not only be targeted and relevant, but also determining the method to best get it to them, given the environment in which they work, is imperative. Last but not least, it is central to know how much they need to interact with and manipulate the content. This much is certain – companies looking to improve distribution of analytics content to “un-tethered” users must find solutions that are simple and intuitive or they will fail.
For these situations, an approach of automated, proactive distribution of business intelligence content that can be downloaded when limited connectivity is available is ideal. Let’s refer to this concept as a “push” approach to business intelligence. A BI-push solution takes business intelligence content (dashboards, query results, data, etc.) from a business intelligence solution and proactively pushes that content to users without them having to request it so that they can access it when they have the time to do so. The typical delivery mechanisms for these solutions are: e-mail, a web/portal solution, or a file download connection when there is access to an Internet connection or a connection to the corporate network for a brief period of time. Once the content is received by the user, it can then be accessed whenever desired, even while disconnected.
Let’s assume that the necessary content for users has been determined and walk through a simple scenario, where a user needs to receive content each morning without having to interact with or manipulate it. The only requirement is a dashboard summary or the results of a query. In this scenario, the first step is to generate the relevant content for the user(s). This requires the corporate business intelligence solution to automatically generate the content each night (most likely by running a query or generating a dashboard) in the form of a static report (for example, in PDF format) or static web page. Once the content is generated and saved, it must next be distributed to the users by a BI-push solution. This can be accomplished in several straightforward ways. The content can be e-mailed to the user as an attachment to the e-mail or in the form of a URL link to a web page that houses the content. The user then simply accesses their e-mail or the web page when they can connect to the company network, and the BI content they need is available.
A more complicated scenario would involve a user receiving content each morning; but, in this case, the user does need to interact with and manipulate the content. For example, one of our clients had a large number of field sales managers who needed better and more frequent access to sales performance metrics and analytics. These sales managers were rarely in a traditional office and, therefore, were not typically connected to the corporate network during the day. When the client reviewed available BI-push offerings from BI tool vendors, they found that the solutions were overkill for the relatively basic information needs of this group of sales managers.
For this client, we developed a hybrid solution. We used a traditional BI reporting tool to generate a series of reports each night. These reports were then converted to static Adobe Acrobat PDF result files, and each file was tagged with metadata describing the parameters of the report it contained. Then, these reports were pushed out to the field sales managers the next time they connected their laptops to the network, according to a set of business rules that specified which reports were to be pushed to each sales manager. The managers simply had to connect their laptops to the Internet in the evening, and the reports would be automatically refreshed for them.
Once these reports were pushed to the sales managers’ laptops, they were given the ability to navigate through the various reports, with the capability to use a basic “drill down” function on them. In reality, they weren’t interactively drilling down – clicking on a link in one of these static reports simply opened up another static report along a predefined drill path. While this solution may be rudimentary compared to true OLAP BI capabilities, it was a good fit for this client’s needs: it provided easy-to-use basic data analysis to a set of disconnected users, at a reasonable cost and with minimal overhead. The take-away from this example is that delivering the right information in an easy-to-access fashion is of paramount importance, even when it’s done with a relatively “low-tech” approach.
If disconnected users require full BI tool capabilities to interact with and manipulate content, then when they have connectivity, the company either needs to push them BI data so they can use it locally on their laptops with local BI tools (with the inherent data quality and security risks that come with having local data) or push them a URL that takes them to a web/portal front-end to a fully interactive query/dashboard or to the central BI solution and its full complement of BI tools. Only the first option is really a “push” scenario, and it has a number of undesirable quality, security and performance risks associated with it. The second option is really just a variation of BI “pull,” simply with a web/portal front-end. If your disconnected users need these capabilities, then you should look at available BI “pull” tools and assess their web/portal access options.
As is the case with all business intelligence solutions, especially in the ever-changing business environment in which we all exist, it is extremely important to regularly re-assess the processes, methods and technologies that are used as well as to solicit feedback from users of these solutions. Also, keep in mind that once these solutions are successfully demonstrated and positive word gets around, the floodgates will open with new requests for additional content for more users. Thus, it’s important to be prepared: even though a low-cost, low-tech solution may be appropriate and attractive initially, it must be able to scale once demand picks up. It’s best to first develop a strategy organized around logical groupings of data, analytics and user requirements before considering what solution will best create more timely, accessible and interactive information to better manage a business at all levels.