Report delivery, alerting must be flexible, functional

Report delivery and alerting is the style of business intelligence (BI) designed to proactively distribute large numbers of reports and alerts to large populations of internal and external information consumers.

Most BI vendors offer products that can centrally distribute e-mails to large user populations, with report enclosures, and on a scheduled basis. However, there are few architectures that support an additional four key areas:

1. Report distribution through any touchpoint.
2. Self-subscription, as well as administrator-based distribution.
3. Delivery on demand, on schedule, or on event.
4. Automatic content personalisation.

With e-mail distribution, the system should display the report directly in HTML, or enclose it as an attachment in the form of an Excel, Rich Text Format (RTF) or Portable Document Format (PDF), with automatic ZIP option and password protection.

In networked printer distribution, reports can be sent directly to networked printers worldwide automatically. The report is formatted in PDF format and sent to all of the printers defined by each recipient user`s preferences.

Bottlenecks caused by a centralised system for report delivery and alerting can be avoided by allowing users to self-subscribe to reports.

Charl Barnard, GM, MicroStrategy

Networked file server distribution allows reports to be distributed directly to archival file servers anywhere on the network. With each new file distribution, the system can automatically create new archive folders, thereby allowing companies to keep a historical record of all standard reports.

With corporate portal distribution, reports can be distributed directly to Web portals. This allows companies to concentrate all information publishing and consumption on their corporate portals.

Bottlenecks caused by a centralised system for report delivery and alerting can be avoided by allowing users to self-subscribe to reports. This eliminates the dependence on a central administrator and allows users to continually change which reports they want to receive and when.

End-users can also subscribe other end-users to receive reports.

Self-subscription over the Web greatly reduces administration costs while enabling applications to be deployed to extremely large user communities.

Centrally maintained security profiles ensure users do not receive any information to which they are not entitled.

Users and administrators can establish a periodic distribution for any report (on schedule). In addition, users can establish subscriptions for themselves as well as for other people, as long as they have the security privilege allowing them to do so.

In addition, users can distribute a report to others immediately (on demand). In this scenario, a user can freely customise a report with pivoting, page by and drill anywhere, and then immediately send that report result to any other user of the system. The recipient will see exactly the report that the sender saw prior to sending.

Report distributions can be triggered based on an event that occurs in the database (on event). This might include new data being loaded into the database or a metric in the database exceeding some pre-established threshold value. The event-based distribution capability is critical for real-time alerting applications.

The system can monitor key performance indicators (KPIs) or analytical indicators, based on business rules, sending an alert to the decision-maker as soon as an indicator reaches its pre-defined threshold or condition level. Users can subscribe to alerts based on any event criterion of their choice.

The alerting functionality is crucial for uncovering problems with high impact risk but low probability of occurrence. It allows decision-makers to fix small issues before they become major problems.

Users can manage their distribution profiles and preferences directly from the Web interface. They can define distribution triggers and delivery touchpoints for any given report distribution service. In addition, they can specify their language of choice, time zone and currency setting.

Most people are only interested in a small slice of data contained in a standard report. This can be achieved by providing four levels of content personalisation:

1. Authentication personalisation, which associates relevant data with the user-stored profile. Users can also receive information from external sources, with user IDs and passwords integrated in external systems` security.

2. Preferences personalisation, which lets users define the specific information they want to receive. Users can choose to specify general preferences for all reports they receive or change preferences for specific subscriptions or reports.

3. Locale personalisation. This allows users to specify how numbers and dates are formatted, which language should be used, and which international character set should be applied to their reports.

4. Security personalisation, which allows any report to be automatically tailored to include data appropriate to each user`s role and group affiliation. The report content can be personalised for regional managers to include data for each manager`s region. Subordinates would see even less data, and superiors might see summarised data appropriate to their interests.

IT administrators can choose to secure data for all reports by individual recipient or to limit an individual`s data access to specific reports and objects. Profiles can be dynamically applied to every report distribution.

IT workloads and development time are greatly reduced because report designers can build one report that can be shared by all users, with the assurance that data will automatically be protected based on each user`s different security settings.

Business intelligence should be pervasive throughout an enterprise. To achieve this, companies need a system that proactively pushes information into the hands of users wherever they might be, and through whatever interface is most appropriate to each person`s usage.