When Tom Petty sang, “Hey baby, there ain’t no easy way out” he wasn’t referring to business intelligence (BI) reporting but he might have been. Current generation reporting engines, AKA data visualization or data discovery tools, market their products with statements like these, emphasizing quick development and ease of use:
- “The democratization of data is here. In minutes, create an interactive viz (sic) and embed it in your website. Anyone can do it— and it’s free.” (Tableau Products Page)
- “Easy yet sophisticated report design empowers your employees to design professional and telling reports in minutes not days” (Windward)
I like these tools, and I do believe that they can provide a leaner, more productive, and more informative approach to BI reporting than some more mature products. However, none is a silver bullet for all data integration and reporting woes.
In his aging but still relevant 1987 treatise on the enduring difficulty of software projects, Frederick Brooks details why there is “no silver bullet. There is no single development, in either technology or in management technique, that by itself promises even one order-of-magnitude improvement in productivity, in reliability, in simplicity [of software development].” Brooks argues that improvements in software engineering up to that time (and likely up to now as well) addressed some of the mechanics of application development but not the core issues, like the inherent complexity of software systems.
Likewise, the BI market finds itself awash in really cool reporting tools (seriously!) whose ads imply quick results with little effort. The reality is that success on the scale envisioned by many customers requires the planning, coordination, and integration associated with an IT project. I hasten to add that this is not a derogatory post on these tools. Rather, understanding the points below will help those adopting a new reporting engine to deploy it in a realistic way and then enjoy the advertised benefits.
I’ll leave the vendors to provide you with the good news. A visit to the Tableau, Qlikview, Spotfire, or Windward site will provide a quick and convincing presentation of the data gathering and visualization benefits of recent reporting tools. Here are a few points that you’ll want to think about that may not be apparent from reviewing their sites:
Data Integration: See my recent post on this topic. To briefly summarize, just glomming together data from multiple sources often creates rather than streamlines work. Different data sources tend to operate by different rules and varying quality standards, and many reporting efforts run past schedule and budget targets as developers reconcile inconsistent data sources and overcome unexpected data quality problems.
Report Standardization: Early adopters of reporting products often buy desktop copies for a few analysts, and then face the problem of separate analysts coming up with their own analyses and measures – and getting different results due to different methods. There’s obvious benefit to basing management decisions on reports with consistent business measures rather than analysts’ inconsistent desktop creations. Furthermore, according to CapTech Manager Cameron Snapp, another key point is the ease of use and efficiency gained by consistent look and feel. “Users creating reports should always consider parameter inputs and drill down interaction so that each report is somewhat reusable and consistent with others developed by other departments.” Standardizing reports and visualizations usually involves purchasing and installing the server version rather than the desktop version of a reporting tool, and having a team build commonly used visualizations with similar look and feel into one or more shared server apps.
Report Distribution: Speaking of shared server apps, demand can ramp up quickly for the easily generated, visually appealing displays that you can develop with these tools. Passing a few reports around by email can quickly grow to managed distribution over multiple servers as demand grows, and as managers recognize the data security, volume, and timeliness implications of informal distribution of critical corporate data. This transition works best if the organization sees it coming and plans early, bringing in experts in report server configuration, security, and management in time for early development to lay the right foundations for growth.
Version Upgrades: After you deploy the new tool, expect quarterly minor and annual major version upgrades of your chosen reporting tool. The new releases will include exciting new functionality but will also bring the need to carefully plan conversion of existing reports to the new version. At a minimum, conversion will involve regression test of all existing reports, and more than likely some will need significant revision to work in the updated tool.
As complex as all this may sound, it isn’t bad news. Although the marketing literature may discount these considerations, they are part of every reporting project. Early plans for data integration, standardization, report distribution, and version upgrades are keys to reporting project success, and enables the organization to reap the benefits of the leap forward in data gathering and visualization that Qlikview, Tableau, Spotfire, Windward, and their counterparts provide.
One response to “No silver business intelligence bullets, but still a bright upside”
[…] so long ago I wrote a piece that attempted to inject a bit of reality into the claims then made by some data visualization tool […]