I've added a static page to the site (see the link at left) that contains links to my posts on the PRIME model from December and January. I feel this is a potentially very useful concept and therefore want to make sure it is easy for you to navigate. If you have not looked at these articles, I encourage you to do so.
Also, I have now completed the "core dump" of ideas that led me to begin this blog and so I won't be posting as frequently. I do plan to make comments on BI and analytics industry trends where I feel there is a business angle and will continue to announce new articles via Twitter (@thisbrucesmith). I'll also continue to tweak the content of this site in an effort to make it more useful and accessible.
Saturday, February 26, 2011
Thursday, February 24, 2011
Data Warehouse 2.0: In Search Of Meaning
Gartner issued a press release yesterday stating their belief that data warehousing is on the verge of major changes beginning in 2011. Two key changes highlighted by the release are increased demand for new types of information formats and applications, and increased demand for real-time data. Data consumers in the most forward-thinking companies are straining to go beyond mere reporting and analysis of business numbers to get at the context or meaning of business activity.
New formats and applications
In order to find this meaning, it will be necessary to go beyond traditional number-oriented data warehouses and incorporate text and media to a much greater extent than ever before. Companies that had toyed with text and media as data sources will now embrace them as integral to their data warehouses. This in turn will challenge business intelligence architects to ditch traditional ideas of reporting and analysis in order to present the new data in a way that helps the decision maker grasp the meaning he or she seeks. One thing I can easily envision is the increased use of tag cloud interfaces similar to what you now see in many blogs.
More data, faster
The other item, which we've discussed elsewhere on this site, is real-time data and business intelligence. This is what is driving the rapidly increasing interest in columnar databases, such as the one that powers Vertica Systems, whose acquisition by HP was announced and discussed here last week.
Behind the search for meaning
Larger companies will be investing heavily in the new technologies in an attempt to match the nimbleness and responsiveness of smaller businesses without adding large numbers of skilled analysts, who will be too expensive or not available. Growing businesses will need to be aware of the need to transition into these kinds of capabilities as they scale up through time and to plan their investments in new technology and data accordingly.
New formats and applications
In order to find this meaning, it will be necessary to go beyond traditional number-oriented data warehouses and incorporate text and media to a much greater extent than ever before. Companies that had toyed with text and media as data sources will now embrace them as integral to their data warehouses. This in turn will challenge business intelligence architects to ditch traditional ideas of reporting and analysis in order to present the new data in a way that helps the decision maker grasp the meaning he or she seeks. One thing I can easily envision is the increased use of tag cloud interfaces similar to what you now see in many blogs.
More data, faster
The other item, which we've discussed elsewhere on this site, is real-time data and business intelligence. This is what is driving the rapidly increasing interest in columnar databases, such as the one that powers Vertica Systems, whose acquisition by HP was announced and discussed here last week.
Behind the search for meaning
Larger companies will be investing heavily in the new technologies in an attempt to match the nimbleness and responsiveness of smaller businesses without adding large numbers of skilled analysts, who will be too expensive or not available. Growing businesses will need to be aware of the need to transition into these kinds of capabilities as they scale up through time and to plan their investments in new technology and data accordingly.
Tuesday, February 15, 2011
HP Acquires Vertica
HP announced yesterday that it is acquiring Vertica Systems, a company I had recently highlighted as a highly successful real-time BI and analytics vendor operating "under the radar" while racking up an impressive client list. I see two takeaways from this:
1. Obviously Vertica is "under the radar" no longer, but more importantly:
2. This represents a big pivot for HP.
HP had been widely rumored to be exiting the BI/analytics market. Their NeoView product had failed to gain traction, and they recently announced an alliance with Microsoft to market a business intelligence appliance running Microsoft BI and SQL Server on HP servers. My guess is this means HP is making a heavy bet on real-time BI and analytics as the next growth business in this segment and are buying in through the Vertica acquisition.
It should be interesting to find out whether this ultimately signals HP's intent to focus on real-time business intelligence and analytics as a niche or if there are more acquisitions in the offing as HP seeks to build a new BI portfolio.
1. Obviously Vertica is "under the radar" no longer, but more importantly:
2. This represents a big pivot for HP.
HP had been widely rumored to be exiting the BI/analytics market. Their NeoView product had failed to gain traction, and they recently announced an alliance with Microsoft to market a business intelligence appliance running Microsoft BI and SQL Server on HP servers. My guess is this means HP is making a heavy bet on real-time BI and analytics as the next growth business in this segment and are buying in through the Vertica acquisition.
It should be interesting to find out whether this ultimately signals HP's intent to focus on real-time business intelligence and analytics as a niche or if there are more acquisitions in the offing as HP seeks to build a new BI portfolio.
Friday, February 11, 2011
Business Intelligence And Privacy
In December of last year, when I wrote about data capture as part of the business intelligence data life cycle, I gave the example of a store clerk asking you for your postal code when you make a purchase. This week, the Supreme Court of the US state of California has ruled that this practice violates that state's consumer privacy law. One's Zip code is "personal privacy information" according to the Court. Unless knowing the postal code is absolutely necessary to complete the transaction (in order to know where to ship goods, for example) it is off limits to the seller.
California has a huge economic footprint; if it were a separate country it would currently have the eighth largest economy in the world, according to The Economist magazine. Thus the ramifications of this ruling are huge as well. In addition, California has long been considered a bellwether for other US states going back to the days of John Naisbitt's Megatrends books. As I pointed out in my December article, retailers use the postal code data they gather (among other things) to target mail advertising. If retailers no longer have the option of asking buyers for location data, is there an alternative?
Look for accelerated growth in loyalty programs. Retailers are increasingly instituting such programs, where customers who are willing to voluntarily give up some personal information can receive preferred treatment and/or discount pricing. This is particularly advantageous to companies with warehouse club membership business models, such as Costco.
As loyalty programs grow in number, watch for the rise of loyalty networks. After all, one can only have so many loyalty cards on one's key chain or in one's wallet. It doesn't take much imagination to see the value added by consolidating all of those loyalty cards on a single network card, or even in a smartphone app that can interface to each retailer. Travel and tourism businesses have been moving in this direction for some time. Websites are springing up to work the retail angle, including RewardPal and One Loyalty Network.
California has a huge economic footprint; if it were a separate country it would currently have the eighth largest economy in the world, according to The Economist magazine. Thus the ramifications of this ruling are huge as well. In addition, California has long been considered a bellwether for other US states going back to the days of John Naisbitt's Megatrends books. As I pointed out in my December article, retailers use the postal code data they gather (among other things) to target mail advertising. If retailers no longer have the option of asking buyers for location data, is there an alternative?
Look for accelerated growth in loyalty programs. Retailers are increasingly instituting such programs, where customers who are willing to voluntarily give up some personal information can receive preferred treatment and/or discount pricing. This is particularly advantageous to companies with warehouse club membership business models, such as Costco.
As loyalty programs grow in number, watch for the rise of loyalty networks. After all, one can only have so many loyalty cards on one's key chain or in one's wallet. It doesn't take much imagination to see the value added by consolidating all of those loyalty cards on a single network card, or even in a smartphone app that can interface to each retailer. Travel and tourism businesses have been moving in this direction for some time. Websites are springing up to work the retail angle, including RewardPal and One Loyalty Network.
Wednesday, February 2, 2011
Performance Management: Methodologies
In introducing Performance Management, we mentioned that it isn't necessarily a simple matter to come up with key performance indicators (KPIs). The underlying drivers of profit and cash flow may vary from business to business and industry to industry. Several methodologies have been developed in an attempt to create an organizing principal for performance management, and this article discusses a few of the better known or historically significant ones.
Baldrige Award
The Malcolm Baldrige Quality Award was established by an act of the U. S. Congress in 1987. The Baldrige Award established criteria for business performance excellence and was the first widely used methodology for performance management in the USA. The award is based on demonstrated adherence to numerous criteria, which effectively become KPIs. Since its inception, the program has diversified from a focus on general performance excellence to include industry-specific criteria in education and health care. For more detailed information see the National Institute of Science and Technology's (NIST) Baldrige Award site.
Six Sigma
Six Sigma is a quality control and continuous improvement methodology that relies heavily on statistical analysis. The key driver here is to minimize total defective parts per million opportunities (DPMO), and other statistical measurements around the quality process may be used as KPIs. Six Sigma is ideally suited for manufacturing businesses but attempts have been made to adapt the methodology to other industries as well. For more information, check the International Society of Six Sigma Professionals.
Kaplan-Norton Balanced Scorecard
The Balanced Scorecard (BSC) methodology, first introduced by two Harvard University professors in the early 1990's, is far and away the most widely used performance management methodology in business today. BSC focuses on four management dynamics: Financial, Customer, Process, and Learning. This methodology is also the source of great confusion, because other non-BSC performance management systems often call themselves "scorecards" as well. Because the Balanced Scorecard is so pervasive, I plan to devote a full article to discussing it in the near future.
Want To Go It Alone?
For those who don't want to be tied to a specific methodology, a resource to check out is The Performance Manager: Proven Strategies for Turning Information into Higher Business Performance. This book breaks down the business by functional area and gives you ideas for KPIs you can implement for each area. It's available in paperback at a very reasonable price from Amazon (but what isn't available on Amazon?).
Baldrige Award
The Malcolm Baldrige Quality Award was established by an act of the U. S. Congress in 1987. The Baldrige Award established criteria for business performance excellence and was the first widely used methodology for performance management in the USA. The award is based on demonstrated adherence to numerous criteria, which effectively become KPIs. Since its inception, the program has diversified from a focus on general performance excellence to include industry-specific criteria in education and health care. For more detailed information see the National Institute of Science and Technology's (NIST) Baldrige Award site.
Six Sigma
Six Sigma is a quality control and continuous improvement methodology that relies heavily on statistical analysis. The key driver here is to minimize total defective parts per million opportunities (DPMO), and other statistical measurements around the quality process may be used as KPIs. Six Sigma is ideally suited for manufacturing businesses but attempts have been made to adapt the methodology to other industries as well. For more information, check the International Society of Six Sigma Professionals.
Kaplan-Norton Balanced Scorecard
The Balanced Scorecard (BSC) methodology, first introduced by two Harvard University professors in the early 1990's, is far and away the most widely used performance management methodology in business today. BSC focuses on four management dynamics: Financial, Customer, Process, and Learning. This methodology is also the source of great confusion, because other non-BSC performance management systems often call themselves "scorecards" as well. Because the Balanced Scorecard is so pervasive, I plan to devote a full article to discussing it in the near future.
Want To Go It Alone?
For those who don't want to be tied to a specific methodology, a resource to check out is The Performance Manager: Proven Strategies for Turning Information into Higher Business Performance. This book breaks down the business by functional area and gives you ideas for KPIs you can implement for each area. It's available in paperback at a very reasonable price from Amazon (but what isn't available on Amazon?).
Tuesday, February 1, 2011
BI Concept: Performance Management
Once upon a time, each Monday the management of a typical business unit huddled around a large stack of reports on green bar computer paper that the data processing department had printed over the weekend. They were trying to figure out from these reports if they had met their sales or profit goals, or some other goal that was critical to themselves and/or to the business.
There is no time for that sort of thing anymore. The paper costs too much (to mention nothing of the trees). And wouldn't you rather have your key people working on improving the business instead of standing around a desk or conference table with their heads in reports?
If you've read elsewhere on this blog, you've come across my PRIME concept of how people relate to data, in short:
They Produce it by capturing, processing, or communicating it.
They Review it, as in a static report.
They Investigate it by drilling, slicing, and dicing.
They Monitor it for changes in data values.
They Extrapolate it to find future opportunities or threats.
The increasing progress toward business intelligence that is more mobile and delivered in real- or near-real time is forcing BI strategists to shift their thinking away from a "Review-oriented" presentation strategy to a "Monitor-oriented" strategy. This means moving from traditional static reports that are updated at regular (but too-slow) intervals to a targeted presentation of the key numbers that will allow a quick assessment of a business' performance, updated as conditions change. This "targeted presentation" is the idea behind performance management.
Performance Management vs. Reporting
Traditional reporting tells you that something happened in the business. It shows you a "bottom line," whether that is income, cash flow, headcounts, etc. It usually has some sort of formal, top-down or aggregate, presentation of the numbers. Traditional reporting often has a "one-size-fits-all" approach, or may have some flexibility to be run by department or region, etc. This may not sound like much, but it's easy to put together if you're familiar with the data and the reporting tool.
Performance management attempts to zero in on why something is happening. Rather than focusing on everything that adds to or subtracts from the bottom line, performance management focuses on those measurements that have the biggest impact on the bottom line. These are often called "metrics" or "key performance indicators" (KPIs). Ideally these metrics are compared against goals so that you can easily see if you are performing well or poorly. This sounds wonderful, but there's a catch: how do you know what's truly a KPI and what isn't? Thus it may take a lot more time (a. k. a. money) to devise a performance management system than to just pump out another profit-and-loss or cash flow report. We'll examine some approaches to performance management in future articles.
There is no time for that sort of thing anymore. The paper costs too much (to mention nothing of the trees). And wouldn't you rather have your key people working on improving the business instead of standing around a desk or conference table with their heads in reports?
If you've read elsewhere on this blog, you've come across my PRIME concept of how people relate to data, in short:
They Produce it by capturing, processing, or communicating it.
They Review it, as in a static report.
They Investigate it by drilling, slicing, and dicing.
They Monitor it for changes in data values.
They Extrapolate it to find future opportunities or threats.
The increasing progress toward business intelligence that is more mobile and delivered in real- or near-real time is forcing BI strategists to shift their thinking away from a "Review-oriented" presentation strategy to a "Monitor-oriented" strategy. This means moving from traditional static reports that are updated at regular (but too-slow) intervals to a targeted presentation of the key numbers that will allow a quick assessment of a business' performance, updated as conditions change. This "targeted presentation" is the idea behind performance management.
Performance Management vs. Reporting
Traditional reporting tells you that something happened in the business. It shows you a "bottom line," whether that is income, cash flow, headcounts, etc. It usually has some sort of formal, top-down or aggregate, presentation of the numbers. Traditional reporting often has a "one-size-fits-all" approach, or may have some flexibility to be run by department or region, etc. This may not sound like much, but it's easy to put together if you're familiar with the data and the reporting tool.
Performance management attempts to zero in on why something is happening. Rather than focusing on everything that adds to or subtracts from the bottom line, performance management focuses on those measurements that have the biggest impact on the bottom line. These are often called "metrics" or "key performance indicators" (KPIs). Ideally these metrics are compared against goals so that you can easily see if you are performing well or poorly. This sounds wonderful, but there's a catch: how do you know what's truly a KPI and what isn't? Thus it may take a lot more time (a. k. a. money) to devise a performance management system than to just pump out another profit-and-loss or cash flow report. We'll examine some approaches to performance management in future articles.
Subscribe to:
Posts (Atom)