December 1, 2010 at 9:51 pm
Comments posted to this topic are about the item The Need for New Technology
December 2, 2010 at 6:17 am
A friend in the business and I were talking recently about BI and data analysis. He works for a very large manufacturing company that recently spent a King's ransom on BI and enhancing their data analysis. Long story short he remarked that the company had spent a fortune to find out what they already pretty much knew. After all, the company has been in business for many decades, long before there were computers, and though they deal with lots of data, keeping the "pulse" on "where they were" actually did not require the huge investment they made.
We came to the conclusion that many young people have a distinct lack of vision backwards... That is, America has had big companies for many decades, long before there were any computers around. How did they survive?!?! They survived by manual analysis of data, and by knowing their businesses. Now we see way too many wonks suggesting that you must make huge investments to "know where you are".
Although I can see a place for BI and data analysis, what good is it to spend say, a half million dollars to tell yourself what you already know? I am not advocating we do away with what good systems can provide us, but I am wondering why so many people think you must make massive investments to know what you probably already know. For example, our company sells a lot of software here and overseas, but I know our foremost market is the USA. I don't need any computer to tell me that - I hear it and see it in the day to day process of my work.
The key phrase in today's editorial is: "spend the money wisely". But, please define "wisely". Is it wise for every company in the world to make massive investments to learn what they already know? That doesn't sound very wise to me.
BI and data analysis are valuable - no question - but that all hinges on what a business thinks it needs to know versus what a business already knows.
December 2, 2010 at 7:17 am
blandry (12/2/2010)
A friend in the business and I were talking recently about BI and data analysis. He works for a very large manufacturing company that recently spent a King's ransom on BI and enhancing their data analysis. Long story short he remarked that the company had spent a fortune to find out what they already pretty much knew. After all, the company has been in business for many decades, long before there were computers, and though they deal with lots of data, keeping the "pulse" on "where they were" actually did not require the huge investment they made.We came to the conclusion that many young people have a distinct lack of vision backwards... That is, America has had big companies for many decades, long before there were any computers around. How did they survive?!?! They survived by manual analysis of data, and by knowing their businesses. Now we see way too many wonks suggesting that you must make huge investments to "know where you are".
Although I can see a place for BI and data analysis, what good is it to spend say, a half million dollars to tell yourself what you already know? I am not advocating we do away with what good systems can provide us, but I am wondering why so many people think you must make massive investments to know what you probably already know. For example, our company sells a lot of software here and overseas, but I know our foremost market is the USA. I don't need any computer to tell me that - I hear it and see it in the day to day process of my work.
The key phrase in today's editorial is: "spend the money wisely". But, please define "wisely". Is it wise for every company in the world to make massive investments to learn what they already know? That doesn't sound very wise to me.
BI and data analysis are valuable - no question - but that all hinges on what a business thinks it needs to know versus what a business already knows.
I think you make an excellent point. One of the things I see is that people are reluctant to tackle time-consuming tasks. This puts a great deal of pressure on all the information technologists to just make things happen. That's fine as far as it goes, but without the ability to identify metrics and understand the relationships, they frequently have us spend time doing things that ultimately have no value. My experience is that if you can't do the job by hand, at least in principle, then there is no way for me to automate it.
December 2, 2010 at 8:51 am
I think many organizations are looking at automated BI to replace humans, wrongly thinking that all analysis can be done by machines, with a final result neatly appearing without effort. However, how much analysis is actually the result of a real human analyst poring through data and making connections that aren't obvious? This type of non-linear thinking is nearly impossible for a machine to perform. I'm thinking specifically of situations where a human sees two disparate pieces of data and makes a connection based on experience and intuition.
This is not to say that BI systems don't have a role, because they do, especially for the straightforward tasks like variance analysis, data aggregation, etc. The systems also help humans organize, sort, and display data, which makes the analysis that much easier. However, BI systems will never be able to replace all of the human qualities.
December 2, 2010 at 10:56 am
I think the need is the automation of the reporting and analysis, so you don't have people stuck doing drudge work. The problem is getting people that really do understand data and technology to implement it correctly.
December 2, 2010 at 3:59 pm
Although America and the world for that matter have had large companies for decades and even centries data analasis has also gone on that long. Sure we didn't have computers but computers were built specifically to help us understand the data. Even before there were vacuum tubed computers filling up large building having their tubes changed out every 8 minutes to understand military trajectories mechanical machines were used for tallying and breaking down complex equations. A large and complicated data warehouse is nothing more than just a nother too. The goal of it is to give your company another edge. Just as ENIAC was used for analyze trajectories to give the US military a competitive advantage.
---------------------------------------------------------------------
Use Full Links:
KB Article from Microsoft on how to ask a question on a Forum
December 3, 2010 at 7:19 am
The promise of BI and data anlysis tools is not to show you where you already are. Rather, it's to help you discover things that you may not know, and to help you discover those things sooner than your competition. If you can discover an opportunity sooner than your competitors then you have an advantage. Sure, you can do all your analysis by hand, but if you have a tool that lets you do it faster than your competition then you may have the upper hand.
That's the promise anyway, but the devil is in the details. If your data warehouse is only telling you what you already know then maybe it's not implemented properly. Maybe, it's giving you a warning that your business is becoming marginalized. Maybe you just really don't need a data warehouse. Maybe you built the warehouse for the wrong reason and your're not asking the right questions. Lots of maybes.
A data warehouse is to help the business answer questions about the business. If you build a data warehouse to answer the question "where are we now?" then that's probably all it's going to tell you. If you build a data warehouse to help answer the question "where are new opportunities for business" then you might get different (and maybe more meaningful) answers. A data warehouse is a tool built for a particular purpose -- to answer a question or questions -- but if your questions aren't meaningful then you don't have have much chance of getting meaningful answers.
December 3, 2010 at 7:35 am
To expand on the "discovering what we already know" point:
A company I used to work for knew that a lot of time was being wasted by back-and-forth between Order Fulfillment and Quality Control. Everyone knew this, everyone groused about it, everyone complained that "the other guy" was either (a) not paying attention to details, or (b) being far too picky, depending on which end of that line they were on. An average order took 9 days to process, and 3-5 of those were back and forth on that line. On the other hand, implementing that line of QC checks had reduced the refunds and lost customers by a staggering number, and the line had to stay in.
So I dug into the database where all the order data, including the quality control checklists, were kept, and discovered a series of patterns to the errors being found. By having a piece of code analyze the order before it was sent to QC by Fulfillment, we were able to reduce the average turn-around to 30 minutes instead of 3-5 days. This gave the company a HUGE competitive edge over all the competition, while reducing the refund and loss rates even further. Stress levels for the employees involved went way down, and it actually turned out to be a very effective training tool for new employees, since the computer could tell them in a simple list, the things they had possibly missed or that required a little extra attention.
Yes, I was essentially analyzing data we already knew. When I outlined the automatic "pre-QC" in a company meeting, everybody just nodded their heads and said things like, "yep", and "of course". But manually collating the data would have taken months, when I was able to get the database to spit out what was needed in a matter of hours, fully analyzing hundreds of details on several years worth of orders.
It was also possible to make it heuristic, in that if the QC line turn-around time started to go up again, it would automatically detect that, and re-analysis could be done very easily, and adjustment to the auto-QC could then be made easily.
Could it have been done manually? The analysis probably could have been, but the automation and heuristics could not.
- Gus "GSquared", RSVP, OODA, MAP, NMVP, FAQ, SAT, SQL, DNA, RNA, UOI, IOU, AM, PM, AD, BC, BCE, USA, UN, CF, ROFL, LOL, ETC
Property of The Thread
"Nobody knows the age of the human race, but everyone agrees it's old enough to know better." - Anon
Viewing 8 posts - 1 through 7 (of 7 total)
You must be logged in to reply to this topic. Login to reply