From the Sage MAS Team

[#1]

Our business applications, in a pragmatic view, do two things.

1. Improve individual productivity by automating everyday tasks. This is where we have typically earned our name in the industry. We build applications (GL, AP, AR, SO, PO, IM, etc.) that enable an individual to carry out everyday tasks. In this process, our applications generate most amounts of structured data a company usually has. It does not take long before a business start to generate good amount of business data.

2. Help an individual make informed decisions. Amount of business data our applications generate aid an individual with the decision making process. Questions such as Who are our top customers; What is our top selling item; Who is our best salesman; Stock situation on our warehouse; Status on Receivables; various trend on a particular topic over some duration; etc. are regularly asked and responded due to our ability to mine the information available in our database. We allow business data to be diced and sliced multiple ways to help make informed decision.

In MAS products, BIE (Business Insights Explorer) – a simple yet powerful tool – come to help business users define what information to view; dice and slice it to meet an individual need; and provide appropriate drill down to know more about the respective transactions. BIE is available for both MAS 90/200 and MAS 500 product.

I will encourage our community to know more about BIE - if you don't already know about it. There are appropriate tutorials available within the product for you to learn and get going.

Just earlier today – I Google web for the MAS 90 BIE and found this video on Youtube introducing the MAS 90 BIE product. Credit for this BIE tutorial goes to Steve Liebreich working at Blytheco.

I will be curious to hear your feedback on BIE. I am particularly interested in knowing how BIE is helping your businesses and its adoption.

I will be sharing more thoughts around analytical abilities of our products in the next set of blog entries.

[#2]

Thanks for the opportunity to provide feedback on BIE. We're running MAS200 4.20.20. We have about 4200 Sales Orders and BIE won't load the Sales Order View. I've had to turn it off for my users because it locks up the user workstation (XP SP2) attempting to load, and I get a support call.

Love MAS, used it for years, easily the best in the market!

Steve

[#3]

I have a number of clients in which BIE won't work for because they have too much data.  The data most customers are interested in viewing that I've seen is invoice history.  What do you do when there are 200,000 invoice history records?  The system takes so long to load that it is unusable.

Dawn

[#4]

Understandable how it could take a while to sort through 200K records with just about any program.

Usually,  a client is not interested in all 200,000 records.  Possibly, history for a certain client or a certain date range is what is important.

Use a filter to narrow down the number of records loaded.  My experience seems to be that the time spent on the large number of records is in loading the data as opposed to searching it.

I would also imagine, that when loading a large number of records, the amount of RAM in the workstation may be relevant?  Not proven, just conjecture.

[#5]

As a long time user of Excel Pivot Tables, I find it very difficult to use the BIE features and thus demonstrate them effectively to the client's users.  They don't seem to be intuitive and straight forward.  In addition, you are locked into the data templates provided by Sage which aren't always what client's are interested in.  The obvious solution to Dawn's delemma is to set up an Excel worksheet with a query linked to a Pivot Table.  Put the spreadsheet on a shared network drive and demonstrate to the users how to open and refresh the table and let them drop and drag to their hearts content.  The obvious downside to Pivot Tables is that you have to have an understanding of the MAS data files which few clients do.  The up to this down side is there aren't a lot of tables that users access (e.g. AR_HistoryHeader, AR_HistoryDetail, IM1, AR_Customer, etc.)

[#6]

I use a 3rd party .NET components from a company called Janus Sys. These controls provide the exact same functionality as BIE but I have full programmatic control. Many of my custom management reports are built with this control. There is a lot of cool stuff that can be done and writing the report can be done in a couple hours. I am organizing a web site and plan to display more of this functionality in that site. If you have a client that could use a simplified version of the BIE please contact me.

[#7]

When entering for Customer History Invoices View, you can hit ESC to stop loading and then enter a date range or customer selection range which goes must faster than the initial load.

If they then save a setting with a date range or selection, it will be easier the second time around.

[#8]

One can experience performance limitation as the need increase to fetch huge amount of data.  Certain things can be done to fine tune and improve performance though like making sure there is enough memory; ensuring ODBC driver is setup in the CS mode;  judicious use of data row limiting; understanding how much data will be fetched and what will user do when large amount of data is fetched back etc. But it is a fair advise to be aware that large amount of data will take time.

BTW - we do plan to keep adding more popular views based on the need. Also, views can be added or customized with some customization skill.

From technology perspective though, and I will be curious to hear your thought - is using the same data repository to perform both enter every day transacations while also leveraging it for analytical purposes requiring trending information that may fetch large amount of data gets tricky. One can slow the system for other need. I have heard suggestions like replicate repository for respective use or do things at different time, etc. 

So what would you say to an idea that suggest we need two data model. One relational transactional data model serving transactional need. Other OLAP data model that near real time keep getting populated with time trend information to allow dicing/slicing of the various key trend across multiple dimensions? OLAP data model has inherent advantage in dealing with multiple dimensions and better suited for analytical need that depends on huge amount of transactional data.

[#9]

We have configured this for two clients, with very large datasets, to replicate to another server that is used as the reporting server. We have done this using SQL 2005 Transactional Replication. This is configured to be near real-time.