Wednesday, November 28, 2007

What is an Analytic Application?

It’s a topic that’s received quite a bit of attention over the years in the business intelligence market. Here’s how we answer this question of “What is an Analytic Application?” at LucidEra…

Enterprise applications like salesforce.com, Oracle Financials, SAP, Quicken, etc. fall into the category of transactional applications. These applications help capture your business transactions and help you implement your CRM or Order Management business processes. They track individual transactions. But if you wanted to analyze this information (for example, which sales reps close deals faster and discount less?) you typically face the following challenges:

Deal information is in CRM system
Order and discount information is in the ERP system
Sales rep names across these systems don’t necessarily match
You need different techniques to read data from CRM and ERP systems because they’re usually from different vendors (although significant market consolidation has certainly taken place!)

Managers are frequently tasked with improving business processes. However, to do so they need to know where they currently stand. What factors will improve the processes? Are their changes creating positive results? An analytic application allows employees in a given function to analyze specific business processes by:

Measuring key numbers like Average Days to Close, Booked Revenue as a % of Quota
Analyzing the factors that affect the numbers like variability by Sales rep, deal sizes, Quarter, Region etc.
Tracking the results of efforts to improve the numbers

We believe that analytic applications should do the following:

Be focused on a specific business process. (For example, the revenue cycle.)
Automatically connect and read data from common applications like CRM, ERP, etc. (similar to reading bank statements automatically over the Internet).
Include a pre-defined format to store data that is optimized for analytical purposes
Contain pre-defined reports and analyses that employees in certain industries and roles need to improve business performance.
Allow you to create the relevant information you need in a self-service way as needed.

Of course, we also believe that the promise of analytic applications will finally be delivered thanks to the software-as-a-service (SaaS) on-demand model. If this is a new topic for you or you’d like to keep track of this market shift, I encourage you to sign up for our RSS feed, or better yet, join our the Friends of LucidEra Group on Facebook.

posted by Darren Cunningham at 4:58 pm

Comments

Ron Dimon

Posted on 29th November, 2007

Hi Darren,

I agree that Analytic Applications are distinct from transactional applications, and I like your 5 ‘ingredients’ to help define them.

I would add that Analytic Applications should add some value to data to help transform it into information.

Pre-defined formats could be one way of doing that, but you would also need to include hierarchy (roll-up) information - like master (or ‘reference’) data.

Other value-add information could be things like plan & forecast information, which usually don’t come from transactional systems, or cost allocations, etc.

Best,
-Ron

Henry Morris

Posted on 7th December, 2007

Hi Darren,

I coined the term “analytic application” ten years ago when I started the BI program at IDC. I still contribute to the program today. I was trying to move the discussion away from the flavors of OLAP which dominated research at that time, to the processes where the technology was used. More than packaged reports and a data model, I emphasized “process support” to note that there were analytical tasks that were repeatable, and provided the opportunity to package in an app.
As far as the types of analytic applications, I used IDC’s taxonomy of applications by business process and postulated what analytic apps would look like in each business area (and later in each vertical industry). At the time, pretty much the only packaged analytic apps were financial ones, specifically budgeting and consolidation.

But to your question as to “what is an analytic app”, here was the original definition I used 10 years ago which we still use at IDC today. It originally appeared in “Analytic Applications and Market Forecast: Changing Structures of Information Access Markets” (IDC #14064, August, 1997).

An analytic application must meet each of the following three conditions:
· Process support. Packaged application software that structures and automates a group of tasks pertaining to the review and optimization of business operations (i.e., control) or the discovery and development of new business (i.e., opportunity)
· Separation of function. Can function independently of an organization’s core transactional applications, yet can be dependent on such applications for data and may send results back to these applications
· Time-oriented, integrated data from multiple sources. Extracts, transforms, and integrates data from multiple sources (internal or external to the business), supporting a time-based dimension for analysis of past and future trends, or accesses such a database

Henry Morris
Senior VP, Worldwide Software and Services Research (IDC)

James Taylor

Posted on 14th December, 2007

Well I mostly agree with Henry here, as I said in this post on my ebizQ blog. I think an analytic application must DO something, take a decision, not just provide reports and cubes.

JT
Author, with Neil Raden, of Smart (Enough) Systems