OLAP for finance users

In this blog post I want to cover the difference between OLAP vs what I’m going to call Relational/OLAP.  If you’re going to stop reading because you think that’s too technical, you don’t know my skills as a writer.  As always, this blog is geared toward the finance user.  Understanding the difference can be vital to deciding between OneStream and Hyperion. I’m going to use People Planning/Workforce to illustrate the difference.

A complaint I’ve heard over the past couple of years has been that Oracle only wants you to use Hyperion their way.  Too many companies have stayed with the outdated on-premise version because they don’t have the same flexibility in the cloud. As a result they’re staying with very outdated software.  I felt the same way.  When looking at the competition, I found one that stood out with the flexibility I used to enjoy.  Not only does OneStream have the old flexibility, it’s managed to expand on it through further development.

Let’s start with the technical stuff.  Stay with me through this; OLAP stands for OnLine Analytic Processing.  aka data cubes. If you started in the Hyperion Planning era, it’s the underlying database behind the scenes.  It’s a straight aggregation engine.  Load the data and it totals up.  The first iteration of these was developed in the 80’s.  They’re still a great technology and widely used, but I’m starting to feel like it’s outdated.

Relational/OLAP is what OneStream uses.  The data can be treated either relationally or as an OLAP cube.  Consolidation can be done as a relational database and then you can take the same data and turn it into a cube.  This is why Hyperion has traditionally had two products (One for Consolidation and one for Planning) but OneStream does them both in one.

Hey, this theory is ever so much fun, but how about a real life example?  Let’s compare Hyperion Workforce to OneStream People Planning

Hyperion Workforce requires you to add each person individually to a dimension.  Benefits can be calculated by person, but I recommend a blended average as more accurate (Look a rabbit hole!  Let’s not go down there).

Each of the people on this form had to be entered individually into the outline.  And then kept there.  For as many years as you want their data to be reported on.  That means lots of maintenance.

OneStream has all the people in a relational table. You don’t need to update the metadata.  Everything is done during the conversion from relational to OLAP.  Benefits calculations can be changed far more easily.  You can still see individual people by drilling in.

A Hyperion form tracks hiring, firing and raises.  These are all things that can be changed in a columnar in OneStream.

That theory has been enjoyable, but how about some practical examples;

Hyperion at Very Large Bank.  VLB had 30,000 employees and wanted to enter them all.  We argued for position based (branches) as being more efficient (something we had done elsewhere).  VLB really, really wanted all of them.  So we put them in and it made a very, inefficient cube.  It took us an extra month making it efficient.  Plus there was going to be turnover in the future. Let’s say 10% a year which is average for banking. If you used that database on 10 years of data, there would be 60,000 names now.

If we had built this in OneStream instead, no problem.  Just load up the names each year.  Adding the new people would not add significantly to the speed and there is far less maintenance.

What if I need to do a cube based version in OneStream?  We recently completed a project for a construction company.  They not only entered the employees and their benefits, but also had complex needs for utilization of their employees.  Constantly adjusting them by month.  So we built a cube based version for them.  In this case it has all the names to maintain, but they understood it was what they were asking for.

So I could do the same in both?  Yes, but I had a choice in methods with OneStream.  That’s the key.  A far more flexible tool.  I can get the job done in either, but there’s going to be more options on implementation and less maintenance in OneStream. It’s not just the implementation, but also easier maintenance after. Hyperion Workforce works if you do things the way they want you to do them, with little option and great difficulty in change.

If we can help you with your Human Resource planning needs, or you’d like to talk about how this also works for financial or operational data, please reach out anytime.

If you’d like to read more, the inspiration behind the OneStream part of this was borrowed liberally from Bob Powers here.