Wednesday, August 9, 2017

Planning Rehabs

It's been quite a while since I last posted.  I had planned to post for frequently but frankly just let it slide to the back burner.  I'll be looking to post more regularly from here.

Recently, I seem to have developed more business rehabbing Hyperion Planning implementations.  Although I have also completed two full implementations (one PBSC and one on-prem) and also worked on a major bug fixing/enhancement project, I've been called into a few clients who's planning systems were challenging or even borderline unusable.  Generally, the responsibility was shared between the company and the integrator.  But when I get the call, the company has tried to move past blame and is interested in salvaging their investment.  Happily, I've been able to re-engineer the systems successfully.  Given the current business environment in which many company's have squeezed budgets and time frames and in which they have been sold a bill of goods that planning systems can be build faster and cheaper than ever, I believe there are a lot of struggling planning systems out there.

Rehabbing such a system can be very complex.  Often documentation was squeezed out.  Requirements were unclear and are now remembered poorly.  As-built quality can be adequate to very poor.  Stakeholders are generally frustrated, discouraged, even giving up.  However, there is a path to recovery.  I recently completed such a rehab for a manufacturing client who's implementation in 2016 was overly complex and frustrating.  Their partner tried to build to every requirement and failed to document much of the system.  The developers showed an understanding of Planning and Essbase at a mechanical level.  But they did not seem to have had a very experienced architect, so the cube and rules designs were sloppy and inefficient.

In addition to needing a rehab, the company had decided to upgrade from to  They created all new instances which provided a clean new platform but also meant additional manual migration since LCM is not completely effective across releases.  This was completed prior to my engagement.  I would have recommended an in-place Dev upgrade and new QA and Prod systems.

My approach started with realistic scope and a clear project plan.  I insisted that the business users be engaged and dedicate time from their schedules as demanded by the project pplan.  Although there were a couple of new major mandatory design requirements, the project scope was limited to those enhancements and the upgrade.  During the project, I was able to redesign specific parts of their application where the bang for the buck was high and as time allowed.  But generally, a significant part of the project was keeping the project in scope and helping the client to start an enhancement tracking system so that out of scope requirements had a landing place.  This helped the users feel that they had been heard and their requests would not be lost.

I recently completed my involvement in the project, leaving behind a functioning system ready for 2018 planning, 200+ pages of documentation, and a trained internal architect to take the reins.

During the project, one of my partners, American Partners (@AMPTalent) offered me a speaking slot at Kscope17.  Given my recent experience, rehabbing planning implementations was a natural topic.  I am pleased to report that the presentation went smoothly and the reviews were very positive (4.5 stars out of 5).  If you would like a copy of the presentation, drop me an email.  If there is a recording, I'll link to it when it's available.

These sorts of projects can be very challenging and require a range of high-end skills (business, technical, management).  But they are very rewarding when completed successfully.

- Jonathan Cohen

Wednesday, February 26, 2014

Hyperion EPMA Data Synchronizations with Substitution Variables

EPMA Data Synchronizations are a fairly simple way to move data between Hyperion applications and to load data into Hyperion from external flat files and interface tables.  The interface is very simple and easy.  But one thing you will not find is the use of substitution variables.  There seems to be a conception that one cannot use subvars in the synchronization filters -- which is quite limiting.

There are a few hints that the design is supposed to support subvars, but there is no mention of it in the documentation and nothing on the boards.  None of this is especially surprising as this is a fairly lightly used capability.  ODI, FDMEE and other technologies have far more features.  Even Essbase partitions are more flexible - only for moving data between Essbase applications.

However, it turns out that subvars are fully supported.  Each dimension in the Synchronization definition can be defined with a filter as one of EQUAL, LIKE, or FUNCTION (from the EPMA Admin Guide):
  • EQUAL—any valid member, such as Actual for the Scenario dimension.
  • LIKE—use wild cards in the filter:
    • * for multiple characters (Only use * wildcard characters at the end of a string)
    • ? for single character
  • FUNCTION—use any valid member selection function, such as BottomOfHierarchy() retrieves the base level members for the dimension.
Nothing about variables.  Turns out, when putting anything in as EQUAL or FUNCTION, the tool surrounds the values with double quotes, and Essbase does not interpret it as a subvar.  But define the filter as LIKE, it is sent to Essbase without.  And Essbase treats it properly.

So, now you know the trick.  A little used but simple tool just became more useful.

For more information, check out the Oracle® Hyperion EPMA Administrator's Guide for version 11.x.  (Oracle EPM Documentation)

Wednesday, December 18, 2013

A New Beginning

Welcome to the first days of Analytics InDepth!  (Full disclosure: actually, this is the first blog posting, but the company is now almost two months old.  Someday, we’ll look back and it’ll all seem the same.)

The back story in brief

So, after 10 years of DW/BI/EPM consulting and climbing the consulting firm ladders, I finally decided to cut out on my own.  I’ve toyed with the idea for a few years and even consulted with a lawyer and an accountant last year.  But I always held myself back, unwilling to take the risk.  However, 2013 was a very mixed bag, as I started the year having a blast - focused on technical architecture and technology strategy that I’ve most enjoyed.  But changes in my firm required movement more into project management.  And put many of my long-term initiatives on the back burner.  After discussion with my firm’s senior management, I decided to move on.  So, I did what I’ve done before - reached into my network and started interviewing.  Two revelations emerged from my conversations.  First, I have developed an unusual and valuable combination of skills in with BI and EPM.  Second, at my level (Director), the boutique Hyperion consultancies had basically similar needs requiring more and more PM/AM/EM work and less overall interesting (to me) technical architecture and strategy.  It came to one day standing outside Bertucci’s ready to go in for lunch, I had a real conversation with our VP of “people”.  After about a half hour, I had a sudden epiphany -- I was wasting everyone’s time trying to hedge all risk by not making a firm commitment.  So, right there I shifted and set my end date for three weeks out.  I knew the outcome of this was going to force to me out on my own.

A business established

And on my own I went.  First stop were my archived files from past exploration into starting a business.  Second, back to my lawyer and accountant to understand more details.  And, of course, into my network to find and develop opportunities (more later).  There were quite a lot of details to get the business established correctly -- especially from a liability standpoint.  Incorporating is important, but going out alone means that the protection afforded in the past by my employers moves to my shoulders.  Here’s the place where trusted advisors are critical.

By the way, if you are in Massachusetts and need excellent, trustworthy advice:

Surinder Ahluwalia, Esq.
Robert S. Fineman CPA

And the race was on:
  • Incorporated (Massachusetts S-Corp)
  • Filed with the IRS
  • Business name
  • Bank account (Randolph Savings Bank - nice service and free business checking)
  • Mailbox (UPS Store)
  • Web site (registrar and host:
  • E-mail/Calendar/Docs (Google apps)
  • Business logo (got a freebie from VistaPrint - nothing fancy - price is right)
  • Business cards (VistaPrint)
  • Checks (VistPrint)
  • Corp Credit Card (AMEX)
  • Linkedin Corporate Page
  • Business insurance (Rodman Insurance)
  • Payroll processing (Advantage Payroll)
  • Accounting software (Quickbooks online)
  • Computer (Dell outlet)
  • MS Office - necessary for Smart View (Amazon)
  • Hard drive for external backups
  • Receipt scanner (Doxy)
  • Cloud-based backup (Crashplan)
  • A blog (duh - here you are)
  • Twitter (@AnalyticsDepth)
  • Fax number (Yes-it’s still needed:

There are still many details like setting up Quickbooks and generating invoices.  Plus, as business starts to flow, I’ll want to get a retirement plan up and going.

Speaking of business, all this activity is wasteful, unless I actually get some work and make some money.


The Oracle EPM/BI space is pretty hot right now and projects to continue to build on the economic recovery.  Companies are flush with cash and looking for productivity enhancements.  Riding Oracle’s coattails is also a good play now, as they make the strong case for Hyperion-powered productivity gains.  Over the last decade, I’ve build up a good network and understanding of the players in Hyperion integration.  Experienced Hyperion headhunters (especially Jim Burke, Dan Counts, and Anthony LaScola) have proven to be active in placement and, as importantly, have been willing to help me get established. 

With hands-on architecture and design experience plus years as a BA and PM, I am able to fill many different roles for clients.  One reason to go independent is to remain technical.  Fortunately, there are quite a few such opportunities.  I’ve reached out to former managers and colleagues; networked with headhunters; and used some job boards (especially Dice).  Given the time of year (October), my primary short-term objective was to land a position that would carry me into the new year, preferably through January.  Through one of the high-end headhunters, I landed a subcontract with one of the big integrations firms.  They had recently gone to production with a large Hyperion Planning system and needed some help with post-production support and enhancements.  Right in my wheelhouse.  It’s been a great four weeks getting my hands dirty and helping the more junior team members support the system.  The people have been great at both the client and the partner.  I’ve been extended two week into January.  And perhaps it’ll go further?

In the meantime, I continue to network for new people, companies, and opportunities.  It’s been fun and keeps me on my toes.

What this blog will be about

20+ years of technology experience -- including 10 years of focus on Oracle technologies, Analytics InDepth helps companies maximize the value of their Oracle EPM (Hyperion) and Oracle BI platforms. With strategic architectural focus, we build, maintain, and optimize your critical Business Analytics systems.

That’s the statement of focus.  And this blog will support that.  But its purpose is even larger.  I intend to use this platform, plus the web site, twitter feed (@AnalyticsDepth), and whatever else to exchange ideas and solutions in the technologies.  I have also learned through the years that having information is meaningless useless it is shared.  So this blog will also offer interesting ideas, technical solutions, and links to useful information.  Sometimes the blog may seem to wander a bit, but it will always be about my business.

I look forward to future adventures with whoever wants to go on the ride.  Your comments are welcomed.

And any business leads or ideas are always appreciated.

Thank you,
Jonathan A Cohen
Managing Principal
Analytics InDepth, Inc.