Big data is not expensive or complex. While ‘expense’ is relative to an organization and its IT budgets, big data projects should not be as complex as to require an army of ERP consultants for installation, tuning, and maintenance. If you don’t grasp the business case or costs, SAP executives suggest you ‘drive a Porsche’ to appreciate the engineering behind SAP HANA. I believe an expensive luxury automobile is the ideal analogy for HANA. It is the cost of software, consultant staffing, maintenance, and other (application) premiums that makes HANA so expensive and complex. My only objection to the Porsche analogy is that users should get to drive their car right away! Unfortunately, you won’t ‘drive off the lot’ with SAP HANA until months of installation, testing, and tuning by ‘mechanics.’ Hope you’re not in a hurry?!
There are affordable and faster solutions for big data. Let’s forgo further use of an automotive analogy and state our big data perspective as, “You deploy on IBM when you don’t want to worry, you deploy on SAP HANA when you don’t want to hurry”
Myth- Big data is expensive and complex
A CEO, CFO, or CMO should be surprised by the number employees and consultants dedicated to deploying SAP HANA for big data. The cost and complexity become more apparent once you read the presentation, “A -to-Z Guide to Implementing SAP HANA: Planning Scoping, Staffing, Budgeting, and Execution: Part I” from SAP Partner Comerit. It offers useful insights into daily rates ($1,600 to $2,300 a day), time frames, and consultant requirements for SAP HANA projects. While many variables exist in implementations, I selected three Comerit scenarios to spotlight consulting fees organizations face when simply migrating existing SAP BW to SAP HANA. We'll use a daily consulting rate of $2,000.
2.7TB and $390,000 for Big Data?
A ‘small team’ HANA migration project required 14 weeks of effort with a core team of 2 company staff plus 3 consultants. The core consultants were dedicated 100% during the first 3 months of this 4 month project. Our estimated costs for those consultants was $390,000. (12 weeks/60 days x $2,000 day x 3 consultants)
5.6TB and $446,000 for Big Data?
A ‘medium team’ HANA migration project required 18 weeks of effort with a core team of 2 company staff plus 3 consultants, 2 Test Team Finance members, 2 Test Team SD & Commissions members, and 2 Test Team: Other Areas. Two core consultants were dedicated 100% during the 4 months of this 5 month project while a third consultant worked only 3 months. Our estimated costs for those consultants was $446,000. (16 weeks/80 days x $2,000 day x 2 core consultants) (12 weeks/60 days x $2,000 day x 1 core consultant)
38TB and $1,600,000 for Big Data?
A ‘very large team’ HANA migration project required 26 weeks of effort with a core team of 3 company staff plus 4 consultants. In this migration case, there were an additional five test teams, each of which required a consultant as “HANA Test & resolution lead” and dedicated 100% for FIVE months! Our estimated costs for those consultants was $1,600,000 (20 weeks/100 days x $2,000 day x 3 core consultants) (20 weeks/100 days x $2,000 day x 5 HANA Test & resolution consultants)
The bad news. You're not done yet.
Data scientists will learn there is more than waiting 14-26 weeks for HANA consultants, tuning, and testing. Beyond the consulting these consulting costs, you still have to buy an SAP HANA license and a hardware appliance. We’ll discuss that in another blog posting.
The good news. Be done in a fortnight.
IBM put our deployment expertise into BI Patterns so you can have big data faster at lower costs.. We created our big data solution, IBM Business Intelligence Pattern with BLU Acceleration on PureApplication Systems so you can have big data deployed and running in only 2 weeks, not 27 weeks!
Disclaimer: I have never owned or leased a Porsche.