I'm comparing SAS VA vs. SAP BOBJ -- main differentiating factor?

Hi,

I suppose the effectiveness of the comparison also depends on the objective
of it. As far as the tools the Sas VA is not in the same category as
SAP/BO, which is more of a query and reporting OLap tool and although VA
has some of such capabilities it has embedded additionally models
with predictive intelligence. Both have space in the same organization as
User profiles are different and call for different tools.

SAS is known for tis statistics capabilities and lately is pursuing some
interisting partnership, sap is one of them for the optimal power of both.

In my company we have BO and just acquired Hana and SAS VA.

Best regards,
Maria Mota

Hi Anastasia (and the community),

Being employeed with SAS and working with Visual Analytics, I would like to provide some input and response to the above.

My point is not to compare BOBJ and SAS Visual Analytics, as my knowledge of BOBJ is too limited. However, in terms of Visual Analytics:

SAS Visual Analytics was developed for business users (end users) for reporting, business analysts building reports and exploring data, and for administrators/IT for managing the environment and providing users a self-service platform for reporting (BI) and more advanced analysis.

Visual Analytics was developed to support self-service BI and for business analysts to get support from more advanced statistical functionality with an easily accessible interface. The point of needing "data scientists" as stated in the above is simply not true. We have other advanced products targetting predominantly the data scientist, however for Visual Analytics being a data scientist is not a prerequisite.

Some customers use Visual Analytics for traditional business intelligence, others for self-service BI, again others for data exploration, and last but not least for hypothesis testing. So it supports different use cases.

Let me know if you have further questions on the specifics of Visual Analytics (or other SAS products).

Have a great day.

Best Regards,
Rune

Anastasia- can you share what you find with the community? Thanks!

Thank you everyone for your very helpful comments and expertise! We own BOBJ Lumira and will be researching more about Predictive Analytics features that it claims to offer. We will be doing more thorough analysis of business requirements and prototypes to make the decision.

Even though I am not an expert (and really know very little ) in SAS and SAP BOBJ, I wanted to give you a reply.
From my experience, if you are big on SAP BOBJ an BW, I would keep that and complement with an analytics db, built for the purpose of predictive models.
It is difficult to have one solution that will satisfy everyone, or a single product that excels at everything.

To replace BOBJ/BW with a new solution would be a very risky project, time and cost consuming, lots of expectations. Moreover, there will be enemies around trying to boicot the project. I would foster a new initiative to build an analytical repository, with the best tools to support a business initiative, and the results from that repository I would bring to your BI/Reporting solution (BOBJ).

Hope my comments are of any help to you.
Best regards
Marco

I agree with Director461.

Elisabete Miranda

Comparing SAS VA to SAP BOBJ is like apples and oranges, these products are from different categories, although there is indeed some overlap. SAP BOBJ is a classic enterprise BI platform whereas SAS VA is a specialty product for advanced analytics and data visualization. The products are actually complimentary, much like traditional SAS which often co-exists with enterprise BI tools.

So it all depends on your requirements as well as user expertise. If you need reporting, OLAP and dashboarding then you need SAP BOBJ or a similar platform like MicroStrategy (which we use). If you’re targeting professional data analysts that need to conduct sophisticated modeling and build fancy visualizations then you need SAS VA.

That is correct, SAS offers statistical and predictive analytics, more tuned for individual data scientist as a desktop user.

BOBJ on the other hand is well tuned for ad hoc and operational reporting, geared for IT to design and publish such reports for business users to consume.

Tool selection is very dependent on your business requirements and the audience. If your business users are not statistical function savvy and do not see value yet to make business decisions based on predictive and statistical models, SAS may not be the right tool for now.

Tool adoption is key for the success of a BI implementation.

Good luck and let me know if you need deeper understanding on tool selection.

Krish.

"We are considering SAS VA on top of HADOOP YARN platform. "
I think that would be an interesting combination. I hear HADOOP all the
time in the BI world. I have some experience in Cassandra but not HADOOP.

"What does SAS VA offer that BOBJ would not, and how can we build a
business case around SAS VA specifically?"
SAS is the market leader in BI market and specifically strong in analytic
area as you have already stated. The second area I would look at is the
data model which is the bridge between data store and analytic tools: how
efficient to construct flexible data models to plug to the algorithms,
third the performance of data retrieval which is significantly important in
production stage.

"Based on what I read so far, SAS VA offers a lot more capability in terms
of statistical analytics, predictive analytics and optimization – something
that BOBJ does not appear to have. Does that seem to be the main
differentiating factor?"
I don't have experience in SAS VA, therefore I cannot comment on it. IMHO,
algorithms are easy to add to a product suite, what makes the major factor
is how smooth and easy to flow data to the algorithms (data model).
Another part I would look at is the presentation layer, how easy report can
be created and intuitive to the business user to create report by
themselves.

I believe SAP VA have their own data storage, the comparison of data
storage should be consider as well such as scale out, efficiency... etc.

Regards,
Ray

People going to SAS must ask themselves first do they have any data scientists to implement the software. It is a leader in the Gartner magic quadrant for statistics for good reasons but to utilise its capacities you must have the appropriate staff to implement it.

You cannot compare SAS to BOBJ in reality as they are diiferent Tools. Indeed SAP now have their own statistic software called SAP Predictive Analysis which now competes directly against SAS. This Tool which came from the KXEN purchase, whilst not as mature as SAS at present offers quicker modelling capabitilities meaning it will be easy apply the solution and you do not have to have the specific thoroughbread data scientists in house.

As you are already a SAP Customer you should contact them for further information.

SAP BW are HADOOPS are of course extremely different beasts

Hi
SAS VA is a data discovery tool where business users can primarily do lightweight data mining activites on their own with the help of IT. Be sure that you have placed it right. It s counterpart in the BO world is Lumira which doesnt belong to BO BI Suite, thus paid extra.

We made a data discovery tool evaluation last year, Spotfire came to my mind at first as a DD tool with broad data mining capabilities. Have you evaluated it?

Can the tool seemlessly integrate to your hadoop Platform and perform well?
What are the data mining capabilities needed, who is going to to that business or IT/BI?
What are the front end capabilities, can the (business )users create reasonable dashboards on their own and publish them?
POC passed?

Those are the questions came to my mind

thanks,
Emre

Dear Anastasia,

Quick answer: correct analysis.

As always, it depends on what you want to do with the data, the kind of data, what results you expect to get and usually the costs of licenses…. SAP and SAS don’t come cheap and both can do a lot.

The core business of SAP is ERP with big data functionality added on. The core business of SAS is (big) data analytics with a strong focus on visualization with their VA solution.

SAP BOBJ / BW can be considered classic BI, SAP uses HANA for big data processing and analysis. www.sap.com/solution/big-data/software/overview.html But you probably know that already.

SAP offers free access to play with their tools so maybe just give it a try and see what it does with your data.

SAS often offers something what they call ‘a day with your data’ to show (potential) customers what VA can do with datasets. Check it out with your local SAS entity.

Note that with a tool like Tableau combined with Alteryx you can also do a lot of (visual) analytics when the amounts of (processed) data are not too large. And if they are, combine with e.g. HP’s Vertica and you can process any amount and file type you want.

Best regards, Oscar Wijsman

Wijsman Interimmanagement & Business Consultancy

in my opinion, SAS offers a much richer set of features that Business Objects offers. SAS requires advanced knowledge of SAS programming and domain of the tool, while BOBJ is more oriented to users less experts in IT. SAS allows, with its various modules, all ETL processes, advanced data analysis and reporting, while BOBJ is more oriented to reporting only, as far as I know.

Elisabete

Hi,
SAS VA (visual analytics) is primarily a data exploration and data
discovery tool. Since it is from the SAS family of products so yes it has
good integration with the predictive and statistical analysis engine of
SAS. In my opinion if you are a big SAP and BOBJ shop then lot of this
functionality is possible using SAP Lumira (it has integration with R).
Personally if your company is already invested in SAP, it does not make a
lot of sense to go with another enterprise class software. Having said that
if it is already decided that it needs to be bought then the key
differentiator (as I know) in addition to predictive and statistical
analysis is the visual analysis which is far superior than SAP and also the
in-memory feature of the SAS LASR analytic server. That is probably the
only difference (assuming you do not have HANA in your landscape).

Hope this helps,

Amit