SAP: What are the key features an analytics platform “must have”?

Kumar Singh, Research Director at SAPinsider, talks with Dave Henry, Senior Vice President, Strategic Alliances, about decision intelligence and how it is revolutionizing the way we make decisions.

In this video, Kumar and Dave examine the critical “must-have” capabilities an analytics platform needs to have for organizations on SAP.

Watch this video to learn:

• What analytics and IT leaders can gain from a unified decision intelligence platform that lets the business query their SAP BW and HANA data directly
• How the migration from ECC to S4HANA is spurring organizations to rethink how their analytics capabilities integrate with their core operational data
• Why integration of Snowflake, Redshift, or other cloud data warehouses need to integrate with existing SAP data
• How TCO can be reduced with a platform approach to analytics
• How decision intelligence can be a win for both business and IT leaders

Kumar: Hi, everyone.  My name is Kumar Singh and I am a Researcher Director with SAPinsider.  In the exploratory new series, we interview leading exports from the SAP technology ecosystem to discuss critical emerging and key topics related to SAP technology environment.  Now, many of you attended SAPinsider 2022 last month.  Out of many sessions focused on data and analytics, we had a few sessions from Pyramid Analytics, one of the enterprise analytics solutions provider, that kind of generated a lot of interest and engagement from our audience and we ended up receiving many, many follow-up questions related to those sessions.

So, to answer a couple of those questions that we ended up receiving from our audience we decided to invite one of the experts from Pyramid Analytics who has been in this talk show before, Dave Henry.  Dave is the Senior Vice President of Strategic Alliances for Pyramid Analytics.  So, Dave, thanks a lot for taking time to engage with us again and welcome to this talk show.

Dave: Great.  Thank you, Kumar.  Good to be with you again.

Kumar: One aspect of one of your Pyramid Analytics presentations was something that really resonated with our audience.  There was a where you guys touched upon [sic].  What were some of the key elements, aspects, or features that should be embedded within enterprise analytics tool that are really critical, must have, kind of important elements.  So, I think the specific slide that you guys presented was titled ‘When it Comes to SAP Analytics Details Matter’ and there were like five really, from my perspective and based on the feedback from our audience, were really interesting and useful insights.  And that’s the first point or like the first aspect that I would love if you could touch upon is what are those five key elements of an enterprise analytics tool or features or aspects that are becoming a necessity in today’s business systems environment.

Dave: Sure.  You know, I think when we first spoke, we talked about SAP Plus, right, is one of the strategies that Pyramid has, so supporting SAP, particularly BW, BW/4 HANA, and exploiting the features there.   But then the Plus aspect had to do with the external world.  So, at SAPinsider we talked about AWS and data lakes and how that fits in.  And that’s critical, but I think the slide you’re referring to deals with how Pyramid supports SAP specifically.  And so, I think we want to talk to that and why it’s important that BW and BW/4HANA features are supported natively and supported thoroughly, right?  That’s really what we want to talk about.

Dave: To kind of set the context for why this is important I want to describe some of the challenges that two groups of people have in SAP shops.  The first, of course, would be your analytics leaders.  What are they really concerned about?  And the second would be IT.  And I think you find when you put these concerns together it makes then a strong business case for accessing SAP, BW, and BW/4HANA natively instead of extracting the data.  So, if we start with analytical leaders I think, you know, it seems kind of mundane but one of the big concerns they have is maintaining and improving their daily operational reports, right?  Their sales reports that go out to department heads, anything related to finance or virtual closing of the books.  These reports get distributed sometimes to hundreds, in some cases we’ve heard of even thousands of individuals, and while this might be considered a legacy requirement it’s mission critical.  And these things have historically run directly against SAP and they need to continue to run against SAP.  So, that would be one goal.

Dave: The second would be really making decision-making accessible or data driven decision-making accessible by all levels of management while maintaining control over sensitive information, right?  So, one of the challenges, of course, when you start extracting data is you have to reimplement all that security logic, right?  So, that’s another case for leaving the data in BW, BW/4HANA natively.

Dave: I think the third thing that analytic leaders are dealing with, and we saw this and talked about it in our case study with Premiere Foods, is emerging requirements.  You know, in the past it was all internal usage, internal data and more or less a monolithic technology stack, but that’s changing.  So, we’ve got now a need to include technologies that are outside of SAP.  So, data that’s going into a data lake in a public cloud.  We’ve got third party data that we need to get insights out of, perhaps to help us comply with regulations or to document our environment’s social and governance initiatives.  So, you know, think about a situation where we’re trying to move things forward and sustain what we have and also deal with new requirements and change, right?  And that’s the analyst’s view.

Dave: And then if you look at it more from an enterprise architecture view, maybe more of an IT infrastructure view, you’ve got a couple big themes, right?  The first is migration, right?  The great migration going from ECC to S/4HANA.  Going from BW to BW/4HANA.  People have to make decisions.  What do I keep?  What do I leave behind?  You know, which of these objects are used, which are obsolete?  And, you know, just figuring that out and decommissioning things can be time consuming.  So, you know, having the ability in whatever new tool you might be adopting for some of these new environments to be able to continue to connect up and see all of the objects and all the content that’s in your existing landscape.  It’s really important, okay?

Dave: I think one that’s been talked about in the industry for several decades now is volume velocity and variety.  It’s really impossible to put all data everywhere.  It’s impractical to put all the data you have into SAP.  It’s impractical to put all the data you have into a data lake.  So, you know, what we’re finding is that customers are saying, okay, I don’t want to put everything in one place.  Help me.  And help me at the same time without forcing my users into different tools for different environments because that’s very, very disruptive.  So, how do I seamlessly incorporate maybe an Amazon Redshift or a Snowflake into my overall, you know, ecosystem and architecture.

Dave: And then I think the other challenge is that people deal with certainly, you know, what they have to give up if they extract data.  So, if you’re an enterprise architect and you’re thinking about moving that data into a data lake it dawns upon you as you start to do this how much value you have to give up when you move that data out of SAP.  So, you’ve got all this business logic that’s been created typically over years, sometimes decades, and you’re moving the data out but what about all the business logic, right?  What about all the security?  What about all the calculations and the formats?  You know, if you don’t recreate that you’re taking something away from your users and generally they don’t like that.  That stuff’s been in there for a reason.  So, thinking about that, dealing with that is certainly becoming more apparent to enterprise architects who are trying to figure out what does this infrastructure look like?  And then, of course, the overriding concern for everyone at C-level in IT is total cost of ownership.  So, how do I do this without breaking the bank?  How can I move forward incrementally logically?

Dave: So, those are the challenges.  And I think with that in mind now we can look at what are the key things that you need to support in terms of functionality?  And a lot of this is about BW and BW/4HANA themselves.  So, look, there’s a tremendous amount of granted proprietary but also very powerful and very useful capabilities that are part of BW, right?  So, the whole info provide concept and how that’s evolved over the years.  The idea of defining, you know, key figures and predefined queries and not only making it convenient for users but also maintaining some central definition over your key performance indicators ensuring that calculations and formulas are consistently implemented and used.  That’s important, right?  And that needs to be preserved.  So, feature number one would be, look, support those info providers and then the features associated with info providers.

Dave: I think a subset of that which we find to be important particularly for larger companies or companies that go through periodic restructurings, maybe they have merger and acquisition activities going on, are going to be, you know, your formatting of your data, right?  And so, there’s formatting and there’s hierarchies.  If we think about formatting and multinational companies, you’ve got multiple currencies, right?  You’ve got multiple units of measurements.  It could be weights, it could be distance, right?  You’ve got different definitions of political geography.  All of these things are important to the analytics themselves, right?  They’re modeled within BW.  You know, particularly when we look at financial reporting and multiple currencies it becomes critical that you have this ability to inspect the data if you’re a front-end vendor, right, thinking from Pyramid’s point of view.  You know, inspect the model and inspect the measures and figure out what the currencies are.  Because if you try to add up measures at a given level and the individual measures are represented by different currencies you can come up with an invalid answer.  There are some circumstances where you don’t want to do a subtotal.  On the other hand, everything’s in the same currency you can.  So, you start thinking about, well, okay, if I pull this out of SAP, I’ve got to all of a sudden model all this currency logic in my data warehouse and it’s a pretty heavy lift.  Why would you do that when you’ve already got that inside of BW?

Dave: So, preserving those formats and, you know, there’s all kinds of other capabilities that are there, captions, alternative captions, support for different languages and locales, that list goes on and on, right?  And we’ve been at this now with customers for about three years and we built out that list that you were talking about really by looking at how customers actually interact with BW in HANA and by helping them stay there in that environment.

Dave: So, the info providers that measure formats and then a couple of the others that are really important, hierarchies, you have to be able to talk about hierarchies.  They’re critical.  So, time dependent hierarchies are something that a lot of organizations use.  Why?  Because your business, your structure of your business or your structure of your product offerings changes over time.  You reorganize, you recategorize things and you can’t just look through it through a current state.  Lens — you have to support those slowly changing dimensions in being able to go back in time and looking at alternative hierarchies becomes really important.  Ragged hierarchies would be another one.  You think about financial reporting.  You think about a lot of slice and dicing tools.  If they’re basic they may not deal very well with a ragged or asymmetric hierarchies.  And yet, if you look at financial reporting in particular it’s super common to have asymmetric hierarchies.  Geographies are asymmetric.  Product categories.  You know, you can have a variable number of subcategories.  It’s asymmetric.  And so, can you pull that out of BW and use it as it is?  That becomes a critical requirement because for all these things if you don’t do that and the users are dependent on it then you are going to have to recreate that in that target environment, right?

Dave: The last one I pointed out that I think that comes up a lot, and it’s a differentiator, is Nuance, right?  Would be the whole question of parameters, variables and parameters, right?  So, they’re really important to implement security, to improve performance by starting the user out querying a subset of data.  You know, you want to be able to use parameters as flexibly as you can.  Okay?

Dave: So, those are just from a, what would you support, you know, I wouldn’t even call it table stakes because not all vendors do that.  What’s really important is support within BW itself.  Now, we can talk a little about, well, how does Pyramid do that, right?  So, I think the first thing people should keep in mind, and you can use different analogies, one analogy would be high fidelity, right?  You know, I’m looking at something through Pyramid and I’ve seen it in another tool that I’ve used, right?  Create business objects, could be analysis for Office, could be other tools.  Can I see exactly what I saw in that original environment?  Super important that we be able to take that direct capability from the engine and surface it.  So, high fidelity is really important.  Yes, and that means, you know, executing BEx queries or HANA views.  It kind of comes down almost to, you know, what you see is what you get.  Is, you know, what I’ve seen over here in this environment and now what I see in Pyramid, can I draw the conclusion that they’re one and the same?  And this is important if customers are doing any kind of migration because oftentimes, you’re changing multiple things.  Maybe you’re moving from BW to BW/4HANA.  You’re also considering new tools on the front end.  I’ve got to go through a QA exercise.  If I have to QA my presentation layer and I’m at the same time trying to figure out maybe the changes I had to implement because I pulled data out of SAP?  It’s just a heavy burden and we find it’s very time consuming, it’s labor intensive.  So, we try to ease that burden for people by just saying, look, if you’ve got it set up in this environment, it should look the same in Pyramid.  So, this is really important to our customers.  And a lot of our prospects, they don’t really believe it until they see it.  So, high fidelity would be a really important thing that we deliver.

Dave: I think another one would be this, again, going back to parameters, is flexible implementation.  So, we decouple the concept of parameters from data sources.  They’re not tied to individual data sources.  They’re reusable.  And so, you know, each time there’s a different query that needs to be run perhaps against a different connection, you know, parameters can be passed to those queries.  And we can do that by user, we can do it by report, we can do it by session.  You can do it by any combination.  So, okay, why is that important?  So, think if you’re an analyst or a developer and you need to look at, again, the information from different perspectives, maybe different user perspectives.  Imagine the time savings you get just from simply using a different set of parameter values when you go to connect, rather than having it be tied to the connection itself.  So, it applies there.  It also applies when you’re doing things like report distributions.  So, bursting of reports.  You’re trying to, you know, deliver publication content to different audiences.  It’s being able to iterate through different sets of parameters for different slices of data.  It becomes, it’s important and it doesn’t really take programming in Pyramid.  These are things you just use out of the box, right?

Dave: So, the next one would be, and I know we touched upon this at SAPinsider in quite a bit of detail, talking about how do you interact with the data source?  So, I think there’s been a general recognition that DirectQuery is important.  Most of the vendors in this space now support some form of DirectQuery against SAP, BW, BW/4HANA.  But, and there’s a big but, there’s two twists to this.  The first is that, so BW uses MDX as it’s underlying language, multidimensional expression language.  HANA uses Sequel.  And how do you deal with that, right?  We have a query engine, and think of this as an engine that takes calculation constructs that are defined at the user interface and maps them into an information model, and then translates that into a query, right?  That’s what we call our Piranha Query Engine.  It can support both Sequel and MDX and it does so universally, meaning that calculations you produce in one environment can be recreated in another.  Meaning that you’ve got this portability of calculations.  And so, whether it’s our Discover tool which is really for self-service analysis and visualization, or its Formulate which is our, it’s not a scripting language so much as a, it’s a calculation creator and editor, right?  Those differences between MDX and Sequel are handled by that engine.  Meaning that you can take views and calculations that are being presented to the user and you can resolve those down to either BW or to HANA very effectively.  And it’s all pushed down.

Dave: So, I think that’s the other thing is that we’re taking this and we’re pushing it down into the engine.  So, if you’ve invested in HANA, for example, you’ve operated from, you know, BW may be running against Oracle underneath it to HANA, you’re going to see an immediate performance improvement with Pyramid.  Why?  Because we’re not trying to calculate these things inside the presentation tier, right?  And so, the middleware tier.  We are pushing this down to the engines.  You get that lift from HANA.  That pop.  You get it day one.  And I think that’s really important.

Dave: The other thing that’s important is there are times when you want to recreate these calculations on data that’s in your data lake.  So, if you’re using Redshift or you’re using Snowflake and you want to have a synchronous or ragged hierarchies support, you can do that.  And one of the few tools that lets you create these models and do what you would consider to be sophisticated relational OLAP queries on these modern analytic databases in the cloud.  So, all of sudden you’re starting to see more parity and what, again, what does this translate to for a benefit?  It really means that the users can learn one environment and then start taking the skills they have, working with their SAP data and translate that over to the cloud.  And it’s not a jarring experience.  It’s very easy for them to translate that experience.

Dave: And then I think the final thing that’s important and it’s becoming really important to, because of what Pyramid has positioned, right.  So again, we’re positioning leave your data in SAP BW, right?  Leave your data where it is because you’ve got a lot of value there.  You don’t want to pull it out.  Is this ability to create publications, dashboards, that are multidata source?  So, you could simultaneously connect up to BW and BW/4HANA.  They’re obviously distinct systems but you could put them side-by-side, link them together, and have a visualization, right?  An interactive dashboard.  You could then if you wanted to add Redshift to that or add Snowflake to that or add PostgreSQL or MySQL, whatever you might have.  And as long as you have a shared dimension, a conformed shared dimension, you can filter and we’ll send queries down to those individual databases.  So, think of this as the modern version of a mash-up without extraction.  And so, this is really, really powerful because if you don’t have to move the data and you had a good data, master data management regime in place, you can just take advantage of that data being where it’s at.  Obviously, that’s a whole topic unto itself and we’ve got some interesting learnings from some of our customers about that.

Dave: But, you know, so you can see it’s this, I guess to summarize, SAP is so important you can’t just not use what you put there, right?  You want to up your analytics game.  You want to give users the ability to get more value out of SAP, BW, you know, leverage those investments, and also expand into these new environments.  We talked about the HANA performance improvements.  I think that’s another big thing to highlight for your users and for our audience.  Bridging these environments becomes important because of the mix and match.  And then, you know, for IT, right, now for the IT requirements it’s, you know, how do we refactor and change all this stuff underneath the covers without impacting our users?  And I think having a platform that spans SAP natively in the cloud, cloud analytic data sources, makes that easier to do, to change things without the users really knowing or caring, right?  The whole concept of automation, of centralizing administration, reducing TCO.  These are all things that are a win for IT.  So, at the end of the day you get a win for the analysts, you get a win for IT and, you know, serving the business you get an upgrade for your end users and the ability to move faster.  So, it’s a nuanced story but it’s one that’s resonating with our customers and prospects.

Kumar: Definitely.  I think the more critical aspects of your answer, Dave, were the ones that I think our audience is really, really going to find useful were the specific aspects are the intricacies or challenges of SAP ONE and that you mentioned, right?  Whether it was the query engine functionality which kind of helps the users of the technology tap into whether it’s MDX or Sequel, blending the BW and HANA aspects together to the query engine or disparate sources of data residing in SAP technology ecosystem and kind of helping create a single point of view leveraging your technology.  I think those are some of the aspects that kind of really highlight the fact that, obviously, it’s technology that’s making it all happen, making it all a reality.  But then it also highlights that at the end of that what a business analytics or enterprise or any tool is about, how it can make the job of end users.  And so, I think you kind of bifurcated it into, like, the end user, the hands-on user of the tool, and the IT and how it kind of, all those aspects that you mentioned, make their job easier.  And I think it all boils down to that, right?