The future of SAP Business Applications

[Former Member]

A senior executive at an SAP customer asked me this week why they should choose HANA over IBM's DB2 database. I work in consulting, so I can make money implementing either solution and I don't currently resell SAP software so I gave an honest answer:

SAP is pouring development dollars into HANA, and the most fascinating innovations in business applications that SAP has to offer, run only on HANA. To name a few:

- The next-generation SAP Smart/Simple Financials and the rest of the Suite that will follow

- All the BW 7.4 Data Warehouse enhancements for logical modeling, data virtualization and simplicity

- The SAP BPC Unified Model

- HANA Live Operational Reports

- 70% of Fiori Apps that combine transactions and analytics

Is SAP a legacy application in your business?

If SAP is a legacy application in your business (or in a specific area of the business), then you would be smart not to spend money on it. I deal with customers who have business operations that are non-strategic, or geographies they look to sell or close. Those customers are often characterized because they moved to support vendors like Rimini Street to reduce the continued cost of ownership.

If this is the case, and you just want to keep the lights on for a particular business, then you would be smart not to move to HANA for this area, because there must be more pressing business concerns. In this case, you should stick on whatever database vendor offers the absolute lowest cost of ownership, which is probably the one you're on already, unless you're running your business on a Mainframe. Buy VMWare and get the cost as low as possible.

Or are you looking to innovate?

If you are looking to innovate, then you need to balance cost and benefit. For some customers, they do this in a point way and innovate in the cloud, using Workday, SuccessFactors, SalesForce, Ariba or one of many others. From the perspective of a specific business objective, the world has moved back to best of breed solutions in the cloud.

However this has not happened for the core of most businesses. Workday and Salesforce are trying to enter the financials market, but there is barely a challenger for the core of SAP's business - the business network and supply chain. When it comes to moving products around the world, between thousands of warehouses, with tens of thousands of employees and thousands of outlets.

Core ERP is alive and kicking.

What about User Experience?

Earlier in the year, I suggested that SAP should free Fiori. My argument was that SAP needed to protect innovation in the core, which was where the maintenance dollars come from. Charging customers to do this would slow adoption, and potentially increase adoption of other solutions. SAP announced this in June, so Fiori now comes without a separate license cost.

Fiori is a wonderful paradigm for end-users, because it makes interacting with a SAP system pleasant. Gone are the green SAP Gui screens, replaced by bright, easy to see tiles that work on any device and which simplify the business process from a user perspective. Fiori can't "simplify" business, but it can make the interaction easier.

Won't other databases catch up with HANA?

Oracle, IBM and Microsoft all have their own "in-memory" version of their database now, and to the onlooker, it looks like a "speeds and feeds" battle which will even out over the next few years. Certainly, for certain simple use cases, Oracle 12c in-memory runs less than one order of magnitude slower than HANA. IBM and Oracle will tell you that they are catching up and there will be parity soon.

What this doesn't explain is that HANA was built from the ground up using all the database knowledge garnered over the last 30 years, and IBM, Oracle and Microsoft are trying to retrofit this into their existing technology. Based on what they have achieved so far, this isn't about "speeds and feeds" and they are falling further behind.

And even if they do catch up from a speeds and feeds perspective, the HANA database includes geospatial, predictive, graph engines that are embedded into the database and their usage is being embedded into the next generation of business applications that, for example, will constantly predictively optimize workforces based on the geographic location of workers, parts, and work locations.

Hasn't SAP has publicly stated that it will support other databases?

Yes, but let's take the example of HANA Live (SAP's real-time operational reporting system for ERP customers). HANA Live is content that resides in the HANA Database, and is HANA-specific. It can't be ported to Oracle or IBM or Microsoft, even if they solve the speed problems that they have currently. If you don't have HANA, you can't have this.

So whilst SAP do support other databases, they aren't spending customers' maintenance dollars on innovating on them. There has been a nod to IBM with some limited support for BW on DB2 BLU, but BLU customers will get some acceleration with none of the modeling, planning, or predictive optimizations that HANA customers get.

So SAP is spending maintenance dollars on HANA, isn't that wrong?

You could view it that way if you wanted to, but I'd put it differently. SAP is in an extremely competitive software environment where customers have a lot of choices. They are at a time when they have a slight dip in core license sales, made up for by the sales of the HANA database and cloud acquisitions.

Investing in HANA provides significant benefits for customers, a lower TCO, and exciting user experience options. If they didn't change the way they operate, remove fat from the business and invest heavily in HANA, the industry would slowly gnaw away at that core license revenue. As it is, there are real reasons to do new SAP projects now, on the HANA platform, to delight users and deliver value. That's good for SAP and good for customers.

What about vendor lock-in? Will SAP squeeze me down the line?

This is a concern for customers who are considering moving their database to HANA - will SAP come and squeeze them for more dollars. My experience of SAP is that they generally have a good relationship with customers. User groups like DSAG and ASUG keep them on their toes, but the relationship is quite harmonious. Not so with other vendors, though I won't name any names.

So yes there is a risk of vendor lock-in, but I would like to point out that the alternative - not investing in HANA and getting the benefits - is much worse than the risk of vendor lock-in. I wrote to SAP this week asking for some clarification on this, and if I get a response then I'll post that here.

Final words

If you are a SAP customer, and your strategy is SAP applications, then you need to align your applications strategy with SAP's applications strategy and roadmap.

That roadmap is HANA - in the cloud, or on premise, but HANA. You don't have to move today, but you should start planning.

[Former Member]

Hi John,

great sum up as usual. I beleive people is trying to understand that SAP is really serious about building future on HANA. Once this perception is digested through the community, I beleive SAP customers will rush into HANA. My prediction, this will happen in second half of 2015. Also H/W cost should be more reasonable in order to support the migration to HANA.

Cheers,

Sarhan. @beyanet_tr

[Former Member]

Thanks Sarhan. Yes I think on the first point you are quite right.

However I don't think they will rush - like Jim always said, it's a multi-year opportunity. ERP is serious business and customers will be cautious, and that's no bad thing!

Do you think HW cost is a serious consideration now? With HANA SPS09 we have multi-tenancy, we have virtualization, we have cheaper non-production hardware with fewer cores. In my customers, HW isn't the major limiting factor. Of course HW will continue to get cheaper, and SAP will support better and more extensive data tiering.

[Former Member]

Ok let me say this way : they need to rush, as you clearly listed reasons above :smile: .

For big enterprises HW is now not the most important barrier, but still important issue for SMEs.

[Former Member]

Great article .. a lot of progress the last few years to make HANA enterprise ready at a big scale. Still some work to be done on the operational management side but it is moving fast.

I would say as usual with SAP their pricing strategy is preventing exponential adoption.

That's a pity from a technical point of view.

HW cost is nothing vs the SW cost ....

It is not the first time we see SAP pricing strategy preventing adoption ... 

Yes it is not simple to decide the right pricing strategy.

As I am focusing more on the technology and not belonging to SAP I would prefer a pricing strategy enabling faster adoption by the  customers.

[Former Member]

Thanks Patrick - for various reasons SAP are not always this explicit.

I'm not sure of your relationship with SAP (Customer/Partner/Independent etc.) but I'm interested in you comment on pricing. Could you elaborate more specifically? What are the specific problems on pricing you would like to see solved and in what context?

Sometimes the wrong price bundle can be put forward by an account rep, which can drive the perception that you describe.

In the customers I am working with, we see software/services/infrastructure including internal costs to be around 1/3 each now for installed base customers. SAP have some pretty neat promotions most of the time - this quarter it is 15% of SMBV (software you paid for), and it includes HANA, ASE, ERP, BW, etc., IQ NLS and lots more between.

Certainly I think SAP intend HANA to have a fair market pricing - 15% is the same as IBM or Oracle.

[andreurban_blumberg]

Good points as usual John. No doubt the decision for new systems or green-field projects is straightforward for HANA. For existing live systems, especially those critical core systems you refer to, that is a lot more challenging given complexity and risks involved in any type of migration project. The sidecar approach doesn't cut it long-term in my view given higher TCO. The risk I see with SAP primarily investing in HANA and often creating artificial restrictions that new features require HANA and not providing DB agnostic offerings in such cases means even traditional loyal SAP customers will go elsewhere for new requirements In this new best of breed (cloud) world. As HANA matures and services skills are more readily available the risks and costs of migrations will hopefully come down and make it easier In future. It will take time. Agree hardware cost is not the issue.

[Former Member]

I do not want to give a very precise answer (confidentiality issue).

Let me give other examples of wrong historical licensing price : SAP screen personas, FIORI. Adobe Interactive Forms ... For the 2 first cases SAP has resolved it but the original licensing model was not right (I highlighted it in a blog comment few months after the initial FIORI launch).

Let me tell you story.

The first we were exposed to SAP R3 (last century) one of my colleague came back from a SAP meeting telling be we could do online invoicing (no need to run in batch).

Do I need to tell then truth after implementation ?

We all know that SAP project is not a piece of cake ....

For HANA introduction on the ERP the ideal would be to introduce it as a technical upgrade. It means the HANA SW license should not be more expensive than my current DB license (it should be even cheaper) ... and it is not the case.

That's SAP is using the business value buzz word to sell HANA.

I see the long term value but it will be easier if the license price could enable a technical upgrade approach.

[petr_solberg]

Hi John,

nice summary.

Agreed, make no doubt about it, Hana is the roadmap, and it's just a case for every customer, when, not, if.

A little birdy told me that enhancements to business suite applications are being prioritised on Hana.

Best regards,

Andy.

[Former Member]

Understood, but remember you have to balance between confidentiality and posting. Comments need to be specific enough to be actionable!

Yes, I agree that sometimes SAP got it wrong. I disagreed with the Fiori approach and SAP listened to their customers, and made it free. Adoption is increasing quickly as a result.

As for the sales story, customers are pretty smart these days: caveat emptor. If you're unsure (even if you're sure), get it in writing and make sure it's in the contract. I don't think the story you mentioned is so common any more, because buyers got smarter. Unfortunately not every person in the field is going to get it right.

I want to pick up on two points, which are common misconceptions.

First, HANA is the same price at SMBV than Oracle and IBM DB2 - 15%. It's not cheaper, but it is no more expensive.

Second, for the same price, you get a LOT more. You get HANA for all Business Suite apps, plus ASE for those apps you can't or don't want to run on HANA. Then you get IQ NLS to reduce the TCO of very large databases.

Even more significantly, you get a set of grants to the runtime license that allow you to run analytics against the database directly, which can be a massive benefit. Plus, you get all of the new apps that run only on HANA.

And to add to that, in our analysis, whilst HANA has a higher TCA (RAM is more expensive), it has a lower TCO than other databases over a 5 year horizon.

[Former Member]

Thanks Andy, a follow-on post might be how to tell when "when" is.

That's what I heard too, and it makes sense: you can do new things that you can't do on other databases with HANA, which makes the Suite able to do new things that customers would like.

[Former Member]

Thanks Andre, and thanks for coming over to SCN! Lot's of good points here...

Sidecars are a short-term solution to solve a specific problem when customers aren't ready to move the whole app on HANA. The increase in cost has to justify a specific pain point.

There is also a lot of nuance between greenfield and brownfield in IT. Most customers are on a 3-5 year refresh cycle and building the business case for any DB migration is much easier when assets are already depreciated.

In most cases so far I've not seen the restrictions as artificial - the features built on HANA won't run on other databases (so far). Do you think there are artificial restrictions?

The cost of moving to HANA has already massively decreased - it's a combination of software, hardware, services and tooling, and the cost of each of those decreases year on year, as you would expect in a maturing market. Tooling is amongst the most significant, because DMO decreases migration cost/risk/time, NLS and dynamic tiering reduce the "hot" RAM size and better optimizations create far fewer memory bubbles.

We do seem to have (cyclicly?) returned to a best-of-breed market, built in the cloud in this incarnation. This means, for example, that Smart/Simple Financials has to compete with Workday and Salesforce, for example. So a customer has to choose between an update to ECC7 EhP7 and a migration to HANA, or a reimplementation on a best-of-breed product.

Make no mistake, SAP *must* make their option the most cost-effective and best of breed, or they will lose customers. For now, Workday and Salesforce have a less mature financials offering, so SAP have a little time. But the big mistake Oracle made with Fusion was to make the move from Siebel/Peoplesoft a complex migration, which meant that every Peoplesoft upgrade is a potential Workday cross-sell and every Siebel upgrade is a potential Salesforce cross-sell. Disaster!

As for the HCM market, Workday and SuccessFactors already fight it out for best-of-breed in the cloud. It's less clear how that will play out in the supply chain market, because there aren't any credible cloud players yet.

[petr_solberg]

agreed John, it was a perfect storm, and let's not forget the JDE customers.

Peoplesoft buying JDE and subsequently Oracle buying Peoplesoft was a birthday gift for their competitors.

Andy.

p.s.

<Me In Recommendations Engine Mode>

If you like this discussion you might like this one

</Me In Recommendations Engine Mode>

[Former Member]

John,

good summary. I find it always quite interesting when people talk about vendor lock in. There is like in anythink in life a up and downside to that. Upside is for me in comparison the Mac or iPhone. It is simply working and has the advantages of being highly integrated (max level) and very well functioning.

For SAP I think it is logical to prioritise HANA since it offers a whole new scope of possibilities which have not been available before.

[Former Member]

Very nice and informative blog John! Even the comments above were quite informative.

One thing is for sure - the years ahead are gonna be dominated by technology and while HANA may or may not remain number one forever, it would possibly remain ahead of the competition for sometime for sure! It would possibly only take a new and innovative technology/competitor to uproot HANA.

[Former Member]

Thanks, and we should all be mindful that nothing is forever, and tech companies in particular rarely have a "second innings".

HANA is potentially the "second innings" for SAP, which keeps them relevant for another 10-20 years.

I remember being in a session with Bill Gates some years back and someone asked "Who do you think will overtake Microsoft?". His response: "If I knew that, I'd buy them"...

[Former Member]

Absolutely, and remember that SAP's database-agnostic design served them well over the last 20 years. When R/3 1.0 was released, most customers ran on mainframes, which make a very small portion of SAP customer databases today.

Over the last 20 years customers moved to Informix, Oracle, DB2, Microsoft, SAPDB, MAXDB. They ran on Operating Systems which have long since disappeared: Compaq Tru64, OSF-1, Dynix, Reliant. Manufacturers have come and they have gone. We moved from an age of Mainframe, to UNIX, to Windows/Linux and Virtualization. We are moving into the cloud era.

So now SAP is saying that the strategy has changed, and it's a testimony to the maturity of customers that they are questioning this. And up to SAP to justify their strategy is worthy of continued investment.

None of which is a bad thing :smile:

[WaldemarFal]

Dear John,

It sounds good but – you know – to convince the business we have to present clear facts. Are probably a benchmarks available?

And while the talk is also about obstacles to migrate I think that the main obstacle to migrate to HANA are not such the things like the unclear or not convincing pricing – in my opinion a lot of existing systems are not ready for migration. They are not even “real time processing” because of “heterogeneous” structures - not optimal interfaces and isolated instances built up through many years.

I am looking for support to convince business to harmonize our system and to plan the migration to HANA.

Regards

Waldek