SAP and DSAG: Clouds, innovation, on-prem, and a bit of AI.

For old time's sake.

Image SAP stock photo.

These days I generally avoid commenting extensively on SAP ERP stuff, even though once upon a time I was Gartner’s lead analyst on SAP. I also worked at SAP for a significant chunk of my career.

My work today as a VC at Acadian means that I spend my time working with and thinking about early stage companies. But how those companies collaborate, enhance, engage and compete in and against the ecosystems of the large enterprise vendors remains an important component of our investment thesis.

But for old time's sake here goes

DSAG isn’t happy with SAP’s latest announcements. SAP’s relationship with its German user group is, as Facebook would say, complicated.

For background and analysis see Josh Greenbaum, Holger Mueller Register piece here. German coverage here. Diginomica has a good earlier piece too.

My reflections on SAP

The vast majority of SAP's R&D efforts over the 30 years have been about gradually and sometimes significantly enhancing the capabilities of R/3. It has successfully hauled the product into the "web" with mySAP.com, and then into enterprise services (whatever those were) with ECC and the Business Suite, then onto the HANA database, and then more recently, into various cloud or cloud adjacent platforms with the S/4 moniker.

Whether SAP could have used that immense engineering capacity to do something else is not up for debate here, but to have cajoled and caressed R/3 through multiple back-end, middle bits and front end renovations is an impressive technical feat. It deserves respect.

SAP’s sales machine has generally been remarkably disciplined at marching or shepherding SAP’s customers onto and through each of these waves, The consulting world has loved this too, as it meant large projects every decade or so.

SAP’s licensing department is probably the most effective in the industry, as those of you that remember enterprise support will attest. More recently SAP has embraced the subscription software movement. My former colleague at Gartner, Yvonne Genovese, brilliantly described these as “Licensing Moments.” SAP has been able to move customers onto new contractual models every decade or so, this is remarkable vindication of the stickiness of the products and SAP’s commercial acumen.

SAP’s marketing department has spent a good portion of its budget over the last 30 years trying to paint each of these renovations as massive innovations that are simultaneously somehow not disruptive. This has worked less well, as it is a Sisyphean task. To those looking for genuine architectural or radical innovation, they have oversold it, and for those seeking incremental advances, renaming, bundling and unbundling products merely creates confusion and mistrust. SAP now has more cloud varieties than a meteorology chart, and this doesn’t aid their cause. Marketing is sometimes guilty of saying what it thinks the financial markets want to hear, rather than what customers need to hear. This is an affliction that impacts almost every large listed software company.

A gnarly paradox

While moving an existing product of this scale to the cloud is remarkable engineering accomplishment, and shifting customers from a perpetual license to a subscription is a commercial innovation that investors laud, SAP has created a paradox that becomes harder and harder to resolve.

Native cloud computing enables software companies to build things that they can’t do on-premise, but if you have spent nearly 15 years describing ported on-premise code as cloud, don’t be surprised when your on-premise customers want whatever you may now have that is genuinely cloud native, on-premise.

This is Clayton Christensen’s innovator’s dilemma.

Speculating on AI

Building anything meaningful with AI is going to make the paradox more challenging. To build compelling enterprise AI solutions, one requires enterprise data to learn upon, ideally from many enterprises. Vendors that grew up in the cloud have built contractual and architectural models that facilitate data and meta date analysis. Few of those vendors have done much radical or disruptive innovation with this data yet. However, recent enhancements in fundamental AI technologies create a tantalising opportunity.

Whereas with cloud, the incumbent on-prem vendors initially fought against the trend (Larry at Oracle most acerbically), and then shifted gradually over time, with AI there has been a rush to embrace it.

SAP made several cloud acquisitions (I worked at SAP SuccessFactors from 2013-2018). These solutions now generate multiple billions for SAP and are still the mainstay of its native cloud revenue. Most of them are successful acquisitions when measured by shareholder returns.

What SAP failed to do was to bring technologies, methodologies and other learning from those businesses into the heart of SAP. Rather than bending and changing ERP to learn from its native SaaS acquisitions, it forced them to adapt to the ERP hegemony. This is why there are so few leaders from those companies still at SAP.

SAP’s acquisition of Signavio now seems astute, as it gives SAP a better window into data and process than it has to date. The cultural fit is good too. Perhaps it has learnt.

Back to DSAG

Over the years user groups have played an important role in course correcting SAP, but in this case I think DSAG are being a tad disingenuous. After all, their organizations (Siemens, VW, Bosch, Daimler etc) use Workday, Salesforce, ServiceNow, Google, AWS, Azure and so on, extensively and they seem perfectly happy to run significant mission critical business processes with those solutions in public clouds.

So they clearly understand the power of cloud software, But when SAP wants to the same, they into go grumpy press release mode. I call this ERPstalgie.

Enough now

Anyway, enough pontification. I like to think understanding incumbents gives me better insights into picking and guiding our investments into those start ups that augment or disrupt HR tech ecosystems. There are many things SAP taught me, and for that I remain grateful.

As is my usual form, I’ll finish with a song that has a tenuous link to the post.