Few at SAP are bet­ter-qual­i­fied to speak to the evo­lu­tion of the company’s cloud ERP port­fo­lio than Jan Gilg, Pres­i­dent and Chief Prod­uct Offi­cer, Cloud ERP.

In his 17-year tenure at SAP, Gilg has served as Glob­al Head of Enter­prise Archi­tec­ture, Glob­al Head of IT Busi­ness Ser­vices, and SVP & Head of SAP S/4HANA. That last posi­tion, which Gilg start­ed at the begin­ning of 2019, expand­ed a year lat­er to encom­pass SAP’s dig­i­tal sup­ply chain port­fo­lio, lead­ing to his cur­rent role, in which he’s respon­si­ble for dig­i­tal sup­ply chain and SAP’s cloud ERP.

As adop­tion of cloud ERP solu­tions con­tin­ues to accel­er­ate, with the glob­al mar­ket expect­ed to near­ly dou­ble (from $64.7 bil­lion in 2022 to $130 bil­lion by 2027) amid the rise of dig­i­tal trans­for­ma­tion, remote work, and stan­dard­iza­tions for busi­ness, SAP is sim­i­lar­ly dou­bling down on cloud innovation. 

Through­out 2023, the com­pa­ny embraced gen­er­a­tive arti­fi­cial intel­li­gence (AI) and con­tin­ued to ush­er on-premis­es cus­tomers toward the cloud, intro­duc­ing the 2023 release of SAP S/4HANA and ​“pre­mi­um plus” pack­ages for RISE with SAP. Now, as the SAP com­mu­ni­ty enters 2024, cus­tomers con­tin­ue to weigh the future of their inno­va­tion paths, includ­ing through ERP sys­tems such as SAP S/4HANA Cloud, pri­vate edi­tion, and SAP S/4HANA Cloud, pub­lic edition.

In a recent inter­view, Gilg dis­cussed the SAP cus­tomer base’s shift­ing atti­tudes toward SAP S/4HANA Cloud migra­tion, inno­va­tion paths for on-premis­es and cloud cus­tomers, and the impact of AI on enter­prise data. Below, in the sec­ond half of our wide-rang­ing con­ver­sa­tion, Gilg detailed the abil­i­ty of cloud ERP to enhance busi­ness col­lab­o­ra­tion, the cen­tral­i­ty of data qual­i­ty to suc­cess­ful­ly lever­ag­ing gen­er­a­tive AI, and much more.

This inter­view has been edit­ed and condensed. 

ASUG: Cloud-based ERP sys­tems are gen­er­al­ly under­stood to be more col­lab­o­ra­tive, net­worked sys­tems for busi­ness­es, but I’m curi­ous what your expe­ri­ence of this has been with SAP cus­tomers this past year. What ben­e­fits are busi­ness­es unlock­ing from cloud ERP? 

Jan Gilg: In gen­er­al, the term ​“cloud ERP,” and the real­iza­tion that it is quite dif­fer­ent in nature from the old­er on-premis­es world that we are leav­ing behind, is real­ly sink­ing in. We see that from adop­tion num­bers for our offer­ings in the pub­lic and pri­vate cloud, where col­lab­o­ra­tion is a key topic. 

On-premis­es solu­tions have tra­di­tion­al­ly focused on the four walls of an enter­prise — as the term ​“enter­prise resource plan­ning” sug­gests — and that’s why all of us at SAP have been say­ing that we have to widen the lens and talk more about the ecosys­tem in which every cus­tomer oper­ates. I think of col­lab­o­ra­tion from two angles. It’s about peo­ple work­ing amongst each oth­er, in a com­pa­ny and with part­ners, a peo­ple-to-peo­ple col­lab­o­ra­tion. And then it’s about com­pa­ny-to-com­pa­ny col­lab­o­ra­tion, through a busi­ness network. 

I’ll admit — the term ​“net­work” sounds like such tech­ni­cal ter­mi­nol­o­gy. That’s why I like to talk more about col­lab­o­ra­tion. Net­works sug­gest cables and wires. We start from a human-to-human per­spec­tive. What we’ve seen is that peo­ple increas­ing­ly work in these pro­duc­tiv­i­ty envi­ron­ments: Microsoft Teams, Zoom, Google Work­space. Com­ing out of COVID, that’s accel­er­at­ed sig­nif­i­cant­ly. We still see those hybrid envi­ron­ments today, and I don’t think they will ever go away. SAP will invest more in bring­ing those envi­ron­ments togeth­er and mak­ing them more seamless. 

Wher­ev­er peo­ple work, even those who only work in Microsoft Teams or Out­look, they can access data out of an SAP S/4HANA sys­tem or cloud ERP sys­tem. They can even trig­ger cer­tain actions in such a sys­tem, often with­out even know­ing that they are using cloud ERP or an SAP S/4HANA Cloud solu­tion. That’s crit­i­cal to dri­ving the num­ber of users because, in the past, ERP has always been looked at as an expert solu­tion, where you have to have exten­sive train­ing to com­plete crit­i­cal tasks, and not many peo­ple have that train­ing. I see that chang­ing; we’ve invest­ed a lot with Microsoft and Google on the engi­neer­ing lev­el to that end.

With the rise of gen­er­a­tive AI, there’s also been a huge push to rethink the role of the dig­i­tal assis­tant. On one hand, the dig­i­tal assis­tant can help you work with­in SAP sys­tems and pro­vide a dif­fer­ent user expe­ri­ence; on the oth­er hand, it can bet­ter enable you to col­lab­o­rate with oth­ers. Its chat and shar­ing capa­bil­i­ties are front and cen­ter in cloud ERP. Extend­ing that beyond the bound­aries of a com­pa­ny, through a net­work, we want to empow­er peo­ple to inter­act, to share doc­u­ments, to send pur­chase orders and sales orders, invoic­es, and deliv­ery doc­u­ments, and so on. We will con­tin­ue to enable more of those sce­nar­ios in the busi­ness net­work. The foun­da­tion for that is cus­tomers’ will­ing­ness to share data, which I have seen increase in this last year. 

Of course, there are ques­tions that must be answered: Is my data secure? What about data pri­va­cy? That’s why we’ve said, right from the begin­ning, in the con­text of AI, that secu­ri­ty and pri­va­cy are very impor­tant to us. But I believe com­pa­nies are much more open to shar­ing data with each oth­er because they see the ben­e­fits; one exam­ple is the auto­mo­tive indus­try net­work Catena‑X. Con­sid­er how expen­sive recalls are for auto­mo­tive com­pa­nies. They often have to recall so many cars, because they don’t know exact­ly which part in which car is defec­tive, because they don’t have this infor­ma­tion, because sup­pli­ers don’t share it with them. If they do share that infor­ma­tion, every­body ben­e­fits. The under­ly­ing foun­da­tion, as always, is data and cus­tomers’ will­ing­ness to share it. 

ASUG: Every intel­li­gent enter­prise depends on data to sus­tain its oper­a­tions, and the impor­tance of clean, well-gov­erned data to cus­tomers ratio­nal­iz­ing data com­plex­i­ties and explor­ing AI can’t be over­stat­ed. How can cloud ERP improve data qual­i­ty, gov­er­nance, and usabil­i­ty for SAP customers?

Gilg: Over the past few years, the nature of an ERP solu­tion in the cloud has evolved from a data-in sys­tem of record toward a data-out sys­tem of record. Assum­ing that the input data is already avail­able dig­i­tal­ly, this sys­tem has a much greater abil­i­ty to crunch data in the moment and play it back to end users, to pull them in if there’s an excep­tion. In the past, that idea has fall­en flat, because cer­tain tech­nolo­gies were not in place, but that’s changed since we based our ERP sys­tem on an in-mem­o­ry data­base. It’s also based on data qual­i­ty, which a lot of com­pa­nies stum­ble over. 

Cus­tomers are eval­u­at­ing brown­field ver­sus green­field approach­es for their ERP imple­men­ta­tion. Green­field is a fresh start; it’s expen­sive and some­times painful, but it pro­vides the biggest lever for progress. Data is always one of the biggest com­po­nents and work­streams to con­sid­er in that dis­cus­sion. You are absolute­ly right: if you don’t have clean data, the algo­rithms on top are not going to help you. SAP has invest­ed in mas­sive data-cleans­ing use cas­es, lever­ag­ing AI to reduce tedious man­u­al work for our cus­tomers. In SAP Inte­grat­ed Busi­ness Plan­ning, we have a mas­ter data appli­ca­tion that looks at data and then pro­pos­es rules as to how this data should be cleansed. The user basi­cal­ly just looks at those rules, eval­u­ates the prob­a­bil­i­ty that the data is cor­rect, and then accepts it, and the rule is applied to the data set, then cleans­ing the data. This gives you a foun­da­tion on which you can run your AI. 

As you can imag­ine, here at SAP, we are heav­i­ly research­ing and invest­ing into how to run gen­er­a­tive AI on top of tab­u­lar data, which is still a nut to be cracked. Today, we see the impor­tance of using gen­er­a­tive AI to address unstruc­tured data, text, lan­guage, video, and pic­tures. Tab­u­lar data is at the core of every ERP solu­tion, and we have some very promis­ing AI use cas­es to that end. Data qual­i­ty is a pre­req­ui­site. We need to help cus­tomers achieve clean data in two ways; it’s part of our RISE method­ol­o­gy for on-premis­es cus­tomers who move into the pri­vate cloud, and cus­tomers in the pub­lic cloud take a green­field approach and start fresh. Clean data is a crit­i­cal suc­cess fac­tor for any AI to pro­duce rel­e­vant and reli­able results. 

ASUG: What are the most crit­i­cal fac­tors for cus­tomers in the midst of build­ing busi­ness cas­es for cloud migra­tion? What moti­vates busi­ness lead­ers to move for­ward with these types of transformations?

Gilg: For exec­u­tives, inno­va­tion con­sump­tion is the most impor­tant top­ic. The last few years have proven, in the line of busi­ness, how a SaaS mod­el brings ben­e­fit to less crit­i­cal process­es, like indi­rect pro­cure­ment, HR, and CRM, but busi­ness lead­ers see the poten­tial to lever­age this mod­el for more crit­i­cal core process­es, and they like a mod­el where they can con­sume inno­va­tion with­out hav­ing to do much for it, so to speak: no big upgrades, no big test­ing efforts, and so on. 

Relat­ed to that, there’s a pen­du­lum that swings back and forth over the decades, and it’s cur­rent­ly swing­ing toward scru­ti­niz­ing every process, to deter­mine if it’s a dif­fer­en­ti­at­ing process or not. In the case that it’s not, why not fit it to stan­dard and — rely­ing on a ven­dor such as SAP, which has been in the busi­ness for 50 years — ask for what­ev­er every­body else is using for cer­tain process­es? You can use that as a forc­ing func­tion, inside your busi­ness, to remove unnec­es­sary approval steps or extra built-in func­tion­al­i­ties. With that rad­i­cal stan­dard­iza­tion, you can dri­ve oper­a­tional effi­cien­cy in your orga­ni­za­tion and also focus on invest­ing in areas where dif­fer­en­ti­a­tion does exist.

One of our cus­tomers, for exam­ple, will run more than 50% of its rev­enue with SAP S/4HANA Cloud, pub­lic edi­tion. That com­pa­ny used the appli­ca­tion as a forc­ing func­tion for their busi­ness trans­for­ma­tion. Respond­ing to what the appli­ca­tion does, the cus­tomer adjust­ed its process­es. This required change, and for peo­ple to get used to doing things dif­fer­ent­ly around process­es that were not gen­er­at­ing val­ue for the com­pa­ny; this allowed them to focus on process­es that do gen­er­ate val­ue, invest­ing in exten­sions and extend­ed capa­bil­i­ties. Total cost of own­er­ship (TCO), of course, plays a role when you look into cloud migra­tion, but it’s not always the main driver.

ASUG: What chal­lenges or risks do you see cus­tomers encounter in mov­ing away from SAP ECC or on-premis­es invest­ments toward the cloud? How do you see SAP help­ing cus­tomers man­age any risks asso­ci­at­ed with that transition?

Gilg: It’s a fun­da­men­tal­ly dif­fer­ent oper­at­ing mod­el, and at the heart of what you’re talk­ing about is trust. This is where I get the most push­back from very large, tra­di­tion­al com­pa­nies that say, ​“Hey, those are our most crit­i­cal process­es. I’m the CIO. I want to be in con­trol of man­ag­ing that. And now I’m asked to give con­trol of that to SAP.” 

SAP must demon­strate how we’ll do that and demon­strate the resilience that we built into the plat­form to oper­ate and believe we can run this much bet­ter at scale than an indi­vid­ual cus­tomer can run their sys­tems, espe­cial­ly as we con­tin­ue to invest in non-func­tion­al top­ics like secu­ri­ty and resilience. Some of our large cus­tomers have thou­sands of peo­ple in their SAP prac­tices, because they have inde­pen­dent­ly devel­oped an oper­at­ing sys­tem that works well for them, at a very high cost. Tech­nol­o­gy is evolv­ing so fast, and cyber­se­cu­ri­ty is chang­ing as well, so they will have to keep up with that. For most com­pa­nies, that’s a sig­nif­i­cant bur­den we can take away, while giv­ing them trans­paren­cy and insight through our mon­i­tor­ing tools. 

Every­thing we can do to give cus­tomers trust and con­fi­dence that we can run their most crit­i­cal work­loads — as well as our SLAs and the secu­ri­ty stan­dards we adhere to — should give them trust in our cloud envi­ron­ment. I see this as a trust jour­ney, pri­mar­i­ly, even more than a tech­nol­o­gy jour­ney, espe­cial­ly for very large com­pa­nies. Small­er com­pa­nies want to offload that work and avoid the need for large IT depart­ments. Large multi­na­tion­al com­pa­nies, how­ev­er, see it as a major step. 

That said, those com­pa­nies expe­ri­ence break­age in the expe­ri­ences they pro­vide for their cus­tomers, and those issues are often because of sys­tem breaks or cer­tain end-data flows, where data doesn’t match and fit togeth­er. That’s always evi­dent to the end cus­tomer, and all com­pa­nies are inter­est­ed in fix­ing that. I believe that you can do that only in the cloud, with a ful­ly cloud-enabled sys­tem. It’s a jour­ney, and it takes time, but you have to get start­ed. The dis­rup­tion that’s hap­pened these last cou­ple of years has cre­at­ed even more urgency for com­pa­nies to get start­ed, lest they fall behind. 

ASUG: How will SAP address the secu­ri­ty con­cerns of cus­tomers mov­ing from on-premis­es sys­tems to the cloud? How can orga­ni­za­tions bal­ance the increased risk of expo­sure to cyber-attacks? 

Gilg: One lev­el of this con­cerns phys­i­cal access to sys­tems and data, which is high­ly reg­u­lat­ed by SAP. It’s not pos­si­ble for some­body to walk into a data cen­ter, gain access to a machine, and con­nect a hard dri­ve to it. Any phys­i­cal access is heav­i­ly restrict­ed. On the devel­op­ment side, we build secu­ri­ty into the process. We have high inter­nal secu­ri­ty rat­ings at SAP, which are scanned every time before any­thing reach­es pro­duc­tion. You have to be inter­nal­ly cer­ti­fied on that, and you can’t ship soft­ware unless you receive the high­est rating.

We also have all the capa­bil­i­ties in our cloud oper­a­tions, while the sys­tem is run­ning, to shield it from cyber-attacks and pen­e­tra­tion. We have reg­u­lar audits in our oper­a­tions and in our devel­op­ment process­es. SAP is spend­ing an extra­or­di­nary amount of mon­ey on the top­ic of secu­ri­ty, data pro­tec­tion, and privacy. 

In Ger­many and Europe, this is espe­cial­ly a big deal with var­i­ous rules, such as Gen­er­al Data Pro­tec­tion Reg­u­la­tion (GDPR). Some cus­tomers don’t want any­body in sup­port to access the sys­tem out­side of the Euro­pean Union, which we can also grant if required. Around data sov­er­eign­ty, we have sov­er­eign clouds in North Amer­i­ca and in Germany.

There are var­i­ous mea­sures we offer, in the con­text of secu­ri­ty, data pro­tec­tion, and pri­va­cy, to make sure that each of the cus­tomers’ require­ments are met. It’s hard for an indi­vid­ual cor­po­ra­tion to mir­ror that. Nobody would ever want to admit their own vul­ner­a­bil­i­ties, but I have to assume many on-premis­es deploy­ments have vul­ner­a­bil­i­ties that SAP can address.

For more from Jan Gilg, read the first half of our ASUG inter­view, dis­cussing the SAP cus­tomer base’s shift­ing atti­tudes toward SAP S/4HANA Cloud migra­tion, inno­va­tion paths for on-premis­es and cloud cus­tomers, and the impact of AI on enter­prise data.