Late on Friday SAP issued a press release about their acquisition of privately held SmartOps. The acquisition is minor and supposed to conclude this calendar quarter.
![]()
The company was founded by Sridhar Tayur, who is a professor of operations management at Carnegie Mellon, and he was the CEO of SmartOps till last year. SmartOps algorithms are supposed to strengthen SAP's Demand Signalling and Advanced Planning & Optimization (APO) products.
Enters SAP Hana, the 'little girl' that SAP - rightfully - is very proud of. When loading all the signals available into Hana - new demand patterns maybe recognized and put into execution plans. And SAP is looking hard for finding uses of SAP Hana in the business application space - which is SAP's core competence.
Despite the pioneering efforts, that functionality will now be largely shelved and re-platformed with the help of SmartOps and Hana. This will not only give thoughts to some engineers, but as well the (my)SAP SCM and APO customer community.
But this is an industry wide problem - as everyone relied and relies on iLog algorithms. And SmartOps only addresses a portion of the algorithm sourcing problem for SCM. Look for SAP to make more acquisitions in this space, as SAP does not have the internal expertise and competence to develop powerful algorithms.
Like with any acquisition - retaining the SmartOps talent will be a key challenge for SAP. But it also uncovers the deeper challenge SAP has going forward. While the company successfully mastered navigating down on the technology stack and providing an innovative in memory database product, it has not been able to move up in the direction of powerful algorithms, or in another buzzword - the true Analytics space. This poses an overall challenge for SAP - ironically in an area where a dozen years ago SAP held both a technical and analytical leadership position. But investment priorities in other areas of the portfolio have both lead to an understaffed and potentially under talented SAP SCM product development team.
Finally great algorithms lead to great analytics, but great analytics need the tie in into execution. And while SAP claims that the Business Suite runs on Hana - very few details on the technicalities have come out so far. The new offering needs to be tied into the operative SAP SCM installations - and if these will be on Hana by the time SAP wants to sell the new SAP Smartops based Demand Sensing rests to be seen. It certainly provides a compelling argument for moving a Business Suite installation to Hana. So SAP will have to choose wether they want to wait for their customers to move their transactional SCM products to Hana - or live with a hybrid approach of a new Hana offering (in memory) and transaction products (on spinning rost a.k.a. HDD). This will be overall an interesting situation to watch as SAP tries to move customers to Hana as the underlying database platform.
An acquisition from the ecosystem
SmartOps has been a key partner for SAP in the supply chain space since 2006, and their customer list reads like a who-is-who of SAP's North American customers.The company was founded by Sridhar Tayur, who is a professor of operations management at Carnegie Mellon, and he was the CEO of SmartOps till last year. SmartOps algorithms are supposed to strengthen SAP's Demand Signalling and Advanced Planning & Optimization (APO) products.
Demand Sensing meets In Memory
SmartOps started out with a more traditional aspect of Supply Chain Management (SCM) - Inventory Optimization, but recently added demand sensing, a technique developed in the more volatile CPG markets, aiming at better predicting product demand. Traditionally this was completely done by using POS data, but with the rise of the internet, social networks, coupon networks, loyalty programs etc demand sensing became more and more a BigData problem. And like many of the best practices out of the CPG space - demand sensing is making inroads in other industries, too. One of the challenges of demand sensing is, that data is available to all market participants (often for a fee) - so speed of analysis and execution will determine the winner between two competitors. So in memory technology looks attractive.Enters SAP Hana, the 'little girl' that SAP - rightfully - is very proud of. When loading all the signals available into Hana - new demand patterns maybe recognized and put into execution plans. And SAP is looking hard for finding uses of SAP Hana in the business application space - which is SAP's core competence.
The SAP SCM Irony
The SAP SCM development team, originally lead by former board member Claus Heinrich, was the first (and pre Hana only) development team to embrace in memory technology for its APO code. Originally developed internally (yes there was a in memory offering at SAP before Hana), the functionality benefited greatly from SAP's acquisition of Sofware AG's adabas database (now SAP DB) and became the SAP liveCache functionality in SAP SCM.Despite the pioneering efforts, that functionality will now be largely shelved and re-platformed with the help of SmartOps and Hana. This will not only give thoughts to some engineers, but as well the (my)SAP SCM and APO customer community.
It's the algorithms, stupid!
The other relevation about the acquisition is, that while the raw speed of in-memory is the enabler - the 'raw iron' to borrow from a SAP competitor, it's the intelligent software on top that matters, more specifically the algorithms to optimize the supply chain. In the early SCM times of the 90ies that company used to be iLog and everybody uses and used their algorithms. But then IBM acquired iLog and though the sharing of the algorithms is secured by (very) long term contracts - SAP may have grown weary of the co-opetition with IBM. With IBM focusing on analytics and being a database vendor - in memory is not far away. Or maybe IBM is working on a Watson for Supply Chain already?But this is an industry wide problem - as everyone relied and relies on iLog algorithms. And SmartOps only addresses a portion of the algorithm sourcing problem for SCM. Look for SAP to make more acquisitions in this space, as SAP does not have the internal expertise and competence to develop powerful algorithms.
Smart Ops - Opportunity or Challenge?
As with any acquisition out of a vendors partner ecosystem, the question is: Why now? And while above is a technology view, it may well be, that there are business aspects, too. SmartOps may have gotten too strategic for some large SAP customers, and these may have pushed SAP to start the acquisition. Where it may have rang a bell at SAP. Or SmartOps was planning this in the longer term, moving away the founder and CEO and putting an executive in place with the stamina (he has finished the famous Vasa Cross Country Ski race) to pitch the company to SAP.Like with any acquisition - retaining the SmartOps talent will be a key challenge for SAP. But it also uncovers the deeper challenge SAP has going forward. While the company successfully mastered navigating down on the technology stack and providing an innovative in memory database product, it has not been able to move up in the direction of powerful algorithms, or in another buzzword - the true Analytics space. This poses an overall challenge for SAP - ironically in an area where a dozen years ago SAP held both a technical and analytical leadership position. But investment priorities in other areas of the portfolio have both lead to an understaffed and potentially under talented SAP SCM product development team.
Finally great algorithms lead to great analytics, but great analytics need the tie in into execution. And while SAP claims that the Business Suite runs on Hana - very few details on the technicalities have come out so far. The new offering needs to be tied into the operative SAP SCM installations - and if these will be on Hana by the time SAP wants to sell the new SAP Smartops based Demand Sensing rests to be seen. It certainly provides a compelling argument for moving a Business Suite installation to Hana. So SAP will have to choose wether they want to wait for their customers to move their transactional SCM products to Hana - or live with a hybrid approach of a new Hana offering (in memory) and transaction products (on spinning rost a.k.a. HDD). This will be overall an interesting situation to watch as SAP tries to move customers to Hana as the underlying database platform.