5 Days of Oracle OpenWorld are over and it is time to look at the takeaways. Given the length of the conference, the number of briefings (Oracle had 38 meetings scheduled for me) and the number of announcements, I will change the Event Report format to my overall top 3 positives and share my top 3 concerns, starting with the technology side (next generation applications) and then hopefully in a few days for the HCM side (Future of Work), as I have already blogged my progress report on the HCM analyst meeting from last Sunday (read here).

Oracle’s vision - It became once more clear what Oracle’s vision is - an integrated technology stack engineered by Oracle, from storage beyond SaaS all the way to value added services like DaaS (if not familiar that's Data as a Service). Operated by Oracle for the customer - or if the customer wants to, operated by them on premises. Even though Oracle is now ‘all in’ on the cloud message - it was last year that Larry Ellison admitted he gave up and would now use the ‘buzzword’ - the company strictly supports the dual deployment capabilities. Interesting enough the deployments options are important for customers, in one of the Q&As Thomas Kurian shared that there were over 20 customer that have gone full circle - moving cloud to on premises and back or vice versa. It is also clear that Oracle sees cost competitiveness on the IaaS layer as critical, as various executives said that Oracle will match Amazon AWS or Google Cloud platform prices, whoever is cheaper. It is clear that Oracle tries to commoditize the IaaS tier as much as possible, making up for revenue and profitability on the PaaS and SaaS side. Certainly an attractive strategy for Oracle.

Computing Eras (from Mendelsohn's presentation)

12c comes along - It is over 2 years now that Oracle has announced 12c. We talked to some customers and they attest it is working and they are seeing benefits. If these TCO benefits were similar or in the range of what Oracle talked about at OpenWorld 2012 remains to be seen. The real scale test for 12c is anyway Oracle’s usage for DBaaS and moving its SaaS apps to it. Andy Mendelsohn multiple times mentioned that Oracle will be moving the Taleo Talent Management products to 12c first - which should make an interesting showcase. But 12c had to be ready also to enable Larry Ellison’s ‘2 click keynote’ of moving both a database and an applications from on premises to the cloud. And it looks like that works - even though Ellison joked ‘it could break anytime as it is live software’. But it was more than the 2 clicks announced in Ellison’s Sunday keynote - nonetheless Oracle has created a huge value proposition for customers - to move on premise applications to the (Oracle) cloud. For customers tired of maintaining their older apps (and paying their DBAs) - there is now (or soon) a viable alternative to run these apps. Ellison was (rightfully) proud that Oracle lived up to its commitment to move customers along, as the company has through all recent technology changes.

Oracle's in memory approach (from Mendelsohn's presentation)

The in memory features of 12c seem to be working well, too. Mendelsohn gave a few database 101 lectures on the pros and cons of row vs column storage and how Oracle enables both - on disk and in memory flavor - at the discretion of the customer. And Oracle was not tired to point out that to take advantage of in memory applications did not have to be re-written (positioning vs SAP HANA) and could write to memory, too (positioning vs Microsoft SQL server in memory option). Surprisingly there were few customer success stories and use cases - but they may not have been surfaced to me. But in general an indication that Oracle customers are conservative to move to new database releases - I heard a number of customers saying that the never move to R1 - but wait for R2. 12cR2 is coming soon so it will be time to check in on the uptake of both in memory and pluggable databases / multi-tenancy. But even if customers may be slow to uptake these features, they are core to power Oracle’s ‘as a Service’ business. So one way or the other the 12c features will get a lot of usage soon.

Database (not Application!) Multitenancy (from Mendelsohn's presentation)

The platform - If you attended and read about OpenWorld and did not hear about platform as a service (PaaS) message something has gone badly missing. In keynotes and sessions Oracle speakers would not get tired of stressing how Oracle is shipping a PaaS - but also uses and makes the same available for customers to do work on higher in the stack products, like SaaS and DaaS. The demos Ellison showed in his 2nd keynote like e.g. creating an employee of the month application for Fusion HCM showed that in action (though much was prepared beforehand). The work horse in the platform is Fusion Middleware, less with its traditional middleware features around SOA and ESB, but more on its composition, mobile and integration capabilities. And there is some merit to the argument, in the past Oracle Applications would certainly use the Oracle Database - but the uptake of Oracle tools and later Fusion Middleware were another story. The challenge was Oracle Applications was always only one customer of the technology stack - an important one certainly - but Oracle platform products usually had to run on faster cycles than the applications could. Enters the SaaS world with multiple releases per year and all of a sudden release - and with that uptake cycles between enabling platform technology and applications consuming those can be synched. It remains a heck of a prioritization effort – balancing the Oracle internal Applications requirements vs best of breed market requirements – but Oracle is deep pocketed enough to fund probably most of both. The vision is certainly remarkable and it’s good to see that e.g. where we have the insight – with Oracle’s cloud HCM products – the work has already begun to uptake, expose, elevate to the Oracle PaaS platform.

Oracle PaaS Portfolio (from Kurian's presentation)

Engineered Top Down? – My impression of the Oracle endeavor has been that it is heavily top down oriented, starting with the Applications. Quick reminder – Fusion started around 2004. So Oracle has been working on SaaS applications much longer than e.g. PaaS / DBaaS / DaaS or IaaS. All the former are lower level components of the tech stack. In the ideal world engineers want to start bottom up normally. When I asked Thomas Kurian about this, his view was that the SaaS applications have been built on these services all the time, citing Java as the example. And that is certainly a valid point – but it would be phenomenal if Oracle would not have to do some kind of re-work to make its SaaS applications run on all the just into live coming technology and platform products.

Oracle SaaS Portfolio (from Kurian's presentation)

Massive Task – The creation of the next Oracle technology effort is a massive effort with a five digit number of engineers working on it. To make sure all these project run well together and synch up on time is a massive task. When I asked Kurian was fair enough to say the latest (and probably newest) effort – IaaS – is what is giving him the most gray hair. So customer need to watch for product maturity and keep an eye on quality, not that they would not anyway – but a more complex (and powerful) technology stack needs appropriate attention. To put it into perspective – only the combined R&D efforts of IBM are in similar scale – and IBM does not claim (and does not need to) for all of it to work together. Oracle has accepted a much tighter locking of its tech stack components. That is more work and risk to create – but has benefits on the upside for customers when it all works together. 15 or so years ago Oracle had a similar vision - the database and the combination of Forms and Reports would run its Applications. Well that never took off back then (separate blog post sometime) - and it's a different age. Both capabilities and challenges are a magnitude 2 2 larger than back then. So Oracle needs to get it done.

Oracle Cloud Scale as of June 2014 (from Kurian's presentation)

Adoption – When even a vendor builds something new – it not only needs to get it done technically with the right quality – but also get adoption in the customer base. And that happens through sales people and partners. It was good to see that Oracle has acknowledged that its salesforce was one of the first sceptics initially, but now the vendor has put in the incentives to get the new products sold. At the same time Oracle is looking at partner enablement, not only on the services side, but also on the product side. Kurian shared that Oracle has already had conversation with over 60 ISVs. Nonetheless Oracle needs to look at this area and produce the numbers in customers and revenue dollars. When I asked Mark Hurd on this he shared that Oracle doubled the applications sales force and will be on top of the challenge. Glad to hear it’s realized – but it needs to be tackled and addressed.


For the longest time I have been critical of how Oracle may be able to match the prices of Amazon and / or Google. But a back of a napkin calculation showed me how: If Oracle can only convert 20% of existing load that it run on premise with DB and Java in the next 5 years – then it is larger than AWS and / or Google, easily. And with that comes purchasing and bargaining power with the suppliers Oracle needs to deal with in its cloud roll out. And we all know cloud is a scale game.

The next area to watch is how well integrated vs. consumable the Oracle stack will be at the end of this exercise. There will be customers who will want to consume it all – and then there will be customers who only want pieces of that technology stack – lock in fear and other reasons will apply. Oracle uses standards – as far as available and applicable at these intersection – but if this will allow Oracle to capture both sides of the business, remains to be seen.

Overall things are coming together for Oracle in 2014 – which was by OpenWorld 2013 blog post header (see here) – a number of announced products are being made available in the next few weeks, more in winter, then spring and summer 2015. OpenWorld 2015 will have in place all that Oracle announced at this OpenWorld – in the meantime there needs to be a lot of precision flying – for customers, partners and Oracle. Mistakes will be costly on either side. Stay tuned.

2012, 2013 & 2014 (C) Holger Mueller - All Rights Reserved