By Holger Mueller
Enterprise Technology | Cloud Platforms
Not all infrastructure platforms are the same and choosing your cloud partner is a critical strategic decision. Holger Mueller, contributing editor at ERP Today and principal analyst at Constellation Research looks at the options for enterprise work loads in the public cloud.
We are on the cusp of the third decade of the 21st century and it is clear that the computing paradigm has moved into the infinite computing era. This era is characterised by overcoming the limited resources of the on-premise compute model, delivered through traditional enterprise datacentres. New use cases needed by enterprises to enable digital business models and modern best practices require the realisation of infinite computing to power the next generation of applications.
And it is the public cloud vendors that provide these platforms, so it is worth taking a look where they stand.
But before that, let’s look at what CxOs look for when considering selecting public cloud infrastructure (IaaS) players:
Deployment Fit. IaaS players not only have to be able to run the new next generation application that enterprises need to build, but also be able to power existing loads – whether they come from standard software vendors (SaaS) or former custom build applications that enterprises need to operate.
Innovation Fit. IaaS players try to be everything to every enterprise but they have specific pros and cons for enterprises use cases and their product DNA matters for a good fit to an enterprise. We highlight these further on here.
Commercial Fit. Needless to say, enterprises need to pay for using IaaS and commercial terms need to be favourable, not only short term – but long term, ideally for the life cycle of enterprise load.
Enterprise Acceleration Potential. Enterprises need to move faster than ever before, so determining the potential of IaaS vendor to help an enterprise become more agile, move faster and at the end of the day acceleration is key.
So where are the IaaS vendors with their platforms? Let’s look at the top three and three contenders:
Amazon AWS – It’s good to be king, but there is a chink in the armour.
None of the players here has bet on the cloud earlier than Amazon. And AWS remains the most compelling business for Amazon as it provides higher and stickier profit than the consumer etailer business. This business dynamic works in favour of AWS, versus for all other vendors the IaaS business shows lower margins than their core business. AWS has built its offering like a menu in a Chinese restaurant – it is long, takes time to read – but probably everything imaginable that is edible is there. Select at your own risk, or rely on your waiter. This approach makes AWS the functional richest IaaS vendor out there. But enterprises need to make sure they can select the right pieces. AWS has been very good at signing up more ‘waiters’ in the last years, in the form of the typical system integrator with which enterprises do business with. The only weakness at AWS at the moment is that it is coming late to the artificial intelligence (AI) party, only unveiling its own AI platform in the last 12 months.
Microsoft Azure – It helps when you speak enterprise.
Microsoft’s original start of Azure was hampered by the Windows everywhere mantra, which current CEO Nadella overcame when heading Azure (moving to Linux). Microsoft had a considerable advantage over fellow leaders AWS and Goggle Cloud, due to its ‘organic’ load of Office that it successfully moved from on-premise to Azure. At the same time Microsoft has been a partner of IT for a long time and usually beats out the same two competitors when it comes to CIO access. A recent string of large-scale wins, also fuelled by some enterprises not willing to partner with AWS for other general go-to market reasons, has given Azure a lot of momentum.
Google Cloud – The third kid is fighting for its space.
Google came late to the IaaS game, and originally with a too ‘opinionated’ approach, i.e. expecting enterprises to build applications like Google does. That didn’t work well and Google re-geared to become more flexible and enterprise friendly. The Google Cloud game plan is firmly rooted into its leadership of making sense of Data, its foundational DNA. Google wants to be the data platform for the enterprise, that powers AI and brings machine learning to all users interested in the form of low / no code platforms. With new leadership under Kurian (ex-Oracle) and Enslin (ex SAP) its enterprise profile has gone up substantially, as well as the development of additional value propositions for enterprises.
Alibaba Cloud – Is there room for an APAC player at the global table?
Alibaba has been at the IaaS game for a while, though not with a global profile. Starting in China, then APAC and now Europe and Africa, Alibaba is taking a ‘surround’ strategy to the biggest cloud market, North America. Its ecommerce DNA makes it similar to Amazon and AWS. Concerns about the outcome of the US / China trade tussles make this a smart strategy in hindsight. Alibaba offers the combined DNAs of Amazon and Google with origins in China. It is a partner to consider when enterprises do business in China / APAC and enterprises need a local player.
IBM Cloud – The early lead has fizzled – can Red Hat fuel a comeback?
IBM has been a trusted partner for enterprises for more than 70 years and it was an early mover to the cloud, leading the ‘monopoly’ race of opening data centres until recently. IBM was also the first to partner with VMware, the leader for running on-premise loads. With its bare metal capabilities IBM is the only IaaS vendor to allow a CIO to move all its workload, no matter what kind of ‘zoo’ it is – to the cloud. But its focus and DNA on services have slowed IBM down, particularly compared to the three leaders. The IBM focus and DNA of (human) services has held back IBM’s early strong position for IaaS. IBM has a shot at gaining relevance thanks to the Red Hat acquisition, that gives it a chance to lead hybrid cloud scenarios. What the public cloud option for these will be is to be determined, but it is that less of the majority will go to IBM’s public cloud.
Oracle Cloud Infrastructure – Being the best cloud for your database is not enough.
Oracle has been late to the public cloud, with founder Ellison dismissing the trend to just another ‘fashion’ of the IT industry. But a few years ago that changed and Oracle is on its second try to get public cloud infrastructure right with OCI. Where Oracle has excelled is in running its database load on Exadata, giving customers the choice between on-premise and public cloud deployments. The recent partnership cross-connecting to Microsoft shows a new approach by Oracle. The key question is – how much of the Oracle SaaS and PaaS load – beyond the database – will OCI power in the next years?
The CxO Takeaway
What do CxOs need to consider when selecting public cloud platforms in 2019?
Craft your AI strategy. AI will be the most transformative technology of the next decade. Getting it right will define the winners and losers in the marketplace. The right partner matters tremendously.
What do my SaaS vendors do? Enterprises will use SaaS vendors who make decisions on which IaaS platforms to build on (see my previous ERP Today article in the July issue).
Data gravity matters. Network costs are real and data egress can be cost prohibitive. The choice of the right IaaS partner is key, as data will attract and create more data.
Avoid lock-In. Only two things are sure: Things will change, and the rate of change is going up. CxOs need to protect their enterprises from expensive lock-in on losing IaaS platforms by making sure their workloads remain as portable as possible. Think standards, virtualisation and even better containers as the first level of defence, Kubernetes as the second, and common standard (e.g. TensorFlow) as the third line.