The Hyperscaler/GSI Relationship Has Forever Changed — Here’s How
By virtue of ever-increasing spending by enterprise customers, the cloud hyperscalers have been elevated to strategic partners of the first rank. This has forced many changes for enterprises and their global systems integrator (GSI) partners. Hyperscalers not only have earned a seat at the table but have gained significant influence on the behavior of their GSI partners.
How so? Increasingly, the GSIs are serving two masters: their end customers and their hyperscaler partners, which they rely upon for go-to-market opportunities. Large GSIs including Accenture, Atos, HCL, Infosys, and TCS among others have established hyperscaler-specific operating units, or at least senior executives that are focused on hyperscaler-specific opportunities. This split loyalty has resulted in a sometimes delicate balancing act for them in maintaining user advocacy in cloud strategies while at the same time helping to further the hyperscalers’ ambitions. This could result in a misalignment of motivations, given that hyperscalers are focused primarily on migration in order to achieve immediate consumption, while GSIs often prefer to get higher margins from modernization activities. Moreover, in contemplating specific customer requirements, GSIs are usually correct in saying “Let’s take a step back” first, instead of moving full speed ahead with lift-and-shift migration.
Migration-oriented technology makes hyperscalers’ motivations clear: Migrate enterprise apps quickly and with as minimal changes as possible (urging companies to migrate first and modernize later using their own services). Examples include Google’s acquisition of Velostrata in 2018, AWS’s acquisition of CloudEndure in 2019, followed by Google’s acquisition of StratoZone and Cornerstone Technology in 2020. More recently, AWS acquired Blu Age in 2021 to accelerate mainframe migration opportunities directly. However, it is also clear that the hyperscalers don’t want to turn away business that comes in through GSIs. For example, around the same time as the Blu Age acquisition, AWS also established a mainframe migration program involving several leading GSIs.
Another example of co-opetition lies in managed services. For example, AWS initially introduced its AWS Managed Services in 2016, previously known as Sentinel, and announced its catalog-based Operations On Demand program last September. Yet it continues to recognize and work with a list of more than 160 managed service provider partners.
Relationships between GSIs and hyperscalers are further redefined by the growing focus on vertical industry frameworks by the hyperscalers. While this does leave sufficient room for services suppliers to extend their deep vertical industry expertise through professional advisory and implementation services, it does delineate a new relationship between them with the potential for some disintermediation of what used to be exclusively the GSIs’ domain.
Will hyperscalers ultimately embrace service delivery to the point where they actually acquire sizable services companies? History shows that product companies have learned that the strategic conflicts inherent in such an acquisition may be too much to bear, not to mention the radical impact on margins and profitability of the hyperscaler business model, and therefore its ability to foster massive, continual ongoing investment, that such an acquisition would be likely to bring.
What is an enterprise customer to do? Acknowledge the new reality and the competing pressures on your GSI partners. Above all, make sure you keep your goals in alignment between the parties. Also, try to validate that your GSI’s hyperscaler-specific GTM units are more than glorified partnering organizations but extend into joint engineering and solution development. Many GSIs rely on their vertical GTM to transcend any potential conflicts, but make sure those new hyperscaler-specific GTM units do not merely reinforce siloed delivery.
Want to learn more? Reach out to email@example.com to schedule an inquiry with me.