5 Enterprise IT Predictions for 2018

I’m a bit late to the VC 2018 predictions blogging party, but have finally put together my own list of enterprise tech forecasts for 2018 based on my conversations and observations from the prior year:

• Serverless computing adoption will accelerate, buoyed by dramatic improvements in performance and observability. To date, serverless infrastructure has been confined primarily to test workloads, non-mission critical microservices or architecting brand new applications, but has rarely been deployed wholesale to replace the infrastructure for an existing application. When it has been, complaints of unpredictable performance abound, leaving engineers skeptical that the promise of cost savings on compute is worth the risk of a laggy or faulty customer experience. However, AWS and other public cloud providers will improve rapidly (as they always have) and will build out the necessary tooling for monitoring and troubleshooting. Simultaneously, a burgeoning ecosystem of startups, such as Stackery, Honeycomb and Iopipe, has emerged to address observability and to ease the transition to serverless architectures.

• Enterprise spend on software-defined and open networking-related solutions will grow dramatically, even as cloud adoption continues to rise. I don’t believe anything can stop the transition to the public cloud in the long term, but a significant % of workloads remain on-premise or run on private cloud (according to this LogicMonitor survey, it’s 37% and 19%, respectively). In the interim, increased pressures to cut costs on existing data centers will push IT execs to adopt commodity hardware with SDN vendor-provided control planes. Additionally, public cloud providers such as AWS as well as managed data center leaders such as Equinix will continue to be buyers of next-generation datacenter equipment and software, even if in the long term the dollars spent here increasingly concentrate among a smaller handful of webscale infrastructure providers as opposed to individual enterprises. I predict that at least one startup in this space (such as Cumulus, Big Switch, Snaproute, Apstra will break out into late stage (>$50M) revenues this year.

• The Chief Data Officer role will increasingly be absorbed into other business units as organizations realize that all execs must be “data execs.” Having spoken to numerous de facto heads of data this past year, it’s clear to me that the perceived responsibilities that a CDO would take on are already distributed across a variety of organizations and titles, whether that’s the VP of Analytics, VP of Data Engineering, VP of Data Governance, VP of Privacy or even the CMO or VP of Marketing, who often oversee hairy data integration and customer 360 projects. Executive roles that are broadly charged with the “data” organization, just like prior years’ Chief Web Officers and Chief Digital Officers, will be absorbed into other business functions as all enterprises become data-centric organizations by necessity. Responsibility for data infrastructure, engineering, analytics and governance/privacy will increasingly fall to all existing C and VP-level executives. This will make it harder for vendors to identify the exact target customer, but will also broaden the # of people who could ostensibly be buyers.

• Predictive maintenance and monitoring solutions for supply chain and manufacturing will gain much broader adoption as pilots that started in 2016 and 2017 mature into commercial deals with new equipment release cycles, but the gains will accrue mostly to platforms run by OEM incumbents such as GE Predix. This has been a hot space for investment over the past few years, but sensor manufacturers will control both the data collection and the distribution / bundling of solutions with hardware. The real opportunity for startups will lies in developing highly vertical-specific solutions in contrast to the OEMs’ generalized platforms. Because the ideal underlying infrastructure for a given workload varies depending on the type/format of the data and the predictive model being used, there remain opportunities to own vertical specific stacks, especially for complex use cases such as genomics and drug testing.

• US-based enterprise IT startups will find large customers in China’s tech industry, and vice versa. As application user bases become increasingly global, IT infrastructures will have to become more globally distributed and performant as well. Several US-based GGV portfolio companies have had promising conversations with potential customers and distributors in China, while we see more and more execs from BAT (Baidu, Alibaba, Tencent) and other internet giants from China visiting Silicon Valley to explore best of breed products and showcase their own sophisticated infrastructures. On the flip side, Chinese infrastructure startups offer their own compelling and differentiated products. For example, Qiniu, which I had the pleasure to meet on my last trip to China, offers cloud hosting, storage and CDN services broadly, but has also bundled solutions to serve the backend needs of specific categories of popular apps, such as livestreaming and social networking. Similarly, UCloud, another cloud infrastructure company, has invested in being especially strong in supporting customers from the mobile gaming industry. For US-based application developers in these categories, partnering with one of these vendors could accelerate deployment and go to market. This is why I believe vendor adoption will flow between the US and China in both directions.

I look forward to revisiting this post at the end of 2018! I’d welcome your thoughts, feedback and/or vociferous disagreement!

#Enterprise #VentureCapital