Making Multi-Cloud Work

For our second annual State of Cloud Strategy Survey, we commissioned Forrester Consulting to expand our respondent universe beyond the HashiCorp contact database and add independent analysis. Even with this refined approach to the survey questions and sample methodology, the 2022 results clearly reiterate and expand upon the key trends revealed last year: the prevalence and success of a common multi-cloud operating model (a consistent framework for adopting cloud services), the ongoing importance of cloud security, and the impact of gaps in critical cloud skills.

This year’s survey sheds new light on how organizations are adopting centralized cloud operations — often, for example, in the form of platform teams driving automation as the path to success. But even as these developments drive good outcomes — 90% say multi-cloud is helping their organizations achieve their business goals — there’s more to be done. This year, we also saw that while organizations are boosting their cloud spending, more than 9 out of 10 are wasting some of their cloud investment. To cope, companies are turning to automated tools to enable central platform teams to leverage scarce cloud skills and enforce and manage a common cloud operating model.

Even though the methodology was slightly different this year, the results reinforce the central findings of our 2021 survey: multi-cloud — using more than one public or private cloud — is the new normal, chosen by more than 4 out of 5 respondents (60% are already multi-cloud, and another 21% plan to be multi-cloud within a year). In HashiCorp’s 2021 State of Cloud Strategy Survey, which polled HashiCorp’s opt-in database with slightly different questions, 76% said they were already multi-cloud, and 86% said they would be multi-cloud within two years.

Critically, multi-cloud is not a hypothetical blueprint without current practical benefits. An overwhelming 9 out of 10 respondents who have adopted a multi-cloud approach say it is already helping their organization advance or achieve its business goals.

That’s a big jump from our 2021 survey, when just 53% of respondents said multi-cloud helped achieve their business goals. Interestingly, the tech industry and large enterprises (15,000+ employees) trailed slightly on this metric, with “only” 87% saying multi-cloud is working for them. Smaller organizations (1,000 - 4,000 employees) were more positive: 92% said multi-cloud was delivering business results. Regionally, multi-cloud benefitting the business was more prevalent in Asia (93%) and Europe (91%), and slightly lower in North America (88%).

Digital transformation was an even bigger driver for multi-cloud adoption this year, cited by 43% of respondents, compared to 34% in our 2021 survey (based on a slightly different methodology). Yet while digital transformation was the top driver in last year’s survey, it came in second to reliability (46%) this year, and was closely trailed by scalability, security and governance, cost reduction, and a broad mix of other factors.

Cloud strategy and operations are so important that the vast majority of organizations have created a centralized organizational group or function such as a cloud platform team or Cloud Center of Excellence (CCoE) to make the most of their efforts. (Note: In this study, we refer to these centralized cloud functions as “platform teams.”)

One reason platform teams are so popular is that they perform a number of critical functions throughout the organization. It's still early days for platform team adoption, and organizations are experimenting with specifically what platform teams can and should do. So while standardizing cloud services, creating and sharing best practices and policies, and centralizing cloud security and compliance top the list, more than three-quarters of respondents rely on their platform teams for all the functions listed.

Given the essential functions involved, it makes sense that many companies that are not yet relying on a platform team are considering implementing one.

Just about everyone knows automation is critical for operationalizing multi-cloud — only 1 out of 100 of respondents thought it was not important — which makes it a key tool for platform teams. Just as telling, 89% thought automation was important or very important:

Popular reasons to embrace infrastructure automation include creating a faster, more flexible, and more reliable self-service IT infrastructure; improving security; better utilization of cloud resources; and faster incident response.

Critically, only about a quarter (27%) of respondents want to build their own tools from scratch — relying on open source or buying commercial tools (either as a service or as running it themselves) are significantly more common choices.

Why are automated solutions so popular? Because consistent, automated tooling can help organizations more efficiently address multi-cloud barriers around people, processes, and tools, including skills shortages, siloed teams, inconsistent workflows, lack of interoperability, and many others. As noted above, platform teams, automated tools, and a consistent cloud operating model help organizations address all of these challenges.

Automated tools also play key roles in many of the factors driving cloud success. Automated security tools are essential to implementing a zero trust security approach, for example, while infrastructure automation is critical to maximizing uptime and availability and providing visibility into cloud infrastructure. 81% of respondents cite the availability of automated tooling as an important or very important factor for their organization’s cloud strategy. 51% of tech firms rated tooling as very important, compared to just 42% of respondents in other industries.

In addition to security and tooling, which were already mentioned above, other keys to cloud success include uptime and availability (86%) and meeting regulatory requirements (82%). Tech firms placed more emphasis on uptime and availability (60% said it was “very important”) than did firms in other industries (46%). Similarly, scalability was rated “important” far more often by tech firms (56%) compared to other industries (37%).

Respondents said automated tools are already bringing significant benefits across a wide range of areas, with scalability (49%), infrastructure as code (42%), cost optimization (40%), and flexible deployments (39%) topping the list. When it comes to benefts respondents hope to gain in the next year, cost optimization (42%) overtakes scalabilty as the most popular choice, and developer productivity (37%), automated workflow-enabled operations (36%), and time to market (35%) also show up in the top five.

In particular, respondents confirmed results from last year’s survey that automated tooling offers significant benefits across four layers of the stack: infrastructure, networking, security, and application deployment.

When it comes to the success of a cloud strategy, nothing is more important than security, which was cited as important or very important by almost 9 out of 10 respondents. Related regulatory and compliance issues were third on the list, named by more than 4 out of 5 respondents, behind delivering uptime and availability. And security was even more important to tech firms, where two-thirds (67%) rated it “very important” (vs. 59% for other industries). For comparison, in our 2021 Survey, security was named the second most important cloud inhibitor (47%), trailing cost concerns (51%).

Data security is by far the biggest concern: The four most-cited external security threats include data theft, ransomware, phishing, and secrets leakage. Internally, data/privacy protection and threat detection and remediation ranked the highest. But the third-place choice, multi-cloud complexity, was ranked as the biggest challenge by 13% of respondents, more than any other choice. (For comparison, data/privacy protection, data theft, regulatory compliance, and software vulnerabilities were the top security concerns in our 2021 survey.)

To protect themselves, organizations are turning to security tools and automation. Data protection/encryption tools were most commonly named (88%) as important or very important to cloud success. In addition, access control and session management and secrets management were tied for third place with infrastructure as code and network infrastructure automation (83%).

Of course, automated tools are also critical beyond security, with tools for CI/CD (84%), infrastructure as code (83%), access controls and session management (83%) and network infrastructure (83%) also ranking high. Not surprisingly, perhaps, CI/CD was particularly valued in the tech sphere, with 61% calling these tools “very important” vs. 44% in other industries.

Finally, almost half (46%) of respondents said automated tooling has or would help improve security and governance (as noted in the Automation section above).

Overspending in the cloud isn’t just common, it’s ubiquitous. More than 9 in 10 respondents noted avoidable cloud spend, most commonly due to some combination of idle or underused resources (66%), overprovisioned resources (59%), and lack of needed skills (47%).

Fortunately, awareness of cloud waste is the first step to bringing it under control — as noted above, 80% of respondents consider “budget” an important or very important driver of cloud success. Notably, many companies already seem to be planning for this and still staying on target for overall cloud spending: less than a quarter (24%) of respondents say they overspent their cloud budget in 2021 (only 5% underspent).

Not surprisingly, the more budgeting and management responsibility respondents had, the more they saw cloud spending meeting expectations. For example, 66% of respondents with titles of director and above said their budget was in line with expectations, as did 69% of respondents who have sole responsibility for cloud strategy. Of course, that doesn’t explain why North American respondents were more likely to report overspending (27%).

In our 2021 survey, skills issues were a critical inhibitor to respondents’ cloud program. In 2022, skill shortages move to the top of the list. In addition, skills-related factors also featured prominently, with siloed teams, lack of training, manual operations, and budget-constrained headcounts also complicating the ability to operationalize multi-cloud.

Working across multiple clouds and hybrid environments, each with its own set of tools and workflows, puts pressure on busy teams to be experts on a vast array of technologies and processes. The top approach to solving this conundrum is to standardize on a common cloud operating model — a set of common tools and automated workflows chosen and enforced by platform teams leveraging the most skilled personnel. In addition to boosting training and working more strategically with partners and consultants, a cloud operating model administered by a platform team helps make the most of limited organizational resources.

Notably, one in five companies (22%) are still struggling to gather the skills needed to staff a platform team. To help, most respondents (59%) are standardizing on a cloud operating model (including shared workflows) to help fill skills gaps. Similarly, since platform engineers are force multipliers, it makes sense to prioritize staffing platform teams.