DevOps
Blog
DevOps12 min

State of DevOps in Swiss Romande 2026: Trends, Data, and Field Observations

Matthieu Robin15 avril 2026

Since 2016, we have been helping Swiss Romande companies through their DevOps transformation. More than 100 organizations, from startups to public institutions, banks, hospitals, and software vendors. What I am sharing here is not an online survey with self-reported answers. It is what we observe concretely in the field, at our clients' sites, day after day.

I wanted to write this article because international reports (DORA, CNCF, Puppet) are excellent for global trends, but they do not always reflect the reality of the Swiss Romande economic fabric. Our SMEs have specific constraints: the nFADP (new Swiss Federal Act on Data Protection), the shortage of French-speaking DevOps talent, and a culture of caution that sometimes slows down adoption of new practices.

Here is what we see in 2026.

Kubernetes Adoption: A Progression in Waves

Globally, according to the CNCF Annual Survey 2024, 84% of organizations are using or evaluating Kubernetes in production. In Swiss Romande, we are below that number, but the progression is clear.

SMEs: Adoption Happens in Stages

Of our 100+ clients, roughly 45% use Kubernetes in production today, compared to 20% in 2022. Adoption rarely happens all at once. The typical pattern we observe: an SME starts by containerizing a non-critical application, tests it for 3 to 6 months, then progressively migrates its core services. The full cycle generally takes 12 to 18 months.

SMEs with 50 to 200 employees are the most active. Smaller ones (fewer than 20 people) still hesitate, often due to lack of internal resources. Larger organizations have generally already made the move or are mid-migration.

According to Datadog's State of Kubernetes 2024 report, the average number of containers per organization grew by 35% year over year. We observe a similar dynamic among our Swiss clients, with a doubling of managed pods between 2023 and 2025.

Public Sector: On-Premise Kubernetes

The Swiss Romande public sector is an interesting case. Sovereignty and compliance constraints push these organizations toward on-premise Kubernetes rather than cloud solutions. We have supported institutions like Fondation des Parkings in Geneva, which manages its Kubernetes infrastructure internally on its own servers. Full data control is non-negotiable for them.

The trend is clear: the public sector adopts Kubernetes, but on its own terms. No US public cloud. No data leaving Switzerland. It is slower, but it is solid.

Regulated Sectors: Hybrid Cloud Under Constraints

Banks, insurance companies, and healthcare players in Swiss Romande predominantly adopt hybrid architectures. Part of the infrastructure stays on-premise (sensitive data, critical systems), while less sensitive workloads migrate to Swiss cloud.

We have worked with pension funds that must balance nFADP/GDPR requirements with the need for modernization. The typical compromise: a Swiss cloud for business applications, and on-premise infrastructure for the most sensitive data. The challenge is to modernize without compromising compliance.

According to the Flexera State of the Cloud Report 2024, 89% of enterprises have a multi-cloud strategy globally. In Swiss Romande, we observe more of a "hybrid-cloud" model with a strong sovereign component, which is a more pragmatic variant of multi-cloud.

CI/CD and Automation: The State of Pipelines

Deployment automation is probably the area where we see the largest maturity gaps between our clients.

GitLab CI Dominates in Swiss Romande

Of our 100+ clients, 65% use GitLab CI as their primary CI/CD tool. It is the de facto standard in Swiss Romande, and this is no accident. GitLab is European (Dutch company), it offers an all-in-one solution (repository, CI/CD, registry, issue tracking), and it exists as a self-hosted version for organizations that want to keep control.

The trend has been stable for 3 years. GitLab holds a dominant position and does not appear to be losing it.

Azure DevOps Growing

About 20% of our clients use Azure DevOps, primarily in the financial sector and large enterprises. The Microsoft ecosystem is deeply embedded in these organizations (Active Directory, Office 365, Teams), and Azure DevOps integrates naturally. Several of our clients in the financial sector use Azure DevOps to manage their deployment pipelines.

GitHub Actions for Startups and Software Vendors

The remaining 15% use GitHub Actions, almost exclusively among startups and software vendors. The advantage: it is quick to set up, the actions marketplace is rich, and most developers already know GitHub.

The risk: GitHub is owned by Microsoft (thus subject to the US CLOUD Act). For sensitive data, this creates a sovereignty issue that some of our clients are beginning to take into consideration.

The Real Maturity Level

Having a CI/CD tool does not mean having a mature pipeline. According to the DORA 2024 report, "elite" teams deploy on demand with a change failure rate below 5%. In Swiss Romande, here is what we observe:

  • 15% of our clients are at the "elite" level: continuous deployment, comprehensive automated testing, automatic rollback.
  • 35% are at the "high" level: daily or weekly deployment, good test coverage, approval processes in place.
  • 35% are at the "medium" level: monthly deployment, partial testing, some manual steps remaining.
  • 15% are at the "low" level: manual deployment, little to no automated testing, no staging environment.

The good news: these numbers improve every year. In 2022, only 5% of our clients were at the "elite" level.

Observability: From Monitoring to SRE

Observability is the area where we see the most progress, but also the most persistent gaps.

Prometheus and Grafana: The Standard

Prometheus and Grafana have become the indispensable duo for metrics. According to the CNCF Annual Survey 2024, Prometheus is used by 86% of cloud-native organizations. In Swiss Romande, we observe a similar rate among our clients who have adopted Kubernetes.

Elastic Stack for Logs

For log management, Elastic Stack (Elasticsearch, Logstash, Kibana) remains dominant among our clients. It is a solid choice but requires resources to maintain. We see a gradual migration toward Grafana Loki for clients who want to simplify their stack.

Centreon for Legacy

Organizations with significant IT legacy (public sector, manufacturing) often use Centreon for monitoring their traditional infrastructure. The challenge is integrating this legacy monitoring with cloud-native observability. We help these clients make the two worlds coexist during the transition.

The Painful Finding

Only 30% of our clients had a complete observability stack (correlated metrics, logs, and traces) before our intervention. Most had basic metrics (CPU, RAM, disk) but no distributed tracing and unstructured logs.

A Datadog study confirms that organizations with a complete observability stack detect and resolve incidents 10 times faster. We see this at our clients: those who invest in observability go from "incident resolved in 4 hours" to "incident resolved in 20 minutes."

The GitLab DevSecOps Survey 2024 indicates that 56% of teams identify lack of production visibility as their main barrier. Among our Swiss Romande clients, that figure is probably higher.

Data Sovereignty: The Swiss Factor

This is the topic that has evolved the most over the past three years.

The nFADP Accelerates Everything

Since the nFADP came into force in September 2023, we have observed a clear acceleration of data localization projects. According to the KPMG Switzerland Cloud Monitor, 72% of Swiss companies consider data localization a top priority. This figure matches our field experience.

The Exodus from Hyperscalers

We are helping more and more clients migrate away from AWS, Azure, or GCP toward Swiss solutions. It is not that these platforms are technically inferior. The problem is legal: the US CLOUD Act allows the US government to access data stored by American companies, regardless of where the servers are physically located.

For a Swiss SME processing health data or financial data, this is a risk it can no longer ignore.

Hikube as a Sovereign Alternative

It is in this context that we developed Hikube.cloud, our sovereign cloud platform hosted in Switzerland. Beyond managed Kubernetes, Hikube offers virtual servers, managed databases, S3 storage, GPU as a Service and streaming (Kafka, RabbitMQ). The idea: offer a complete alternative to hyperscalers with a contractual guarantee of data residency in Switzerland, with no CLOUD Act exposure.

According to IDC, the Swiss cloud market is growing at 25% per year. We observe that the share of sovereign solutions within this market is growing even faster.

The Number That Surprised Us

40% of our new engagements in 2025-2026 include a data sovereignty component. That was 15% in 2022. The nFADP is not just a legal topic -- it has become a first-tier technical criterion in procurement processes.

Persistent Challenges

Not everything is rosy. Here are the obstacles we encounter regularly.

DevOps Skills Shortage in Switzerland

The DevOps job market in Swiss Romande is tight. Experienced profiles are rare, and there are more open positions than candidates.

The Puppet State of DevOps Report 2024 confirms that teams with mature practices manage 4x larger infrastructure with similarly sized teams. The problem in Swiss Romande is building that first mature team.

This is one of the reasons why managed services are gaining ground: when you cannot find the skills, you outsource them.

Resistance to Change

Traditional organizations (banks, public sector, manufacturing) have a culture of caution that slows adoption of new practices. "We have always done it this way" remains a phrase we hear regularly.

Change does happen, but it takes time. We have learned that it often takes a major incident (extended outage, failed compliance audit) to trigger a real transformation.

Cost of Training and Tool Changes

Moving to Kubernetes, setting up CI/CD, deploying an observability stack, migrating to Swiss cloud: each step has a cost. For a 50-person SME, a comprehensive DevOps transformation program costs between CHF 20,000 and CHF 150,000 over 18 months (training, tooling, consulting, infrastructure).

This is a significant investment. But the DORA report shows that organizations investing in these practices achieve 4 to 5 times better operational performance. The ROI is real, but it takes time to materialize.

Our Predictions for 2027

Based on what we observe today, here is what we anticipate for next year.

Platform Engineering Will Replace "DevOps" as a Term

The term "DevOps" is starting to wear thin. More and more of our clients talk about "Platform Engineering" to describe their approach. The idea: build an internal platform (based on Kubernetes, CI/CD, observability) that developers use as self-service, without needing to understand the underlying infrastructure.

According to the CNCF, Platform Engineering is the most cited trend in 2024-2025 surveys. In Swiss Romande, we see the first concrete implementations at software vendors and scale-ups.

AI Will Automate Part of Monitoring

Observability tools are increasingly integrating artificial intelligence for anomaly detection and incident correlation. Gartner estimates the observability market will reach $62 billion by 2026, with a growing share powered by AI.

In practice, we are starting to see clients use AI assistants to analyze logs and suggest root causes. It is still experimental, but the potential is real.

Multi-Cloud Will Become the Norm in Switzerland

Today, many of our clients are still on a single cloud provider (often by default rather than by choice). We anticipate an evolution toward multi-cloud architectures with a mandatory Swiss component. The typical pattern: sensitive workloads on Swiss infrastructure, non-sensitive workloads on European public cloud.

Sovereignty Will Become a Competitive Advantage

Swiss companies that can demonstrate complete data sovereignty are beginning to use it as a commercial argument. "Your data stays in Switzerland, under Swiss jurisdiction" is becoming a differentiator, not just a regulatory constraint.

We already see this among our fintech and healthtech clients: data sovereignty is mentioned on their website, in their pitch decks, and in their responses to procurement processes.

What This Means for Your Company

If you are a Swiss Romande SME that has not yet started its DevOps transformation, here is what I recommend:

  1. Start with observability. It is the foundation for everything else. You cannot improve what you do not measure. Prometheus, Grafana, structured logs. Start there. Our observability guide details the steps.

  2. Automate your deployments. Even if you do not adopt Kubernetes right away, set up CI/CD. GitLab CI is a solid starting point.

  3. Clarify your position on sovereignty. With the nFADP, you need to know where your data goes. Our article on the nFADP will guide you through the audit.

  4. Evaluate Kubernetes when you have 3+ services to manage. Not before. Kubernetes is powerful, but it has an adoption cost. Our Kubernetes guide for SMEs provides a realistic roadmap.

  5. Consider managed services if you lack internal skills. This is not a sign of weakness -- it is pragmatism. Our SLA vs Managed Services comparison will help you choose the right model.

Methodology

This article is based on feedback from over 100 client companies we have supported between 2016 and 2026. These companies span the finance, healthcare, public sector, logistics, IoT, and software vendor sectors across Swiss Romande (Geneva, Lausanne, Neuchatel, Fribourg, Valais).

Internal figures (adoption percentages, tool distribution) come from our CRM and engagement reports. External statistics are sourced from the CNCF Annual Survey, the DORA report, the Flexera State of the Cloud Report, Datadog, Puppet, Gartner, KPMG Switzerland, and IDC.

This is not a scientific survey at the scale of Switzerland. It is a field-based experience report, with the biases that implies: our clients are companies that have chosen to invest in their infrastructure or on our cloud, which likely makes them more advanced than the average.

But that is precisely what makes these observations useful: they show where companies go when they take the subject seriously.

Related reading:


Found this article helpful? Discover how Hidora can help: Professional Services · Managed Services · SLA Expert

Does this article resonate?

Hidora can support you on this topic.

Need support?

Let's talk about your project. 30 minutes, no strings attached.