Internal development platform adoption to exceed 90% within the next five years
Introduction
When Amazon CTO Werner Vogels coined the phrase “You build it, you run it” back in 2006, he ushered in the burgeoning DevOps development framework. With DevOps, organizations could more efficiently bridge the gap between software development and IT operations.
Since then, many organizations have embraced agile workflows and DevOps to speed up deployment times and reduce siloed development processes.
However, DevOps approaches remain imperfect and often fall victim to the same bottlenecks and silos they should prevent. Inconsistent development environments, inefficient automation of processes, or even a lack of internal resources can cause breakdowns in DevOps workflows. This can — and often does — lead to increased burnout among developers who are asked to take on more responsibilities to complete the same tasks.
A different approach is required, and platform engineering presents an alternative to traditional agile models that aims to solve DevOps’ most significant shortcomings. But is a platform engineering framework as effective — or even desired — as some have claimed? Are teams using them now, or do they plan to use them in the future?
We at DuploCloud wanted to answer these questions. We surveyed 500 IT professionals to learn about their development approach and determine what they’ve found most effective. We hope this survey provides the foundation for improving your organization’s development framework, enabling you to unlock its full potential.
Executive Summary
- Cloud adoption is prominent. Over 85% of respondents are on the cloud or currently migrating; less than 15% are strictly on premises.
- Internal development platform (IDP) adoption is set to exceed 90% within the next five years. Platform engineering solutions are highly desired, and organizations are working to implement them.
- Structural problems cause burnout. 34.2% say burnout is their team's number one issue — ahead of internal bottlenecks and silos (32.2%) and open-source component approval or licensing issues (32.4%). Heavy workloads, inefficient processes, and poor allocation of resources cause burnout.
- IDPs are a potential solution for solving employee burnout. 50% of respondents would feel less stressed and more satisfied at work if their organization shifted to a platform engineering model.
- Adoption roadblocks include concerns about lengthy iteration times and entrenched processes.
Methodology
We surveyed 500 actively-employed IT professionals at organizations in the United States. Respondents come from a variety of backgrounds and fields, including:
- Data scientists
- AI/ML architects
- IT solutions and services providers
- Software engineers/developers
- DevOps engineers/Operations managers
- Other
Survey answers were collected in February 2023.
DevOps vs. Platform Engineering
Before we dive into the survey data, we must first define the difference between traditional DevOps approaches and platform engineering.
DevOps combines software development and IT operations to make organizations more efficient. With DevOps, teams are responsible for finding and becoming experts on their tools and workflows, which adds to each employee’s cognitive load as they take on more tools and microservices to perform their jobs.
Platform engineering shifts these individual responsibilities onto a centralized platform. Platform teams build and integrate tools that dev and ops teams need into an internal development platform. With platform engineering frameworks, developers don’t need to be experts in the tools — they can just use them, knowing they’ve been vetted for use in the central platform.
The Cloud and Platform Engineering Go Hand in Hand
Current Cloud Usage Stats
The cloud is no longer the future of software development — it’s the present. Virtually all development teams embrace it to improve processes, data access, and deployment efficiency.
Over 85% of respondents are either on the cloud or currently migrating. 31.2% of developers work within a multi-cloud infrastructure, with most others either working in public or hybrid clouds or migrating. The number of organizations utilizing only on premises configurations continues to shrink — less than 15% are strictly on premises.
Along with cloud computing, 69.8% have most or all services on containers, using them as the foundation for their DevOps processes. Plus, a further 18% are currently migrating, with only 12% reporting no plans to migrate to containers.
Terraform and Kubernetes are the most common orchestrations, surveyed at a combined 58.44%. Because developers are most likely to know how to use either Terraform or Kubernetes, implementing these orchestrations will bring newly-hired employees up to speed faster.
Mesosphere and Docker Swarm are roughly half as common, each at 16.3%.
78% of developers deploy code at least once per week, if not more. Thanks to a combination of DevOps processes, cloud computing configurations, and containerization, developers are no longer afraid to make regular deployments. Deployment times greater than a week are a thing of the past.
Current Platform Engineering Adoption Rates
While standard DevOps practices are currently in use at most organizations, platform engineering is the natural progression forward to enhancing those practices.
91.2% of developers either already have an IDP in place or plan to implement one within the next five years. It’s clear that development teams already see IDP implementation as a necessary aspect of cloud-native development to meet modern business demands.
Regarding what employees prefer to work with:
- 33.6% prefer working in the platform engineering model — it’s the most popular model cited.
- 27.4% prefer dedicated dev and ops teams.
- 18.2% prefer the DevOps model.
- 13% prefer a stratified model.
In reality, most organizations currently utilize dedicated dev and ops teams at 37.5%.
These results show the discrepancy between where developers are and where they want to be. The current DevOps model isn’t working for many organizations, and those that shift to platform engineering prefer it more. Many development teams currently using DevOps or other structures would prefer a change to a platform engineering framework.
Satisfaction and Failure Rates
It’s clear that the DevOps framework has enhanced developer efficiency and accuracy. But how do those outcomes compare to those that have implemented an internal development platform? We compare several key performance metrics below.
Most frameworks are reasonably effective at preventing failure. 64.5% of all developers cite a failure rate under 20%.
However, teams using a platform engineering framework see more optimal failure rates. 70.48% of developers currently using an IDP cite failure rates under 20%.
All frameworks are generally good at recovery. 53.6% of developers report that their mean time to recovery (MTTR) is between an hour to a full day.
However, platform engineering models provide more reliable recovery times. Organizations currently implementing IDP systems have seen their MTTR rates over a day shrink from 13.2% to 7.4%.
Regardless of the framework teams use, 71.4% of respondents have a positive outlook on the future of their organization.
When employees get to work with IDPs, this positive outlook increases to 90.75%. Platform engineering makes developers happier!
Burnout Concerns
34.2% of respondents say burnout is their team's number one issue. Open source component approval or licensing issues are very close behind at 32.4%. Burnout and resource bottlenecks are causing roadblocks in day-to-day operations.
When breaking down these results by employee or industry, they appear even starker.
- Software and DevOps engineers face a significant risk of burning out — 42.7% cite burnout as a substantial issue impacting their team’s productivity.
- Burnout is the number one issue impacting senior engineers (employees with over six years of experience), at 48.9%.
- Fewer junior engineers (employees with less than six years of experience) cite burnout as a problem, at 35%, but still believe it’s their number one issue.
- Employees who work for small businesses also cite burnout as their number one issue, at 37.5%
- Burnout hits large enterprises especially hard — it’s their number one issue, reporting at 58%.
Unless organizations take burnout seriously, these numbers will only climb — and as a result, organizations will encounter declining productivity rates, poor performance, dissatisfied employees, and staff leaving in droves for less stressful environments.
When examining the source of burnout employees face, 47.2% say burnout is caused primarily by high workloads. The next most commonly cited cause of burnout is inefficient processes, at 35.2%.
Modern-day engineering teams are asked to do too much to keep pace with business demands using current methods, especially when they have to build, learn, and manage the tools and workflows they use.
These stats break down accordingly:
- 56.3% of software and DevOps engineers cite high workloads as the leading cause of burnout. Inefficient processes come next, at 38.5%.
- 54.4% of senior engineers cite high workloads, while 30% attribute burnout to inefficient allocation of resources and staff. They’re also more likely to work on large teams (nearly 40% work on teams with over 200 people) and still have too much to do — this speaks to their need for greater efficiency, as people aren’t being optimized.
- Junior engineers are experiencing many of the same stressors as their senior cohorts; 57.7% cite high workloads, and 41.2% cite inefficient allocation of resources and staff as the causes behind burnout. Their junior status insulates them from the effects of burnout — at least for now. It’s only a matter of time before these employees spend too much time suffering heavy workloads and buckle under the stress.
Developers Want a Platform Engineering Model
Many employees have learned about platform engineering and want to try it out to improve processes and reduce stress. 50% of respondents would feel less stressed and more satisfied at work if their organization shifted to a platform engineering model.
These results break down accordingly:
- 57.8% of software and DevOps engineers would feel less stressed and more satisfied at work by shifting to a platform engineering model.
- Even with years of experience in DevOps platforms, 53.3% of senior engineers would feel more satisfied in their work by transitioning to a platform engineering model.
- While junior engineers experience lower burnout rates, they are more likely to feel less stressed and more satisfied after shifting to a platform engineering model, at 61%. Given the similar stressors as senior engineers, it’s likely that switching to an IDP will improve junior engineer growth and retention.
- Over half of all employees working at small businesses would feel more satisfied working with an IDP.
- Employees working at large enterprises especially want to make the jump — nearly three out of four employees on teams with more than 200 people would feel less stressed with a shift to the platform engineering model.
Platform Engineering Adoption Roadblocks
While most respondents either want to transition to an IDP or already have, 16.6% do not plan to implement a platform engineering model or don’t know if they will.
32.5% of respondents with no plan to implement an IDP see lengthy iteration and development cycles as their team's biggest issue. The next most significant concern is talent shortages and recruitment at 27.5%. 37.5% of respondents further claim that burnout is minimal at their organization.
To break these statistics down further, 50% of non-adopting software and DevOps engineers believe security vulnerabilities are the most significant issue faced by their team, followed by talent shortages and recruitment at 40%. 50% also believe that burnout is minimal at their organization.
Even beyond the reasons surveyed here, there’s likely an entrenched reluctance to change from legacy systems:
- When restricting the cohort of respondents with no plans to implement an IDP to just engineers, they become some of the most senior of those surveyed. 60% of engineers have six or more years of experience.
- Non-adopting engineers are split evenly between on premises (30%) and multi-cloud (30%) — that’s double the on premises percentage of any other cohort.
- The majority of this cohort also has no plan to migrate to containers at 40%.
These answers are largely out of alignment with other respondents and the industry, likely speaking to an “old school” way of thinking and doing things. This reluctance will see these organizations left behind by businesses that can embrace platform engineering processes that improve productivity while reducing employee burnout.
Essential Takeaway
Platform engineering is the future for DevOps and is set to exceed 90% adoption rates after doubling within the next five years. If teams aren’t currently using an IDP, they either plan on implementing one or dream about using one.
Organizations that have made the shift are already seeing benefits. IDPs help reduce failure rates, decrease mean time to recovery rates, and boost an employee's overall outlook of the organization’s future.
If your organization isn’t considering implementing an IDP, chances are people want it. 50% of all respondents say a platform engineering model would make them feel less stressed and more satisfied at work. Do not let institutional roadblocks or old ways of thinking prevent you from adapting.
About DuploCloud
DuploCloud is an all-in-one DevOps Automation platform and can be the basis for your IDP. The platform enables developers to automate infrastructure provisioning and cloud orchestration with built-in security and compliance configurations, reduce bottlenecks, enable developer self-service, and easily spin up instances that meet PCI DSS, HIPAA, SOC 2, and GDPR specifications. Plus, organizations can gain peace of mind with ongoing compliance monitoring and audit-ready reporting. DuploCloud enables teams of all sizes across industries to deploy cloud-native infrastructure ten times faster and reduce operational costs by 75%.
Want to see DuploCloud in action? Contact us at info@duplocloud.net or visit duplocloud.com to request a demo.
