New Relic Now ¿Sueñas con innovar más? Comienza a vivir tu sueño en octubre.
Reservar plaza
Por el momento, esta página sólo está disponible en inglés.

In this section, we explore observability deployment plans for next year and the next two to three years, AIOps (artificial intelligence for IT operations) usage plans, and what steps organizations are most likely to take in the next year to get the most value out of their observability spend.

Future of observability highlights:

83%

anticipated deploying at least 1 new capability next year

82%+

expected to deploy the 17 observability capabilities by 2026

47%

planned to train staff on how to best use existing observability tools

Overall, organizations continue to see the business value of observability—and expect to invest more in it.

Observability deployment plans

In addition to asking survey-takers about their current observability deployment, we inquired about their deployment plans for next year and the next two to three years.

Deployment plans for next year

Looking out to 2024, the results showed that:

  • Most (83%) expected to deploy at least one new capability in the next year.
  • More than half (51%) expected to deploy one to five.
  • Nearly a third (32%) expected to deploy six or more.
  • Just 17% did not expect to deploy any new capabilities.
  • About three in 10 expected to deploy machine learning (ML) model performance monitoring (31%), distributed tracing (29%), and synthetic monitoring (29%).

When we look at the summary for one year out, capability deployment expectations are at least 90% for capabilities like network monitoring, database monitoring, security monitoring, and alerts.

Even with some of the observability capabilities that respondents were less likely to have already deployed (like synthetic monitoring, Kubernetes monitoring, distributed tracing, serverless monitoring, and ML model performance monitoring), we see numbers in the high-60% to mid-70% range.

These results indicate that respondents have big plans to deploy additional observability capabilities again this year. By mid-2024, at least two-thirds expected to deploy each of the 17 different observability capabilities (currently it’s 23% or more).
Number of new observability capabilities planned for next year
83%

expected to deploy at least 1 new capability next year

Regional insight
Those in North America were the most likely to say they don’t plan to deploy new capabilities in the next year, but they also had more capabilities currently deployed than other regions.

Organization size insight
Small organizations were the most likely to say they don’t plan to deploy new capabilities in the next year, which is surprising since they had the lowest current deployment rates.

Industry insight
Government respondents were the most likely to say they don’t plan to deploy new capabilities in the next year (28%).

Deployment plans summary

Forward-looking enterprise leaders continue to implement observability as a business imperative. It’s interesting to see how aggressively respondents once again expect to have most capabilities deployed in the next year and the next two to three years.

By mid-2026, 82% or more expected to deploy each of the 17 different observability capabilities. Very few of our survey respondents did not expect to deploy these observability capabilities (up to 16%).

This stated intent to deploy a large number of observability capabilities is once again one of the most eye-opening results from this study as it suggests that most organizations may have robust observability practices in place by 2026. The finding highlights the current state of observability and growth potential in the near future.
82%+

expected to deploy the 17 different observability capabilities by mid-2026

Observability capabilities deployment summary for 2023 through 2026

AIOps usage plans

We wanted to know where organizations fall on the spectrum between the deployment of manually configured incident detection (alerting) and completely autonomous, AI-led incident detection. So we asked respondents to what extent their organization anticipates using AIOps in its incident detection and remediation workflows one year from now and found that:

  • Almost three-quarters (70%) planned to rely more on manually configured incident detection, while 52% planned to adopt a more AI-led approach.
  • Only 16% said it would be mostly manually configured.
  • Just 8% said it would be mostly AI-led.

Of the respondents who said they plan to deploy AIOps in the next year, 40% anticipated it being more manually configured, and 25% more AI-led.

These results indicate that while incident detection deployment plans for next year were more heavily weighted toward manual configuration, there’s an increasing demand for automatic detection. Therefore, we believe that there’s an opportunity for observability vendors to increase the adoption of automatic detection as they build customer trust in its accuracy, reliability, and relevance.

Regional insight
Asia Pacific plans were more AI-led, while North American plans were more manually configured.

Organization size insight
Large organization plans were more AI-led, while small organization plans were more manually configured.

Industry insight
Energy/utilities and IT/telco plans were more AI-led, while healthcare/pharma, government, and financial services/insurance plans were more manually configured.

Anticipated deployment of manually configured incident detection (alerting) or completely autonomous, AI-led incident detection for next year

Value maximization plans for observability

We were interested to know what steps organizations are most likely to take in the next year to get the most value out of their observability spend. The survey results showed that:

  • Nearly half (47%) planned to train staff on how to best use the observability tools they have.
  • Roughly two in five (41%) planned to consolidate tools.
  • About a third planned to optimize their engineering team size (33%) and/or reduce spending across the board (31%).
  • The rest planned to use open source (27%), switch to a more affordable vendor (26%), or observe less of their tech stack (20%).
  • Just 3% said they won’t take steps to get the most value out of their observability spend.

Of the respondents whose organization had achieved full-stack observability (by the report’s definition), 47% said their organization is most likely to consolidate tools in the next year to get the most value out of their observability spend compared to 38% whose organizations had not achieved full-stack observability.

These results indicate that once organizations reach the milestone of achieving full-stack observability, the next logical step is to optimize it further by consolidating tools. In addition, they show that training is a key factor in maximizing value for observability spend.
This is a clear sign that organizations are facing financial pressure and seeking to lay off staff or make the best use of what they have by training staff and consolidating where practical.
47%

planned to train staff on how to best use existing observability tools

Regional insight
North American respondents were more likely to maximize value through tool consolidation, while Asia Pacific respondents were more likely through reduction in spend.

Industry insight
Industrials/materials/manufacturing respondents were the most likely to say they plan to consolidate tools in the next year (46%), followed by services/consulting (44%).

Most likely steps to take in the next year to get the most value out of observability spend

I think having an observability tool is important, but how you use it is even more important. It should be doing the job for you. Because, if you don’t know how to use it, over time, you start realizing, that the cost is high but you’re not really getting any value.