Professional Documents
Culture Documents
Executive summary 3
Key findings 5
Conclusion 20
The 2021 survey found that, overall, observability adoption is on the rise.
Organizations on the higher end of the observability maturity spectrum are
realizing benefits, such as:
• Higher productivity
• Improvement in code quality
• End-user satisfaction
• Software developer retention
Teams further along the observability maturity spectrum not only realized
additional benefits; they also realized those benefits with higher-quality
business outcomes—for example, virtually all the Advanced and nearly 9 in 10
of Intermediate respondents provide customer experiences that leave their
customers either “Always Satisfied” or “Sometimes Satisfied” with their services.
In short, teams on the higher end of the observability spectrum realize clear value
in several areas, including an improved ability to identify problems and solve
those problems faster, more satisfied customers, and more confidence deploying
frequently to production when compared to their less-mature counterparts.
The 2021 report details how observability can help your organization achieve
production excellence and gain a competitive edge.
This survey was conducted from December 9, 2020 – January 11, 2021. Invites to participate in the
survey were distributed via a Honeycomb email list and social media outreach. A total of 405 respondents
participated in the survey. While the margin of sampling error cannot technically be calculated for online
populations where the relationship between sample and universe is unknown, the margin of sampling error
for equivalent representative samples would be +/- 4.9%.
While observability adoption increased this year, shifts in maturity were only
significant at the upper and middle parts of the spectrum (a 4% increase in
Advanced teams and a 2% increase in Intermediate teams). Respondents
indicated several barriers to progressing their goals, but those in the middle of
the maturity spectrum disproportionately cited lack of skills to implement as a
barrier to adoption.
Fig. 1
Devops/SRE
Developers
Architects
Managers
CIO/CTO
Operations
Figure 1: When asked which members of their team were responsible for driving decisions to practice
observability, respondents primarily chose Developers, DevOps, and SREs. Some respondents identified
as the person in their team who introduced observability practices; their roles were 21% developers, 36%
DevOps or SRE, 14% managers, 16% architects, 10% CIO/CTO, and 3% operations.
• Immediately identify the problem when something breaks and the impact it
has on other systems (69%)
• Understand their entire system, from high-level trends to specific events and
outliers (63%)
Fig. 2
Figure 2: Respondents were asked to indicate whether they agreed or disagreed that their teams had
various capabilities that gauge a comprehensive understanding of their production systems.
Most respondents fall in the evolving middle, where they practice observability
processes or tooling, but rarely both. They also report some, but not most, key
capabilities, such as identifying and resolving bugs before and after deploying
to production. Lastly, at the time of the survey, approximately one in five
respondents were not practicing or using observability tooling, but have plans to
do so within the next year.
Our observability maturity model evaluates the extent to which teams practice
and realize the benefits of observability, and scores them along an observability
maturity spectrum. The observability maturity spectrum identifies five distinct
groups:
Fig. 3
Figure 3: This figure shows a spectrum of observability maturity that correlates with how well teams
report they are able to demonstrate five key capabilities.
In the Advanced and Intermediate groups, over 10% of those surveyed reported
a combination of practices and tooling that reflect a highly observable system.
Another 13% of the Intermediate respondents reported practices, tooling, and
outcomes consistent with relatively sophisticated observability practices,
down from 17% last year—but that 4% shifted to the Advanced group on the
observability spectrum. In other words, we’re seeing groups higher on the
maturity scale continue to improve.
Most survey respondents fall into the Novice group. Participants report
observability processes or tooling, but rarely both. There may be several reasons
for this misalignment, but we believe this may reflect mismatched expectations
among practitioners in this group.
Analyzing responses, this group is more likely to self-report that they are
practicing observability because they are using tools like logs, metrics, and
traces. However, they also do not report having the key capabilities associated
with an observability practice. They report that they do not have a comprehensive
understanding of their systems. They also report some, but not most, key
capabilities. Additionally, one in two respondents in this group indicated that
practices like observability and DevOps are siloed within their organization or
only practiced on a team-by-team basis.
Planning
Teams in this group reported practices and tooling that suggest identifying and
resolving problems faster are top priorities. More specifically, in this group, one
in four are at the very beginning of their observability journey and are starting
to practice on a team-by-team basis (26%), while one in three respondents plan
to practice observability in the next 12 months (33%). Approximately one in five
respondents do not currently practice or use observability tooling, but have plans
to do so within the next year.
Of the 17% of respondents that are unpracticed, more than half are enterprise-
sized (52% with 1,000+ employees) financial services (22% finserv) companies
and report higher Azure use (39%).
Fig. 4
Figure 4: This year, we saw a shift from Intermediate to Advanced as well as from Planning to Novice.
However, no similar shift was seen from Novice to intermediate, suggesting that adoption barriers may
be stalling progress for Novice teams.
• Code that is well understood, well maintained, and fewer bugs than average
• The ability to visualize context-rich events that allow efficient, focused, and
actionable on-call processes
• The ability to set and measure service level objectives, resulting in better
alignment between engineering and business goals
Figure 5: Groups in the Novice stage of maturity disproportionately report lack of implementation
skills and too much technical debt as barriers to adoption.
Fig. 6
Advanced/Intermediate
Novice
Planning
No Plans
Fig. 7
Advanced/Intermediate
Novice
Planning
No Plans
Figure 7: 97% of Advanced and Intermediate teams report confidence in catching bugs before
production vs. 7% of teams with No Plans. 58% of No Plans teams report little to no confidence
catching bugs before production.
Teams and organizations that are further along the observability maturity
spectrum realize better business outcomes, including more satisfied customers.
High maturity correlates to a more comprehensive understanding of the
entire system, which increases confidence in catching bugs before and after
production.
Figure 8: Intermediate & Advanced teams are three times more likely than teams who have not started
down the path of observability, and twice as likely as Novice teams, to deliver customer experiences
that always satisfy customers.
1. Does your organization understand the state of your entire system from end
to end, at any given time?
2. Can you identify the source of problems in production without shipping new
code?
3. When something breaks, do you have the ability to understand who is being
impacted?
4. How confident are you in your team’s ability to catch bugs BEFORE being
released to production?
5. How confident are you in your team’s ability to catch bugs AFTER being
released to production?
• Accelerate: The Science of Lean Software and DevOps: Building and Scaling
High Performing Technology Organizations, by Nicole Forsgren, PhD; Jez
Humble, and Gene Kim
• Site Reliability Engineering, by Betsy Beyer, Chris Jones, Niall Richard Murphy,
and Jennifer Petoff
• OpenTelemetry Project
Finally, take the time to bring all software engineering team members together
and practice observability on a daily basis in order to realize the meaningful
benefits. Adopting a culture of shared ownership during incident response and
post-incident reviews will help address system-wide issues and also reduce risk