Spring 2021

Building blocks to
boost your DevOps
transformation
Contents

Introduction 2

Chapter 1

Empowering people, changing culture 6

Chapter 2

Making progress with process improvements 13

Chapter 3

Tooling up for transformation 23

Conclusion

Ready for that next step? 28

1
 Introduction

DevOps has gone mainstream. According to HGS Digital’s A Closer

Look at DevOps, only 3 percent of software development businesses
surveyed said they had never heard of DevOps in 2018, and 91 percent
said they did have plans to adopt it into their workflow. More recently,
Puppet Lab’s 2020 State of DevOps Report, showed that 79 percent of
companies were using mid-level DevOps practices, with just 5 percent
operating at a low-level.

Rather than asking themselves whether DevOps is something they

should be prioritising, senior leaders are already onboard. Now
they’re focused on ensuring that they’re using DevOps to their
advantage. But, with so many organisations confidently jumping on
the bandwagon, many claiming to be dedicated converts, what does
DevOps really look like on the ground? And how intrinsically has
the framework been adopted?

2
What does it mean to do DevOps?
‘We have an experienced DevOps team’ or ‘We’re hiring a DevOps
engineer’ are the types of sentiments commonly bandied around, but
they’re big red flags. DevOps isn’t a tool or individual process
restricted to a few people or isolated teams; it’s a mindset shift, much
like agile, and applies to IT as a whole. It shouldn’t be a standalone
department, rather a continuous culture that takes root and filters
throughout the business.

Adoption of DevOps changes a business’s culture, allowing teams to

work fast and with autonomy. As author and entrepreneur Gene Kim
said: ‘[DevOps is] really about enacting technology practices and
architecture that help organizations create a very fast – or even
continuous – product flow, from dev through testing through
operations, all the way to deployment, while often increasing reliability,
stability, and security.’

This is directly opposed to more traditional models. In the past, a

software company would have to involve hundreds of developers over a
period of weeks to deploy small upgrades or fix bugs. A waterfall model
– a linear approach where changes are batched up – made releases
unwieldy and risky, not to mention the time it would take to release a
new product. DevOps developed as a direct response to these
inefficiencies: slow turnaround, poor quality, lack of accountability and
poor teamwork in particular. Today, high-level DevOps adopters can
release reliable products at a much faster pace.

3
DevOps is going beyond agile
Agile is a well-established approach – it’s celebrating its 20th birthday
in 2021. By now, most organisations have made or are in the process of
some kind of agile transformation. But however hard you’re waving
your agility wand to accelerate product development, drive value, and
overhaul your organisation, it’s really all for nothing if you can’t get your
products to market at the same rate. No matter what comes before it, if
your software is still worming its way through traditional sign-offs and
endless testing, building up to a big release, then you’re not doing agile
properly – and you’re certainly not doing DevOps.

DevOps is agile’s spunky younger sister. Where agile software

development focuses on optimising the development life-cycle, DevOps
works to accelerate delivery, making it faster while maintaining quality
and ensuring continuous development with tight feedback loops. It
helps you carve a path from ideation to production, uniting software
development and tech operations for an end-to-end pipeline that
includes every stakeholder in the process.

As a combination of organisational philosophies, values, principles,

tools, and practices, DevOps requires people to change the way they
work and the way they think across all aspects of development and
operations – from front-end developers to back-end IT infrastructure
and security professionals.

4
The three tenets of DevOps
Three fundamental principles underpin DevOps:

Continuous
Flow Feedback learning and
experimentation

Together they ensure teams are united towards a common

goal and reaping the rewards of continuous integration (CI)
and delivery (CD). Its success rests on three core
components:

People Processes Tools

Over the following chapters, we’ll take a look at each in turn,

providing an overview of the key building blocks you need to
consider and providing our top tips to help you step up your
DevOps practices.

5
 Chapter 1

Empowering people,
changing culture

Looking back, old-school software development seems a little silly.

With its siloed developers and system architects waving over a
metaphorical wall at the operations engineers and database
administrators on the other side, it’s questionable that these two
fundamental facets of getting software out the door haven’t always
been on the same page.

While it might sound like a logical solution to overcoming process

inefficiencies, for some people, transitioning to DevOps is a big ask.
And it’s way out of their comfort zone. This is particularly true of
operations teams, who are used to working to reduce risk by
minimising change and taking things slow. Now they’re being expected
to speed up with continuous integration (CI) and continuous delivery
(CD), release regularly, and operate in a fail-fast pipeline. DevOps aims
to bridge the gap between developers’ change-ready ideology and
operations’ more cautious approach.

For DevOps to be successful, everyone has to be on board. If your

organisation understands and supports that a compatible culture will
develop and the new hires, tools, and tech you need to make it happen
will emerge in its wake. In this chapter, we’ll explore what a DevOps
culture looks like, how to get yours on track, and the people and skills
you need to help it flourish.

6
A conducive culture
No one said DevOps was easy. In fact, according to Gartner, through to
2022, 75% of DevOps initiatives will fail to meet expectations. The
reason won’t surprise you – issues around organizational learning and
change. Simply showing up one day and imposing a new set of
practices on your employees is dooming DevOps to fail.

Employee buy-in is paramount – knowing why the business is

expecting something new from them and how it’s going to make their
work better and easier will reduce resistance significantly. People might
buy in to your DevOps initiative initially, but they’re not going to stick
around and proliferate its ideologies and rituals for no reason. They
need to understand the deeper thinking and the benefits – both to them
personally and the wider group – to get stuck in.

7
Here are some key considerations to build buy-in and transform
your culture.

• Educate with feeling

Sharing information about how DevOps will make a real difference
is the first step to getting people onboard. Implement a change
management strategy that considers the relative understanding,
resistance, and experience of all teams. Everyone needs to come
along for the ride, but it will be bumpier for some than others.
The main goal is to give people a clear idea of how it’s going to
feel to work differently. The only way they will know this is by
doing. So share actionable insights which they can apply to their
work and observe the results, no matter where they are in the
pipeline. Take this a step further with a program that simulates the
experience of working in a fully DevOps environment.

• Do it gradually
No one likes a sudden shock to the system, and DevOps is no
different. A gradual change where employees can merge into this
new mindset will get better results in the long run. While driving
change from the top is vital, actual change will start at the bottom.
By implementing DevOps at the team level, it enables individuals
to realise the possibilities, discover potential pitfalls, and
overcome them – crucially before the whole business has made a
shift. This slow-burn approach allows people to redefine their roles
at a more realistic pace too.

8
• Collaboration is key
Being on the same page means actually getting stuck in together.
Without collaboration – from individual developers to C-suite
leaders –DevOps just doesn’t work. In larger corporate structures
where people and thinking are siloed, this is a big challenge.
Communication plays a huge part in this. Employees should feel
empowered to share knowledge with each other, for the
betterment of everyone. Consider using community forums or
collaboration tools (see chapter three).

• Banish blame and embrace trust

Forget the blame game; it won’t help you there. It’s vital that people
feel they can share information about failure and frustration
across the wider organisation without fear of the pointed finger.
Start from a place of trust. In the past, people were so confident
in control frameworks, checklists, and risk mitigation, that they
didn’t really need to trust their colleagues. But with DevOps trust
is earned through collaboration andsmall wins (that quickly
scale). Trust your teams to act responsibly, share their failures
honestly, and to uphold the principles and requirements outlined
by the organisation.

• Good enough is really good

People need to know that perfection is not the goal and that
improvement – in process, quality, and speed – is what they’re
striving for. ‘Good enough’ shouldn’t be considered a dirty phrase.
It’s what allows people to put something out there quickly, see how

9
 it works, and then iterate any issues away based on real-world
feedback. This is why fluidity and flexibility in DevOps are so
important: systems should support a fail-fast way of working
rather than stifle it.

• Budget for the best scenario

You’re going to need really good people to enable CI and CD
pipelines to work effectively. Don’t underestimate your needs:
make sure experienced, senior people are involved from the
outset, and invest in the right tools too. It’s going to take time to
achieve your goals, but you’ll get there much faster if you budget
accordingly rather than sidelining your requirements at the start.

• Rewards based on reality

A thriving, motivated workplace culture is one where people feel
rewarded and strive towards common goals. Rather than reward
teams for following protocols and meeting targets – especially
targets that might even conflict with those of other teams –
DevOps uses strategic metrics to determine success. Throughout
the organisation, objectives should be centred around business
outcomes and measured accordingly.

10
The right stuff
If it’s not clear by now, people are the most important component of your
DevOps transformation. Making sure you have the right talent to ignite
the spark and carry the torch is half the battle. Despite the tech-heavy
nature of software development, when it comes to DevOps, soft skills –
like communication, collaboration, flexibility, and decision-making
– should be prioritised alongside security, coding, and infrastructure
knowledge. Don’t restrict your DevOps skills audit to IT teams alone:
make sure this mix of skills is present across the organisation.

When hiring or upskilling your existing teams, don’t forget the following:

• Expertise takes time: employees are not going to get up to speed in

a week. There are new methodologies and tools to learn, so have
patience. But don’t be naive and expect people to pick things up on
an ad hoc basis – self-learning during downtime won’t cut it.

• Program your people to win: dedicate the right time and resources
to helping your teams transition their skillset. One option is a
DevOps training program that helps them become familiar with the
tools, tech, and methods pivotal to transformation. FedEx took this
approach and upskilled 25,000 programmers with its own bespoke
training solution.[1]

[1] https://devopsinstitute.com/10-devops-trends-to-watch-in-2020/ 11
• Know when to bring in new people: it’s important to strike the right
balance between upskilling your existing employees and bringing
in new people to boost your DevOps capabilities. Your existing
employees might already have in-depth knowledge about your
organisation, but there are lots of benefits to bringing new hires on
board. Keep in mind the cost of hiring in DevOps skills and that
you’ll need a certain level of expertise to pick the right people too.

• Assess your technology solutions: make sure the tooling you’re

choosing is flexible enough to support the wider needs of the
organisation. Everyone needs to get up to speed on DevOps, but
only some people will develop an expert knowledge of tools.
Others can simply benefit from automation and efficiency.

People play a huge part in your transformation, so don’t sideline them

or overlook their needs. And bring everyone along for the ride. Make
sure teams are continuously informed, encouraged, and supported as
they learn and acquire DevOps skills, and you’ll reap the rewards.

12
 Chapter 2

Making progress with process

improvements

No matter the people in your teams or the tools you choose to use,
successful DevOps transformation can’t happen unless you step away
from traditional IT’s ‘trust but verify’ methodology to a new outlook of
continuous integration and delivery, incorporating a set of essential
processes. And a pick ’n’ mix approach won’t do here. While
organisations using DevOps will incorporate variations on the theme,
they will all be utilising the following at a minimum: collaboration,
automation, continuous integration and delivery, and monitoring.

These fundamental processes will form the foundation of your

DevOps pipeline, allowing everyone to work more efficiently and
effectively towards achieving faster, high quality software to suit the
end users’ needs.

13
Collaboration: achieve more together
Collaboration is more than just a cultural facet – it’s a process-driven
practice that you must build into the way you work; in particular, it helps
IT organisations with globally distributed teams to concentrate on
aligning to achieve a common goal, whether that’s delivering a release
by the end of the day or a wider business objective for the year. Here are
just some of the problems collaboration can help you overcome:

• Operational inefficiencies, including limited automation and

release delays

• Problematic handoffs caused by different time zones and

disparate teams

• Redundant work and limited team problem-solving because

of communication silos

• Impromptu changes that impact product stability

• Inbox overload and unnecessary meetings

• Uncertainty over roles and responsibilities

14
 Top tips for collaboration

• Use the right tools

Feedback and communication are key to ensuring collaboration
is a constant and continuous process. Tools like Slack, Microsoft
Teams, and Google Chat make it much easier to stay in touch
and on top of what everyone’s working on.

• Set clearly defined goals

Make sure organisational, team-level, and individual goals are well
defined with clear objectives. Big-picture goals should be the
same across the board so everyone’s priorities are aligned.

• Embrace diversity
Your team might be geographically spread out and culturally
diverse, so inclusivity and sensitivity are key to smooth
collaboration. Don’t make assumptions, and embrace different
points of view; it will help to unite the team in the long-run.

• Emphasise equality and unity

For everyone to feel invested, they need to know their contribution
is valued. Unifying language – face to face and through digital
communications – and leaders who inspire honesty and respect
help create a sense of oneness and equality.

15
Automation: repetition right at your fingertips
What sets DevOps apart is automation. It’s only possible to achieve this
level of efficiency by relying heavily on automated processes to lighten
the load. For that, you need to have the right suite of tools in place (see
chapter three) for automation, building, integration, testing, environment
configuration, deployments, workflows, data synchronisation across
systems, and more, leaving people to focus less on repetitive tasks and
more on complex problems.

If you already have some level of automation in place or are considering

automating something new, consider how the following techniques can
help improve your efforts and boost your productivity.

• Have a clear vision

Make sure your automation strategy is aligned with wider
business goals. Having a defined purpose behind why you’re
using automation will ensure it can really deliver value – whether
you’re improving operational efficiency, empowering employees,
or ensuring process quality. And it’s vital that key stakeholders are
involved in strategic planning, particularly the IT team. If they see
automation as a core business objective, they’ll support it and
make sure you have the resources you need to deliver.

• Plan for scale

Think ahead about how automation can scale across the
business. Will you take a centralised approach that incorporates
the entire organisation or retain automation within a particular

16
 function with a central team that controls standards? If you
choose the former, you might want to create a ‘centre of
excellence’ to lead the way in exploring and adopting new
technology or practices. Whereas that latter might require
delivery pods that are responsible for developing their own
automated processes.

• Fastest benefits first

Pick processes that automation can help generate big benefits for
at pace. These are the kinds of processes prone to human error,
that require a large number of manual and repetitive tasks, or that
are impacting customer experience. Before automation takes
place, you might be tempted to streamline a process. Consider
whether automation alone might be enough, because
improvements are costly and might have little impact.

• Take your time to define

Don’t rush to robotics – before you even start to automate
processes, gather all the right information to help mitigate any
problems. That means bringing subject experts along for the ride
and holding a process walk-through for the people who matter
most. Make sure you document as you go and share with the team
how automation will differ from the human process you previously
had in place. Once everyone’s happy, it’s time to get sign-off from
key stakeholders and conduct peer reviews.

17
• Pick people wisely
You’ll want a skilled and experienced lead developer to run any
automation implementations, which might mean hiring in some
new talent or working with an external partner. Make sure your
people receive training where necessary – exceptional development
and process analysis skills are vital. And unless you want standards
to slip as the team grows, consider appointing people to oversee
quality and manage the robots.

• Deliver with care

Once development is underway, don’t skip any steps. Sign-off
testing helps get everything up and running effectively before
unleashing automation on the business. When everything’s working
as you want it, don’t let the robots have all the fun. Build in business
referrals or exceptions, so that the operations team can intervene
where necessary or if things don’t go to plan. Automation can only
work with a strong supporting IT infrastructure with capacity to
support where necessary.

18
Continuous integration and delivery:
an agile approach
Continuous integration (CI) and continuous delivery (CD) forces
developers to merge their source code updates with others’ far more
frequently than they might otherwise choose to, and operations teams
to make sure they’re not creating a bottleneck with manual testing, to
get releases out to users.

Unlike waterfall-style development, this brings conflicts and issues to

light much earlier in the pipeline. Meanwhile, continuous testing (both
automated and manual) ensures a high level of quality is maintained,
exposing any security or reliability problems in the process. But
automation alone doesn’t tick CI and CD off your list. Here are a few
other workflow and cultural implications to consider:

• How to shift left

CI/CD is all about ‘shifting left’ – testing early to prevent rather
than detect problems. It helps to increase the quality of releases
and avoid nasty problems at the end of the development cycle.
By working side by side, development and operations can
recognise and iron-out differences in deployment – a common
cause of failures in production – to create standard procedures.
You can reduce the failure rate further by making all environments
in the pipeline closely reflect production using cloud and
pattern capabilities.

19
• Sooner rather than later
Integration should happen early and often. Too long with one
developer and you run the risk of countless conflicts when their
code merges with the main repository. And if they integrate too
late in the process, it’s hard to understand conflicts, leading to lost
time. Regular integration also allows developers to share valuable
knowledge in bite-sized chunks.

• Stories should include tests

Developers should be involved early on in user stories, helping
them understand business requirements and build relationships
with product managers. It also means every feature can have a
plan for automated tests built in – it will cut down time spent
fixing bugs on each iteration and give you confidence to make
changes by quickly checking all previous features are working as
they should.

• Roles will be redefined

While they will still carry out some exploratory testing for complex
use cases, automation means QA engineers won’t need to waste
time testing trivial capabilities manually. It frees them up to
support developers and help build better testing strategies, tools,
and datasets.

20
 • Overcoming automation fears
Some people will be wary of CD – they’ll see automation as the
enemy as it’s essentially taking over parts of their job. Build
enthusiasm and confidence into your team with the help of
experienced automation cheerleaders. If you have the right people
for the job, they’ll soon realise that repetitive tasks are more
suited to automated tools, and that automation frees them up to
work on more interesting projects.

Metrics: measure your progress

Metrics provide the information you need to figure out the bottlenecks
and make all your DevOps efforts worthwhile. Underpinning all these
processes is the mantra ‘implement, measure, improve’. There’s no
point having a speedy, heavily automated CI/CD process with a truly
collaborative culture if you don’t take the time to measure what’s
working and take actions to improve what’s not.

There are three different kinds of metrics: time-based, quality, and

automation. The first will help you save time and ship code as quickly
as possible. Quality metrics, perhaps the most important of the three,
ensures that no matter the speed, quality doesn’t suffer. While the
third set measures the impact automation has had on your
deployment process.

21
Together, these help to measure software performance and
availability, improving stability in the process. By continuously taking
measurements, you can identify causes quickly, preventing outages
and minimising disruption for users. Some typical data points you
might consider measuring include time to market, deployment
frequency, change success/failure rate, security test pass rate, and
mean time to recovery from pipeline failure.

Make more of your value stream

Metrics go hand in hand with value stream management (VSM).
A value stream is everything in the delivery lifecycle – from inception
to production – that’s required to deliver your products to customers.
A product’s value is determined by its customers. To maximise ROI,
each step in the value stream should create value, according to the
customer’s wants and needs. Processes like value stream mapping
bring together stakeholders from across the business to document all
the activities and hand-offs involved.

By effectively managing software and service delivery this way, you can
more easily identify which steps are adding value and which aren’t.
As well as leveraging real-time metrics, VSM helps you break down
operation silos, encourage cross-team collaboration, and coordinate
and automate workflows.

22
 Chapter 3

Tooling up for transformation

As teams grow and DevOps processes scale across your organisation,

consistent performance is paramount. The tools you choose need to
be available around the clock, and capable of performing at a high level
to support your collaborative pipeline. But using DevOps tools doesn’t
automatically mean you’re applying DevOps principles. Only by using
these tools holistically with a unified culture and the processes
explored in chapter two will you be able to fully transition the way you
develop and deliver your products.

As you build your toolchain, you’ll need to include those that step up
your collaboration, keep context-switching to a minimum, aid
automation, encourage visibility, and allow for effective monitoring.
You might choose an all-in-one solution that doesn’t integrate with
third-party tools or an open toolchain that you can customise with the
pick of the bunch, tailoring your toolset more closely to your
organisation’s needs.

Whichever approach you use, the tools you choose will correlate with
the processes mentioned in the previous chapter, notably planning and
building, continuous integration and delivery, metrics, operations, and
feedback. Given the fast-moving nature of DevOps, your individual
needs, and the rapid development of new tools, we’ve outlined some
key considerations rather than making specific recommendations.

23
Build up your toolbox
• Primed for planning
Use tools like Jira, Confluence, and Slack to make work more
manageable and learn from users at pace. Collaboration is a
priority, as is sprint planning and the ability to gather feedback
and organise it into task-based outputs. All team members should
have access and be able to share and comment on everything
with ease. Integration features are also important to convert user
stories into your development backlog.

• Collaborative coding
Get up to speed on source control and collaborative coding tools
like Bitbucket and GitLab. These help you store your code in
chains, giving better visibility of every change and making
collaboration much smoother. Initiate pull requests to tell your
team about the changes you’ve pushed, speeding up peer review
and getting your code in the mainline much quicker.

24
• Keep it continuous
Make sure CI and CD stay top of mind with tools that make it
easier to check code in a shared repository multiple times a day
or as regularly as your organisation deems necessary. You’ll need
a production-identical environment to develop in so everyone’s on
the same page when they’re coding and a platform that
automatically applies tests to development branches and offers
real-time chat alerts – providing feedback from your team.

• Testing times
Automation rules the roost when it comes to testing tools – they’ll
help you speed up development, reduce risk, and test efficiently
with minimal effort. Think UI, security scanning, and load testing
all taken care of. What’s more, these analyse data and produce
reports that help your teams identify problem areas and course-
correct as appropriate. You’ll also want tools that can handle
exploratory testing and orchestration, as well as wallboards that
give everyone visibility over results, including operations.

25
• Hassle-free deployment
For automated deployment, you’ll need to follow good engineering
practices – deploy to your lowest-level environment first before
replicating up to production to point out any difference and
generate standardised actions. There are a number of useful tools
out there to help you automate deployment, including GitLab,
Microsoft’s Visual Studio, and Bamboo Server.

• Make more of monitoring

For monitoring to be effective, it must be ongoing – only by
recording data round the clock will you be able to understand your
software’s overall health. You need monitoring tools that integrate
with your group chat and software development platforms so
notifications, changes, and projects are all tracked in the same
system. This will allow you to identify and resolve problems
more quickly.

• Figure out your feedback loops

You’ll need multiple tools that encourage and support continuous
feedback, can deliver insight from all that information, and enable
everyone to have access. Harness the power of dashboards that
integrate with your code repository and deployment tools and let
you pull all your information together in one place, avoiding
unnecessary stress prior to a release.

26
Big tools, big decisions
When it comes to the tech and tools available, there’s a lot of choice.
Figuring out what will make most sense for your organisation can be
hard. In general, people pick and prefer solutions they feel most
comfortable with, which can be counterintuitive to transformation.
This approach can also lead to a tool binge – bringing on too much
technology where integration falls by the wayside and you end up
with nothing that really does the job you need. Take the time to
consult, and don’t lose sight of your budget – the wrong choice could
put your team behind competitors, prevent agile working, and make it
harder to scale.

27
 Chapter 4

Ready for that next step?

A successful DevOps transformation rests on an in-depth

understanding, wholehearted acceptance, and organisation-wide
adoption of the principles and practices outlined here. DevOps is
centered around the continuous improvement of software, moving
away from rigid methodologies. But, this agile mindset needs to apply
to the people, processes, and tools that make it happen too.

As you drive change across your teams, increasing reliability, stability,

and security of your products, you’ll be forced to consider whether to
build up or buy in the skills you need, assessing your people differently
for how they can help boost your DevOps capabilities. You’ll need to
take a long, hard look at your workplace culture to ensure it supports
your new processes, and learn which tools will help rather than hinder
your digital transformation.

28
Lots of companies want to implement devops, but don't really
understand why. And it’s important to keep in mind who it’s designed to
benefit most – the customer. But it’s not a simple case of ‘the
customer is always right’. DevOps helps organisations to strive for
greatness and offer the customer something they didn’t even know
they needed. These key principles allow you to better understand
customer pain-points, working efficiently to put things right. They help
you stay solutions-focused – releasing features that address real use
cases and removing barriers that impact the customer experience.

At Adaptavist, we help you stay ahead of the digital transformation

game with our DevOps solutions. With our strong and dedicated team
of experts, we combine the right mix of strategic-led consultancy and
technology-led solutions that place people, process, and tools at the
heart of your business strategy.

To discover DevOps in a whole new light

Contact us

29
 About Adaptavist

We help organisations transform to Platinum

Solution Partner
continuous change being their business as US GOVERNMENT

usual. We do this by supplying technology, Platinum

providing advice, and delivering change Marketplace Partner

through modern, iterative approaches to Platinum

development, deployment, and application Solution Partner

lifecycle management.

Adaptavist is Atlassian’s largest platinum

partner, supporting more than half of the
Fortune 500. We are uniquely placed to
provide our experience, expertise, and insight
to help your business.

Whether you want training for your team, to

build a software platform for your company,
or to automate your existing tooling, we can
help you. If you want to unlock the full power
of Atlassian and transform your business
at scale, get in touch with our team today.

Visit our website to learn more

adaptavist.com

Copyright © 2021 Adaptavist, All Rights Reserved