The astonishing value of speed

Slide 1

Slide 1 text

The astonishing value of speed Why you cannot afford not to go fast in IT Uwe Friedrichsen – codecentric AG – 2013-2021

Slide 2

Slide 2 text

Uwe Friedrichsen CTO @ codecentric https://twitter.com/ufried https://www.speakerdeck.com/ufried https://ufried.com/blog/value_of_speed/

Slide 3

Slide 3 text

Going fast

Slide 4

Slide 4 text

What do you mean with “going fast”? How fast is “fast”?

Slide 5

Slide 5 text

What do you mean with “going fast”? How fast is “fast”?

Slide 6

Slide 6 text

Features / Unit of time

Slide 7

Slide 7 text

Cycle times * Deployment frequency * Time from a new idea until the customer experiences its implementation and can provide feedback

Slide 8

Slide 8 text

With “fast” I refer to cycle times Deployment frequency simply must not compromise cycle times

Slide 9

Slide 9 text

What do you mean with “going fast”? How fast is “fast”?

Slide 10

Slide 10 text

With “fast” I mean the delivery speed (and quality) of the elite performers of the State of DevOps (*) report (*) https://services.google.com/fh/files/misc/state-of-devops-2019.pdf

Slide 11

Slide 11 text

https://services.google.com/fh/files/misc/state-of-devops-2019.pdf, p. 18

Slide 12

Slide 12 text

https://services.google.com/fh/files/misc/state-of-devops-2019.pdf, p. 20

Slide 13

Slide 13 text

https://services.google.com/fh/files/misc/state-of-devops-2019.pdf, p. 21

Slide 14

Slide 14 text

We don’t need it

Slide 15

Slide 15 text

What about post-industrial markets?

Slide 16

Slide 16 text

Post-what?

Slide 17

Slide 17 text

Formal part of value creation Solution: machine Dynamic part of value creation Solution: man sluggishness/low dynamic high dynamic high dynamic The historical course of market dynamics and the recent rise of highly dynamic and complex markets The dominance of high dynamics and complexity is neither good nor bad. It‘s a historical fact. t 1970/80 today Age of crafts manu- facturing Age of tayloristic industry Age of global markets 1850/1900 Spacious markets, little competition Local markets, high customi- zation Outperformers exercise market pressure over conventional companies We call the graph shown here the “Taylor Bathtub”. The “bathtub” curve Source: BetaCodex Network Associates, “Organize for complexity”, BetaCodex Network White Paper 12 & 13

Slide 18

Slide 18 text

Formal part of value creation Solution: machine Dynamic part of value creation Solution: man sluggishness/low dynamic high dynamic high dynamic The historical course of market dynamics and the recent rise of highly dynamic and complex markets The dominance of high dynamics and complexity is neither good nor bad. It‘s a historical fact. t 1970/80 today Age of crafts manu- facturing Age of tayloristic industry Age of global markets 1850/1900 Spacious markets, little competition Local markets, high customi- zation Outperformers exercise market pressure over conventional companies We call the graph shown here the “Taylor Bathtub”. Pre-industrial era Source: BetaCodex Network Associates, “Organize for complexity”, BetaCodex Network White Paper 12 & 13 Tailor-made solutions Mastery is key to success

Slide 19

Slide 19 text

Formal part of value creation Solution: machine Dynamic part of value creation Solution: man sluggishness/low dynamic high dynamic high dynamic The historical course of market dynamics and the recent rise of highly dynamic and complex markets The dominance of high dynamics and complexity is neither good nor bad. It‘s a historical fact. t 1970/80 today Age of crafts manu- facturing Age of tayloristic industry Age of global markets 1850/1900 Spacious markets, little competition Local markets, high customi- zation Outperformers exercise market pressure over conventional companies We call the graph shown here the “Taylor Bathtub”. Industrial era Source: BetaCodex Network Associates, “Organize for complexity”, BetaCodex Network White Paper 12 & 13 Cost-efficiently scale production Getting more done with less people is key to success

Slide 20

Slide 20 text

Formal part of value creation Solution: machine Dynamic part of value creation Solution: man sluggishness/low dynamic high dynamic high dynamic The historical course of market dynamics and the recent rise of highly dynamic and complex markets The dominance of high dynamics and complexity is neither good nor bad. It‘s a historical fact. t 1970/80 today Age of crafts manu- facturing Age of tayloristic industry Age of global markets 1850/1900 Spacious markets, little competition Local markets, high customi- zation Outperformers exercise market pressure over conventional companies We call the graph shown here the “Taylor Bathtub”. Post-industrial era Source: BetaCodex Network Associates, “Organize for complexity”, BetaCodex Network White Paper 12 & 13 Continuously respond to changing demands Continuous market adaption is key to success

Slide 21

Slide 21 text

Industrial Post-industrial Dictated by Supplier Driven by Consumer Market development Demand >> Supply Supply >> Demand Demand-Supply Ratio High Low Success certainty Long Short Refinancing period Market characteristics Wide, sluggish Narrow, fast Cost-efficiency & scale Fast feedback & adaptability Preferred strategy

Slide 22

Slide 22 text

In post-industrial markets not those survive who produce as cost-efficiently as possible, but those who adapt to the ever-changing needs and demands of the customers faster than their competitors. This requires fast feedback loops with your customers.

Slide 23

Slide 23 text

What does that have to do with IT?

Slide 24

Slide 24 text

1960 1970 1980 1990 2000 2010 2020 Complicated (Business functions) Complex (Business processes) Highly complex (Business nervous system) Software crisis Software engineering PC LAN Internet Business Support of IT Selective Holistic Complicated Complex “Moore’s law” Mobile IoT

Slide 25

Slide 25 text

IT has changed a lot over the decades ...

Slide 26

Slide 26 text

1960 1970 1980 1990 2000 2010 2020 Complicated (Business functions) Complex (Business processes) Highly complex (Business nervous system) Software crisis PC LAN Internet Business Support of IT Selective Holistic Complicated Complex “Moore’s law” Mobile IoT Software engineering ... but still we strive to control our IT of today ... ... based on the concepts we developed for an IT almost 50 years ago

Slide 27

Slide 27 text

Formal part of value creation Solution: machine Dynamic part of value creation Solution: man sluggishness/low dynamic high dynamic high dynamic The historical course of market dynamics and the recent rise of highly dynamic and complex markets The dominance of high dynamics and complexity is neither good nor bad. It‘s a historical fact. t 1970/80 today Age of crafts manu- facturing Age of tayloristic industry Age of global markets 1850/1900 Spacious markets, little competition Local markets, high customi- zation Outperformers exercise market pressure over conventional companies We call the graph shown here the “Taylor Bathtub”. Remember the “bathtub” curve? Source: BetaCodex Network Associates, “Organize for complexity”, BetaCodex Network White Paper 12 & 13

Slide 28

Slide 28 text

Slide 29

Slide 29 text

The role of IT today

Slide 30

Slide 30 text

The role of IT • Nervous system of the business • Enabler of (disruptive) new business models • Integral part of the business model (“digital transformation”) • Medium for the continuous customer communication

Slide 31

Slide 31 text

You cannot change anything in your business without touching IT. IT delimits how fast you can get feedback from your customers.

Slide 32

Slide 32 text

Today, business and IT are the same side of the the same coin. The other side is the market.

Slide 33

Slide 33 text

But my market is not post-industrial

Slide 34

Slide 34 text

There is still huge value

Slide 35

Slide 35 text

Higher organizational performance https://services.google.com/fh/files/misc/state-of-devops-2019.pdf, p. 5

Slide 36

Slide 36 text

Better work/life balance, reductions in burnout https://services.google.com/fh/files/misc/state-of-devops-2019.pdf, p. 6

Slide 37

Slide 37 text

Improves stability, reduces burnout https://services.google.com/fh/files/misc/state-of-devops-2019.pdf, p. 6

Slide 38

Slide 38 text

Shorter cycle times do not lead to more stress and chaos, but on the contrary to less stress and higher performance. This alone should be worth aiming for short cycle times.

Slide 39

Slide 39 text

Still not convinced …

Slide 40

Slide 40 text

There are even more advantages

Slide 41

Slide 41 text

Cutting off manual tasks • Short lead times require high degree of automation • Additional effects of rigorous automation • Less errors happen • People are happier • More capacity for valuable work • Get more done with the same – happier – people • Saves money, improves productivity and robustness

Slide 42

Slide 42 text

Simplifying tasks • Aiming for short cycle times ruthlessly points out waste • Makes visible if a task has become an end in itself, e.g. • Bloated requirements validation processes • Complex design rules from the ivory tower • Documentation without a reader • … • Shows the degree of autopoiesis * * In its sociological interpretation by Niklas Luhmann

Slide 43

Slide 43 text

Valueless routines, ends in themselves, bloated procedures and alike do not mix with short cycle times

Slide 44

Slide 44 text

Simplifying tasks • Focus on cycle times means focus on value • Simpler processes • Less useless procedures • Focus on purpose and value, not on routines and habits • Saves money, reduces stress, improves productivity

Slide 45

Slide 45 text

Configuration frenzy • Most applications contain lots of configuration options • Multiple execution paths • Complex decision logic • Additional configuration option management frontends • Additional configuration option storage • Additional logic to access configuration at runtime • Sometimes up to 50% of overall solution complexity

Slide 46

Slide 46 text

Configuration frenzy • Added complexity comes at a price • Code harder to read • Code harder to understand • Code harder to test • Code harder to debug • Changes becomes slower and more expensive • More bugs creep in • Robustness in operations goes down

Slide 47

Slide 47 text

Most configuration options are only needed due to long lead times

Slide 48

Slide 48 text

t Timespan available for delivering change Timespan needed by IT to deliver change Change via configuration options needed due to this time gap Change need detected Change delivery needed at the latest Earliest delivery date via software change

Slide 49

Slide 49 text

What if lead times were shorter than needed change delivery timespan?

Slide 50

Slide 50 text

t Change delivery needed at the latest Timespan available for delivering change Timespan needed by IT to deliver change Configuration options not needed for change delivery Change need detected Delivery date via software change

Slide 51

Slide 51 text

Configuration frenzy • With intraday lead times, most configurations not needed * • Saves time and money, improves robustness * You will not get rid of all configuration options, but you may not need quite a lot of them anymore

Slide 52

Slide 52 text

Project overheads • Long lead times lead to a huge backlog • Requests pile up while organization is busy processing current batch of projects • Simple queue not sufficient to manage amount of requests • Instead, more complex organization needed

Slide 53

Slide 53 text

Project overheads • Requests are bundled in projects to manage backlog • Projects must be defined • Projects must be budgeted • Projects must be prioritized and approved • Projects must be set up and staffed • Projects must be implemented, controlled and readjusted • Huge additional effort • Requires minimum project size • Often reinforcing loop

Slide 54

Slide 54 text

What if you could implement requests so fast that you do not need to manage a huge request backlog?

Slide 55

Slide 55 text

Project overheads • Intraday lead times would dramatically reduce backlog • Much simpler organization would be sufficient • Imagine a big Kanban board for all requests * • Project overheads would disappear • Regular organization sufficient to handle all requests • Business and IT owners would be more relaxed • Stable teams are possible * Usually not that simple, but still simple solutions possible

Slide 56

Slide 56 text

Project overheads • Saves money, makes you faster, reduces stress

Slide 57

Slide 57 text

No “emergency taskforces” needed • Widespread separation between “project” and “maintenance” • “Project” implements new functionality • “Project” typically is organized in long-running projects • “Maintenance” takes care of bugs and small changes • “Maintenance” usually uses a flow-style process • Most experienced engineers usually assigned to “project”

Slide 58

Slide 58 text

Project A Project B Project C Maintenance Senior Engineer Junior Engineer Ops Expert

Slide 59

Slide 59 text

Now a serious production error occurs …

Slide 60

Slide 60 text

Project A Project B Project C Maintenance Senior Engineer Junior Engineer Ops Expert Hey, we’ve got a serious production problem here!

Slide 61

Slide 61 text

Project A Project B Project C Maintenance Senior Engineer Junior Engineer Ops Expert Oh, that’s intricate. We don’t know how to fix it.

Slide 62

Slide 62 text

Thus, an “emergency task force” is set up

Slide 63

Slide 63 text

Project A Project B Project C Maintenance Senior Engineer Junior Engineer Ops Expert Emergency Task Force That’s intricate, indeed. But we can fix it. Coordination & management

Slide 64

Slide 64 text

Meanwhile …

Slide 65

Slide 65 text

Project A Project B Project C Maintenance Senior Engineer Junior Engineer Ops Expert Emergency Task Force Coordination & management Crippled (significantly reduced progress) Crippled (significantly reduced progress) Mostly standing still Mostly standing still Additional overheads

Slide 66

Slide 66 text

Often, these “emergency task forces” run for days and need to be formed on a quite regular base. This means a significant productivity loss, not to mention the stress for all people involved.

Slide 67

Slide 67 text

Now, assume an IT organization with intraday cycle times

Slide 68

Slide 68 text

Senior Engineer Junior Engineer Ops Expert Capability Team 1 Capability Team 2 Capability Team 3 Capability Team 4 Capability Team 6 Capability Team 5 Capability Team 7 Note: This is one possible organization supporting short cycle times. There are a lot more ways to implement an effective organization supporting short cycle times.

Slide 69

Slide 69 text

Now a serious production error occurs …

Slide 70

Slide 70 text

Senior Engineer Junior Engineer Ops Expert Capability Team 1 Capability Team 2 Capability Team 3 Capability Team 4 Capability Team 6 Capability Team 5 Capability Team 7 Whoa, something blew up in production!

Slide 71

Slide 71 text

Senior Engineer Junior Engineer Ops Expert Capability Team 1 Capability Team 2 Capability Team 3 Capability Team 4 Capability Team 6 Capability Team 5 Capability Team 7 Yeah, we need to look into it asap!

Slide 72

Slide 72 text

Senior Engineer Junior Engineer Ops Expert Capability Team 1 Capability Team 2 Capability Team 3 Capability Team 4 Capability Team 6 Capability Team 5 Capability Team 7 Swarming on production issue

Slide 73

Slide 73 text

Meanwhile …

Slide 74

Slide 74 text

Senior Engineer Junior Engineer Ops Expert Capability Team 1 Capability Team 2 Capability Team 3 Capability Team 4 Capability Team 6 Capability Team 5 Capability Team 7 Swarming on production issue No exceptional procedures required. Can be handled using standard process. No exceptional ad hoc organization required. Can be handle by regular organization. Rest of organization not crippled. No exceptional coordination overheads needed.

Slide 75

Slide 75 text

No “emergency taskforces” needed • With intraday lead times, production errors become “just another task” (with a high priority of course) • Saves money, reduces stress, improves productivity

Slide 76

Slide 76 text

Less worthless work • With long lead times, most code is not used * • ~20% frequently used • ~30% rarely used • ~50% not used (often disliked by users) * see, e.g., https://www.martinfowler.com/articles/xp2002.html#BuildOnlyTheFeaturesYouNeed

Slide 77

Slide 77 text

t Specify the new system!

Slide 78

Slide 78 text

t What could be needed for the new system? ? ? ? ? ? ?

Slide 79

Slide 79 text

t I think we need I want Wait. I’ll do some market research In the future, we might need Let me ask some more stakeholders

Slide 80

Slide 80 text

t Do you really have everything in it? Remember: We only have that one shot!

Slide 81

Slide 81 text

Slide 82

Slide 82 text

t Budgeting

Slide 83

Slide 83 text

t Approval

Slide 84

Slide 84 text

t Implementation

Slide 85

Slide 85 text

t Deployment

Slide 86

Slide 86 text

t That’s not what I need!

Slide 87

Slide 87 text

Now, assume an organization with intraday cycle times

Slide 88

Slide 88 text

< 1d

Slide 89

Slide 89 text

< 1d

Slide 90

Slide 90 text

Less worthless work • Intraday lead times enable fast feedback from users • Allows identifying and cutting off non-successful features • Allows creating more value with a smaller team • Enables outsmarting demographic change • Saves money, makes you faster, increases revenue

Slide 91

Slide 91 text

More flexibility • Companies often aim for better “business agility” without understanding the effects of digital transformation • Business and IT have become inseparable • Business and IT are the same side of the same coin – the other side is the market • IT delimits speed and thus flexibility of your business • Intraday lead times allow for great “business agility” • Improves flexibility, enables increased revenue

Slide 92

Slide 92 text

Do you still think you do not need it?

Slide 93

Slide 93 text

Is there really a different path worth following?

Slide 94

Slide 94 text

Hmm … I’ll think about it

Slide 95

Slide 95 text

Summing up

Slide 96

Slide 96 text

Effects of going fast 1/2 • Essential for highly dynamic, post-industrial markets • Significant cost savings • Simpler, more effective organizations • Higher productivity • Increased robustness of the IT landscape

Slide 97

Slide 97 text

Effects of going fast 2/2 • Higher employee satisfaction (with all its positive effects) • Increased company resilience • Improved business agility • Higher user/customer satisfaction if played right • Relief for the omnipresent shortage of IT experts • …

Slide 98

Slide 98 text

There is no sensible alternative to going fast

Slide 99

Slide 99 text

What are you waiting for?

Slide 100

Slide 100 text

Uwe Friedrichsen CTO @ codecentric https://twitter.com/ufried https://www.speakerdeck.com/ufried https://ufried.com/blog/value_of_speed/