Developer Experience Values

Values are strategies to help us meet our needs. As a DevEx team, choosing values that help meet stakeholders’ needs also helps us meet our own.

While each stakeholder will phrase their needs differently, we can leverage a framework to ensure we don’t miss any during the improvement cycle or adoption curve.

Product Mindset (AirBnB, Google[dora], IBM, Snowflake, LinkedIn) Operates in incremental experimentation, learning, and improvement cycle. Outcome Orientation: Measuring success by the impact on users and business goals, rather than just by project completion or output Persistent model: A mental model of the relationships between high level inputs and outcomes. Used for developing hypotheses and strategies to influence desired outcomes, then to evaluate those strategies’ effectiveness. DORA called these Structural Equation Models. Data-driven prioritization and decision making Utilizing user feedback, telemetry, and data analysis to prioritize improvements by expected input change (according to the model). Cross-functional collaboration (including users): In traditional product mindset, includes platform teams. Successful DevEx programs (AirBnB, Google[dora], IBM, Snowflake, LinkedIn), include users too. User-centric: Focused on understanding and prioritizing user needs, not specific tools or features. Teams prioritizing user needs achieve 40% higher organizational performance DORA 2023 In a DevEx team, we have to consider many “users” needs. Company Leaders, Product and Engineering Leaders, Product managers, Platform and Product team leads, product managers, and engineers use our data, reports, dashboards, surveys, tools, and processes. Linkedin (Champions), AirBnB, and Snowflake (influencer volunteer). Champions, CAP, (AirBnB version). Why? It doesn’t directly reduce friction, increase flow, hasten feedback loops, or reduce cognitive load. Needs: Predictability is the human need to optimally know the relationships among things and events in the world: what causes what, follows what, and belongs with what. It builds trust and a sense of control. Competence is the human need to gain skills that change outcomes in the world. Competence builds a sense of control and sense of worth. Acceptance is the human need for positive social interactions. It builds trust and a sense of worth. Profit is a company need - profit for companies is like air for humans.

  Needing others makes us vulnerable.
  > Trust is choosing to make something important to you vulnerable to the actions of someone else. - Brené Brown

An effective improvement and adoption loop rely on trust. Trust comes from two things: Predictability and Acceptance. Acceptance arises from synchrony and reciprocity. Value1: Trust

Counterexample - Speed: Doesn’t im

Improvement and adoption struggle without trust. EXAMPLES.

Trust comes from

First we’ll use an example to highlight gaps, look at some companies who fill that gap, suggest a collection of values that eliminate gaps based on those companies practices. (linkedin, AirBnB, Snowflake).

Example Value with Gaps - Speed

stakeholder need value cycle phases that value meets need in 0,1,2,3,4

Product Engineers, Continuous Improvement Cycle - which needs are met by which values at which stage? Predictability Listening: Listening to product engineers’ pain points has no direct impact on product engineers’ predictability. Prioritizing: Prioritizing faster does not improve predictability or competence. Improving: Improvements may increase predictability, but improving faster by itself does not. Reporting: Reporting faster does not increase predictability by itself. Competence: Listening: Listening to product engineers’ pain points has no direct impact on product engineers’ predictability. Prioritizing: Prioritizing faster does not improve predictability or competence. Improving: Improvements may increase predictability, but improving faster by itself does not. Reporting: Reporting faster does not increase predictability by itself. Acceptance

Speed
    Listening to product engineers' pain points has no direct impact on product engineers' predictability.
    The process of listening has no direct impact on predictability.
    Listening faster, may yield tech predictability solutions faster, which may be prioritized, which may be improved. 1/2/2/2
    Listening faster, may yield interaction predictability solutions faster, which may be prioritized, which may be improved. 1/2/2/2

    Listening, to speed solutions has no impact on predictability, except maybe a 1/5 effect on faster responses having less time for information in them to change.
    What is listened to may impact predictability.
  Prioritizing (& feedback): Prioritizing speed has no impact on predictability.
  Improving (& feedback): Improving speed has no impact on predictability (though certain improvements may)
  Reporting (& feedback): Improving speed has no impact on predictability (though certain improvements may)

    We send out a survey to understand developers' experience. On submission, developers receive "your survey was submitted". They also get a message that they can expect results for their team in a week or so. A week later, they receive results for their team.
    The analysis shows a problem among teams with similar workflows - slow local builds, and team-specific metrics (code reviewer response times are significantly below average for teams with similar workflows).

  Prioritizing (& feedback)
    No immediate impact on speed.
  Improving (& feedback)

  Reporting (& feedback)   Competence   Acceptance   Profit   Listening (to product engineers).
Listening temporarily slows them down.

They decide to set a team review time SLA, and implement nudges for reviews that approach it.
Relationships between their survey and their work: 1/2
Skills gained: TBD - no action taken yet.

They decide as a team
Relationships possible
Relationships between their survey and their work: 2
At this stage cognitive load
Relationships between events understood: 0
Skills gained: 0
Cognitive Load reductions: -1
Speed improved: -1

Skills gained
Flow states meet...
Feedback loops in listening are "your survey was submitted". No skills gained. No relationships between that and their work. No positive interactions yet. Competence That meets the meet...
Predictability (no)
  Feedback loops

Competence starts being met when developers communicate opportunities to improve.  and feel heard. , feel heard, and predictably see results. can improve their skills and performance.
Acceptance
Profit

Prioritizing (with product engineers) Improving (with product engineers) Reporting (to product engineers)

Product Engineers
Platform Engineers
Leaders
Product
Company

Prioritizing Improving Reporting

In the adoption curve, which values meet which needs for which stakeholders at which stage? Early Adopters Early Adoption

At the core of a Developer Experience program is cyclical continuous improvement process. The core of an effective improvement process is trust.

Trust is choosing to make something important to you vulnerable to the actions of someone else. - Brené Brown

As stakeholders in engineering contexts, we need to trust each other to understand and meet our needs. A developer experience team facilitates that trust, improving both stakeholder and business outcomes in the process.

What values facilitate stakeholder outcomes during a developer experience cycle?

What problems to solve? detractors/laggers

PM Skills Groups: Early Adopters

Detractors/Laggers

Transparency Closure Responsive Predictable Reliable

Needs Framework Detail Optimal Predictability: is: the need to optimally know the relationships among things and events in the world: what causes what, follows what, and belongs with what builds: Trust, Sense of Control requires: Things and events. Stable relations between them. Time and mental space to learn those relations.

Competence: is: the need to gain skills that change outcomes in the world builds: Sense of Control, Sense of Worth requires: Clear understanding of desired outcomes Opportunities and time to acquire and practice skills that achieve those outcomes.

Acceptance/Connection: is: the need for positive social interactions builds: Trust, Sense of Worth requires: Synchrony (i.e. clicking): spontaneous rhythmic coordination of actions, emotions, thoughts, and physiological processes across time between two or more individuals Actions (how: response rhythm, attention) failing to close loops breaks attention synchrony Emotions (how: empathy, emotion validation) Thoughts (how:staying on current conversation topic with active listening, perspectives validation, and avoiding judgement, advice, opinions, interruptions) Mutual Purpose: Clear shared understanding of desired outcomes Physiological Processes: We cannot practically control these, and they don’t apply in async communication Reciprocity (actions): Equivalent what: Will you reciprocate equivalent value? Equivalent when: Will you reciprocate when I expect? Infants as young as 2 to 3 weeks of age recognize when mothers violate the reciprocity of a social interaction, as when they temporarily become unresponsive. This doesn’t change. We feel it when people don’t respond in a similar cadence. Equivalent how: Will you reciprocate how I expect? Equivalent where: Will you reciprocate in the context I expect? Equivalent whom: Will you reciprocate to whom I expect (e.g., to me, vs to the company)?

  Directed Graph of actions: Trust comes from previous direct (2-edge cycles) and indirect/generalized (n-edge cycles) actions towards me. My actions depend on personal utility (including meeting own, other, or group expectations).

  [A unified framework of direct and indirect reciprocity](https://doi.org/10.1038/s41562-021-01114-8)
  [Social exchange theory: Systematic review and future directions](https://doi.org/10.3389/fpsyg.2022.1015921)

  Reciprocal: equivalent exchanges
    Direct when engaging directly (e.g., volunteers, stakeholders)

Needs like water and connection are universal. Values are individual. Values are strategies that we believe will meet our needs. Product Mindset (AirBnB, IBM, Snowflake, ) Connection (AirBnb, Snowflake, ) Experimentation Feedback loops

How do these play out in improvement? Listening Prioritizing Improving Reporting

How do these play out in adoption?

Questionable Values: Efficiency/Productivity: what needs does this meet, for whom?

Efficiency Low cognitive load (to make mental space for new skills) Feedback Loops (that are timely and accurate, to learn from experiments and practice) Flow States (to experiment with, practice and learn skills efficiently and effectively) Low cognitive load (to make mental space for new events, things, and relations) Feedback loops (that are efficient and accurate, to learn from observing new things and events) Flow States (to mentally integrate the new relations)

AirBnB Developer-centric (Customer focus) Restructuring teams around developer personas rather than tools, and formalizing stakeholder engagement through regular feedback loops, surveys, and advisory boards. Pulling in some of the higher level ICs within our company, within engineering, who had opinions and suggestions, making them partners in this journey. So rather than us trying to convince and report, having them partner with us And really, of course, deep understanding of the full product life cycle, what it takes, having the product mindset from the start, understanding the requirements and so on. And also the full life cycle to actually getting software out of the door and rolling things out and what it means then after the product launched, et cetera. So kind of having a holistic understanding and experience with best practices and all this good stuff. So I think that’s as far as the qualities that are necessary for this role. And of course, because you are in the middle of everything, then communication skills, storytelling, working with stakeholders and partners, that becomes even more important. Common cases easy, uncommon possible Laser focus Deliver incrementally OneMedical Developer Centric (Customer Focus) Snowflake DevProd and Operational Excellence Trust Product Mindset (Collab between Engineering and Product, includes Customer Focus) Closed loops Customer Involvement (influencer volunteers to both help and demo the improvements, and share context) Linkedin Closed survey loops CarGurus IBM we kind of have a product mindset as much as possible where we go out and we talk to both individual developers and spend time with them and understand kind of their friction points. Personas: And then we can kind of like break out personas in some of this stuff. So like one of the ones that we go to a lot is what’s a new hire persona.

Adoption Curve: Collaborations: Persona Champion: linkedin Cargurus: Champions Local Volunteers: snowflake

Partners: AirBnB

Improvement Cycle: Listening (transparent, empathetic, respectful, non-judgemental, prompt)

Predictability requires from DevEx:
  Transparency: When, where, why, how we listen, and to whom?
  Reliability: Responding promptly and closing the loop.
  Feeling heard (snowflake)
  Feeling understood (snowflake)
  Commitment to help

Acceptance requires from DevEx:
  For close relationships with leaders, high-influence early adopters and detractors
    Feeling heard
    Feeling understood
    Empathy: Experience the pain with that person (why DevEx PMs need to be technical)
    Non-judgement
    Respect
  Understanding: the pain points preventing competence
  Note that this is more important for close interactions since they fall under direct reciprocity more than indirect.
Transparency:
  As devs, we cannot control or at least predict our context unknowns, which inhibits all their needs.
Reliability
Without reliability:
Predictability requires from DevEx:

  To accurately anticipate a future thing.
  This also covers transparency, since we can't predict what we don't know.
  And do those things reliably.
  We cannot predict what we do not understand, so predictability includes certainty.
Certainty:
  To accurately understand a current thing without doubt.
  Communicate When, where, why, how we listen, and to whom?
  If we're unclear about these, we won't trust
  To minimize uncertainty we must respond promptly and close the loop.
Feeling certain

Feeling heard (Empathy)
  Experience the pain with that person (why DevEx PMs need to be technical)
  Non-judgement
  Respect
  Understand
Responsive
  Promptly
  Closing the loop
Maximize feeling heard

Prioritizing (transparent, holistic, collaborative, respectful, unbiased) Clear priorities Close the loop on all listening transparency around which activities align with which priorities and how that alignment is determined prioritization around strategies then metrics vs metrics then strategies

Encourage collaboration, empower devs who want to decrease their pain.

Implementing (transparent, collaborative, incremental, reliable) Follow through on the prioritized items. Close the loop on all prioritized items. Implementing measurement or improvements Encourage collaboration, empower devs who want to decrease their pain. Snowflake’s Customer Advisory Panel (CAP) Customers (Devs) embedded in team efforts to make things better

Analyzing and Reporting Results (transparent, accurate, precise, unbiased, safe, relevant, actionable)

Trustworthiness

Trust is choosing to make something important to you vulnerable to the actions of someone else. - Brene Brown The DevEx cycle requires good stewardship of others’ trust. ICs and Leaders need to trust those responsible for DevEx to have their best interests at heart. Trust values are broken down by DevEx cycle phase. They synthesize DORA-highlighted cultural attributes (high-trust, low-blame, high-learning, experimentational), types of trust, and various other sources relevant to that step. In listening (transparent, empathetic, respectful, non-judgemental, prompt) Maximize feeling heard, maximize useful feedback, maximize buy-in In prioritizing and planning (transparent, collaborative, respectful, objective) Encourage collaboration, empower devs who want to decrease their pain. In implementing (transparent, collaborative, incremental) In measuring and reporting (transparent, accurate, safe, relevant, actionable) Experimentation Fosters innovation, data-driven decision-making, and reduces the risk of large-scale failures by testing new ideas in a controlled environment. Developer Workflow Focus Aligning platform efforts on holistic developer workflows instead of individual tools/technologies helps maintain a big-picture view, prevents conflicting team goals, prevents silos that resist change, and prevents optimizing a tool at the expense of larger workflow gains.

Written on June 30, 2025
Tagged: