DEV Community: bffjossy

DevRel Collective 2021 Salary Survey - Introduction

bffjossy — Mon, 27 Sep 2021 23:13:33 +0000

The DevRel Collective (DRC) has been running a survey over the past few quarters regarding the salaries and benefits paid to DevRel professionals around the globe. The DRC exists to facilitate sharing information, resources, and encouragement among the professional Developer Advocates, Evangelists, and Technical Community Manager community with ~2000 members and counting.

In this report we take a look at some of the results from the survey with a view to summarising the key themes.

The purpose of the survey

A challenge faced by many practitioners has been a relative scarcity of reference materials pertaining to career growth and salary trajectories. The sometimes ambiguous nature of titling across companies, the recent growth of DevRel as a formalized profession, and the frequently small size of teams, along with the variation of where DevRel “sits” within an organization all contribute to a lack of clarity around compensation.

The survey is intended to offer insight into the state of compensation in DevRel. We examine roles including and adjacent to “developer advocate”, “evangelist” and “technical community manager.” Our twofold goal in its creation was to: help individual contributors (non-managers) ensure that their salary is comparable for their geographic region and experience level; and assist hiring managers as they calculate appropriate budgets for headcount.

It is our hope that this report can function as a balanced resource to both employees and managers to arrive at an understanding around norms.

Survey design

We spent a great deal of care and time on survey design. The full survey questions have been added as an appendix.

Some of the questions are qualitative and some are quantitative. That said, it was perhaps inevitable that the responses occasionally begged additional questions which were beyond the scope of the survey. We discuss why those choices were made in the requisite sections.

Areas of inquiry included:

Cash salary
Benefits and perquisites (also known as “perqs” or “perks”)
Total compensation
Country respondent is based out of
Sentiment on compensation
Job titles
Length of tenure in industry and time at current position
Career level and manager/non-manager status
Ethnicity
Gender
Age
Sexual orientation

The survey topics are of a sensitive nature. It was therefore important to preserve anonymity of, and restrict access to, the raw data. The raw data is viewable only by the admin team and data scientist. We only use it to create visualizations that will show (among other things) average compensation for specific geographic areas.

We also spent a lot of time digging into the questions themselves to identify the real underlying questions. For example, some questions are very cut and dry such as: what country do you live in? While others are sentiment based: what’s the cost of living in your area? The reasons for that are cost of living (CoL) can be relative, and also some regions vary widely in pricing over a small geographic area. So the real inquiry is not about the cost of typical rent in your area, it’s about the lived experience.

For every question we added or deleted, we considered what story its data could tell in combination with other fields.

A note on the data

This sample is small (~150 replies,) and highly skewed to the US (~80 replies). It is therefore difficult to draw general conclusions from this data that can be applied to the wider DevRel community, and care should be taken when doing so.

It is our hope that future iterations of the survey will be more geographically diverse across non-US markets. Please see the location sections if you would like to learn more.

The survey looks at data collected in October-December 2020. Some later responses trickled in after that time (11 in total). The survey link was promoted in the DevRel Collective Slack, and also posted to Twitter, LinkedIn and dev.to as well as several industry newsletters.

There were some areas where it was important to understand the potential impact of marginalization but were hindered by a small sample set. We sometimes made the decision to report on the dominant group versus the rest of the responses. After some consideration we decided it was preferable to show what we were able. This is in no way meant to flatten the experiences of people, create an improper hierarchy, or “other” anyone. It was the best we could do to bring back some insights without de-anonymizing the data.

For the results below, we have worked to summarize what we have from the survey responses only. To validate any of these findings more generally would require more focused work, which we may spend additional time on in future iterations.

It’s important to remember that the lumping together of disparate groups into the “non-X” category can be misunderstood as non-X == Y. Actually non-X is a catchall category. In other words, what we learn is mostly about the majority group. We’re hoping this will change in future interactions. In the meantime, this data can offer some amount of insight to people interested in equity. This is important because a key factor of marginalization is that members tend to have less access to information available to them versus the information shared across in-groups.

A reminder about the models contained in this document: please bear in mind that these are simple models, and you should use them to augment your own experience. The sample is small. We would want to look at more robust analyses before we make any sweeping claims for a specific individual.

Processing

Some steps had to be taken to make the data usable. Due to the small sample size, it wasn’t possible to process each country in its own currency. Given that the majority of the survey responses came from the US, recent currency exchange rates were used to convert the `Salary` responses into US Dollars.

We filtered one or two unrealistic replies (over $700,000) to prevent skewing the results – these seem unlikely to be real responses (although well done to whoever that is if they are real!)

Before diving in: the larger conversation

As mentioned above, our goal for this survey is to get information to people on both sides of the table to help move forward productively with conversations grounded in data. To that end, we aren’t selling anything: no products, memberships, or courses. We don’t charge to view the report or require signing up for a mailing list to view our results.

If you would like to offer some kind of support, here are some ideas:

The best way to do so is to help support is by simply getting the word out: Polywork, Twitter, LinkedIn are great places to start
Social media is great but so is old-fashioned word of mouth (for example at the hallway track of your next event)
Please let your colleagues and managers know about it
Finally, if you have contacts at any of the industry “salary database” resources and can facilitate introductions we would love to speak with them about how to get DevRel represented in meaningful ways. This would also help streamline conversations for managers who currently resort to plugging in (commonly conflicting) values of engineering and marketing at review time

The more people who know about this survey, the richer the results will become with every iteration! We can be reached at team@devrelcollective.fun

This request is completely optional, of course. If you choose to engage privately with the results, that’s totally fine — either way, we are happy to have been a resource in your career journey.

SPECIAL THANKS TO OUR DATA SCIENTIST, GREG SUTCLIFFE

Greg Sutcliffe has been working in community management for a decade, and is currently the Principal Data Scientist for the Ansible Community. He's interested in how appropriate use of data can inform the development and governance of communities, especially with regards to open source projects. He also likes cooking.

DRC Survey: Appendix B - DevRel Salary Survey questions

bffjossy — Mon, 27 Sep 2021 23:13:14 +0000

2021 DevRel Salary Survey

(originally deployed as an Airtable form)

DevRel Salary Survey

The goal of this survey is to provide industry data about the standard salary ranges for DevRel professionals, from Technical Community Managers to Developer Advocates, Developer Experience Professionals, DevRel Managers, and everything in between.
It’s our goal to help individual contributors (non-managers) ensure that their salary is comparable for their geographic region and experience level, as well as assist hiring managers as they calculate appropriate budgets for headcount.

All of the information collected here will be anonymized, analyzed, and then shared in a manner which keeps your specific information private. If you have any questions, feel free to reach out to the DevRel Collective admin team at team@devrelcollective.fun
Don’t overthink your answers. If you need to estimate, that’s fine!

What’s Your Base Salary? *
Please only include your gross annual salary in whole numbers – we’ll ask about the rest of your total compensation later.

What currency is your salary in? *
We pulled the top 30 currencies according to exchange-rates.org. Please email team@devrelcollective.fun if your currency is missing.

What country are you based in? *

Do you receive a bonus every year? *
Yes / No

Do you have RSUs (or something like them)? *
RSUs give an employee interest in company stock but they have no tangible value until vesting is complete. Upon vesting, they are considered income, and a portion of the shares is withheld to pay income taxes. (Investopedia)YesNo

Do you have stock options (or something like them)? *
Options are financial instruments that are derivatives based on the value of underlying securities such as stocks. An options contract offers the buyer the opportunity to buy or sell—depending on the type of contract they hold—the underlying asset. (Investopedia)
Yes / No

Do you own shares in the company (or something like them)?
Restricted shares are awarded outright, and their owner has the same rights and privileges as any shareholder. However, the shares may be vested, and the company may reserve the right to buy back unvested shares if the employee leaves the company. (Investopedia)
Yes / No

Is your company privately owned or publicly traded? *
Private/Public

How many employees does your company have? *
1-20 / 21-50 / 51-100 / 101-300 / 301-600 / 601-1,000 / 1,001-3,000 / 3,001-20,000 / 20,001+

How’s the cost of living in your area? *
Please evaluate based on your local state/province or country.
Low cost of living / Medium cost of living / High cost of living

Do you feel like you are fairly compensated for your job? *
Extremely Undercompensated / Somewhat Undercompensated / Mildly Undercompensated / Neutral–Fairly Compensated / Mildly Well Compensated / Somewhat Well Compensated / Extremely Well Compensated

What is your job title? *

Is coding required as a part of your role? *
Yes / No

Are you a people manager? *
Do you have the ability to hire and fire?
Yes / No

Are you responsible for strategic decisions? *
Yes / No

What level are you at in your career? *
Entry-level / Mid-level / Senior-level

What type of employee are you? *
Full time / Part time / Contractor / Volunteer

Is this an update to a job you’ve already reported in this salary survey? *
Yes / No / Unsure

How many years have you been at this company? *
If it’s a brand new job, please put 0.

How many years have you been in Developer Relations? *
We recognize DevRel is a relatively new title. Feel free to list how long you’ve been in DevRel-like positions.

What technical communities are you focused on in your current role? *
Please select all applicable options.
– DevOps
– Open Source
– Databases
– SaaS
– Cloud
– API
– Security / SecOps
– Blockchain
– Healthcare
– Developer Tooling
– FinTech
– IoT
– Gig Economy
– Internal DevRel
– Embedded Systems
– Game Development
– AI/ML
– Data Science
– Mobile
– Web

What is your closest major airport code? *
Be patient – this list takes a moment to load 😊

Are you required to have a presence at the office? *
(pre or post-COVID answers, please)
Remote / In-office / Occasional visits (e.g. monthly or quarterly)

How would you describe your ethnicity? *
US-centric, feel free to choose “prefer not to say” or choose as many as are applicable.
– American Indian / Alaska Native
– Black and/or African American
– Hispanic, Latinx, or Spanish Origin
– White
– Asian
– Native Hawaiian or Other Pacific Islander
– Middle Eastern or North AfricanAfrican
– Prefer not to say

What is your age range? *
Under 18 / 18-24 / 25-34 / 35-44 / 45-54 / 55-64 / 65+ Prefer not to say

How would you describe yourself? *
– Female/Woman
– Male/Man
– Genderqueer or Gender non-conforming
– Non-binaryIntersex
– Prefer not to say
– Other (custom text)

Do you identify as transgender or genderqueer? *
Yes / No / Prefer not to say

What is your sexual orientation?*
– Lesbian
– Gay
– Bisexual
– Straight
– Queer
– Pansexual
– Asexual
– Prefer not to say

SPECIAL THANKS TO OUR DATA SCIENTIST, GREG SUTCLIFFE

Greg Sutcliffe has been working in community management for a decade, and is currently the Principal Data Scientist for the Ansible Community. He's interested in how appropriate use of data can inform the development and governance of communities, especially with regards to open source projects. He also likes cooking.

DRC Survey: Appendix A - Other plots (company size, trading status, and regional spread)

bffjossy — Mon, 27 Sep 2021 23:13:09 +0000

Other plots

Here’s a few other things we investigated that didn’t turn up much to learn from in the themes above, but might have some value to some readers:

Size of Company

As you can see, this plot is all over the place. If anything, it speaks to the value of finding the right role rather than optimizing for a particular company. There is a slight trend of higher compensation at larger companies, but we can also see that it’s possible to make a healthy living across many levels.

Public/Private Company

For the purposes of this report,”public” refers to companies which are publicly traded, making the result fairly unsurprising – bigger companies go to IPO; bigger companies pay slightly more; so that result shows up here too.

Top Five Airports

Many practitioners of DevRel are high-mileage travelers, making their “home base” airport a quality-of-life factor. Respondents answered the question “What is your closest major airport code?” as follows:

SFO
AUS and LHR (tie)
SEA and JFK (tie)
DEN, TXL and YYZ (tie)
PDX, MSP, BER (tie)

Given its proximity to Silicon Valley and status as an international airport, it’s unsurprising that SFO is the most-flown-from airport. After that, the results quickly filter across a wide geographic area.

SPECIAL THANKS TO OUR DATA SCIENTIST, GREG SUTCLIFFE

Greg Sutcliffe has been working in community management for a decade, and is currently the Principal Data Scientist for the Ansible Community. He's interested in how appropriate use of data can inform the development and governance of communities, especially with regards to open source projects. He also likes cooking.

DRC Survey IX: DevRel Salary Takeaways, 2021

bffjossy — Mon, 27 Sep 2021 23:13:03 +0000

Conclusions

Looking over the plots above, the take-away is mostly things you’d already expect:

Know what you’re worth (which hopefully the plots here give you a better handle on)
If you’re a member of a marginalized group, don’t limit your research to people who are just like you; seek out information about the highest pay overall
Make sure the role matches your unique skillset
Keep moving on if you really want to climb the “*salary* ladder.”

We call out salary specifically because while we acknowledge that total compensation includes more than just your paycheck, we found it near impossible to analyse elements like stock options, RSUs, etc, since they have variable value. It’s also important to consider your non-monetary quality of life, such as remote working, flexible hours, an excellent team or manager, and how happy you are in your current role. As ever, use your judgement!

Future survey rounds

As the data began to come together, we discovered there were additional questions we would have liked to ask. For example, “are you happy in your job?” would have made for an interesting lens. In the next iteration, we also plan to ask what part of the organization DevRel reports to.

It would also be interesting to look at parental status along with genders and career progression. We plan to ask about future plans, such as where the respondent sees themself in five years. And we are always seeking to understand better the ways in which job title could possibly play into salary.

Are there other questions you’d like us to ask or results you’d like to see? Be sure to let us know by emailing team@devrelcollective.fun.

Wrapping up

That wraps up our first take. These results suggest that we’re doing the right thing with this survey – there seems to be confusion around appropriate pay or compensation, and perhaps even duties, for DevRel folks. This confusion perpetuates unhealthy expectations, and if these results can help to break that cycle, we believe that may lead to positive change.

That said, the sample is small. We would want to look at more robust analyses before we make any sweeping claims for a specific individual.

SPECIAL THANKS TO OUR DATA SCIENTIST, GREG SUTCLIFFE

Greg Sutcliffe has been working in community management for a decade, and is currently the Principal Data Scientist for the Ansible Community. He's interested in how appropriate use of data can inform the development and governance of communities, especially with regards to open source projects. He also likes cooking.

DRC Survey VII: DevRel Coders, DevRel Managers, and Salary

bffjossy — Mon, 27 Sep 2021 23:12:58 +0000

Knowing what you’re there for

In addition to career progression, the nature of the job is relevant too. There were two questions asked that uncovered interesting findings regarding the expectations of the company and/or the employee. These were “Does your job require coding?” and “Do you manage others (i.e. hiring/firing power)?”.

Here’s a look at how salary varies across these two dimensions:

Let’s step back and make some sense of this. We’ve got 4 scatter plots, one for each cell of the 2×2 grid for “Coding” vs “Manager”. What are some of the basic conclusions we can pull from this data?

People who code are typically paid more than those who don’t
Likewise, managers are paid more
Interestingly, those who do manage but don’t code are typically paid more than those who do both

It seems that trying to be “all things to all people” is counterproductive.

For the quadrant where coding is required for a hiring/firing role, perhaps we’re looking at someone who works for a company that doesn’t know what they need, advertising a “kitchen-sink” role. Alternatively, perhaps we have a new hire who isn’t sure what they’re worth and agrees to a lower price despite being highly skilled.

Our hypothesis based on this data is that knowing your skill-set and where you can contribute the most value is an important part of being able to ask for (and receive) a higher salary.

SPECIAL THANKS TO OUR DATA SCIENTIST, GREG SUTCLIFFE

Greg Sutcliffe has been working in community management for a decade, and is currently the Principal Data Scientist for the Ansible Community. He's interested in how appropriate use of data can inform the development and governance of communities, especially with regards to open source projects. He also likes cooking.

DRC Survey VII: DevRel Salaries and Sexual Orientation

bffjossy — Mon, 27 Sep 2021 23:12:43 +0000

Sexual orientation

For the question What is your sexual orientation? Available options were:

Lesbian
Gay
Bisexual
Straight
Queer
Pansexual
Asexual
Prefer not to say.

The top three answers in descending order were:

Straight
Prefer not to Say
Bisexual.

Five respondents selected multiple options, and the charts in this section should be read with that consideration.

Sexual orientation and representation

Sexual orientation and US salaries

There is a difference of about 11% between Straight-only responses and all others combined, but caution should be taken to consider the sample size. As a result, we’re not drawing further conclusions at this time, but it’s an area we’d like to explore in the future with additional data points.

US-based manager salaries by sexual orientation

For US-based respondents who are people managers, responses showed a difference of 21% between those who marked only “Straight” versus all others combined.

	Straight-only, US-based	Everyone else, US-based
	MEDIAN SALARIES
People manager	$170,224	$140,000
Strategic manager	$161,000	$150,000

SPECIAL THANKS TO OUR DATA SCIENTIST, GREG SUTCLIFFE

Greg Sutcliffe has been working in community management for a decade, and is currently the Principal Data Scientist for the Ansible Community. He's interested in how appropriate use of data can inform the development and governance of communities, especially with regards to open source projects. He also likes cooking.

DRC Survey VI: DevRel Salaries and Ethnicity

bffjossy — Mon, 27 Sep 2021 23:12:35 +0000

Ethnicity

The survey’s demographic questions regarding ethnicity are US-centric. We acknowledge that’s not ideal, but it gave us a baseline to start with. For this and other reasons, the ethnicity section was optional, with “prefer not to say” as an option. This is to respect that the locus of identity is within the individual.

In response to the multi-choice question “How would you describe your ethnicity?” 7.4% declined to state; 70.3% answered White; 9.4% Asian; 3.4% Black and/or African American; 2.7% Hispanic, Latinx or Spanish origin, 2.0% Middle Eastern or North African (MENA), and 1.3% African. Multiracial responses account for 3.4% combined; all of the multiracial responses included White as one of the ethnicities.

Race and ethnicity are not synonymous. For example, it’s possible to be ethnically Latinx, but Latinx is not a race⁴. While understanding that the measures currently in use are imperfect (and that the choices themselves are US-centric as noted previously) we wanted to give people options that were meaningful to them on their own terms.

Regarding the “Asian” category, Asians and Asian Americans have been shown to have drastic variation in income when national origin is factored in⁵. Asians simultaneously must battle the harmful model minority myth⁶. Because of these issues, we initially planned to disaggregate the data but due to the small sample size, that was not possible.

Other studies have shown that Asians and Asian-Americans are underrepresented at the management level in the tech industry, with a 2017 Harvard Business Review report finding that despite making up 12.5% of the professional workforce (and 5.6% of the overall US population) Asians were the group least likely to be promoted to management⁷. We would have liked to see if this all carries through to our specialization, but the sample size was too small for any meaningful conclusions at this time.

⁴ For more on this topic, see:
— “Being Asian and Latino”, NPR
— “Afro-Latino: A deeply rooted identity among U.S. Hispanics”, Pew Research Center
— “The Many Dimensions of Hispanic Racial Identity”, Pew Research Center

For more about AAPI workplace issues including salary and promotion data, visit
— ⁵ ”Why Disaggregate? Disparities in AAPI Income”, AAPI Data
— ⁶ ”How the Model Minority Myth of Asian Americans Hurts Us All”, TIME Magazine
— ⁷ "Asian Americans Are the Least Likely Group in the US to Be Promoted to Management”, Harvard Business Review

Ethnicity models

To split the data, we first note that (unlike Gender) there are some multiple-replies (5 total) which have both “White” and “something else” selected. We’ve put replies that had only “White” selected into one group (104 replies) and all the rest into the second group (44). This is clearly not ideal, for a variety of reasons. We hope to have an increased number of responses in the future to allow more nuanced reporting.

The data we collected shows the gender pay gap remains around 7% with ethnicity included.

As for the ethnicity pay gap, it’s there but smaller than what we see for gender :

That’s around a 4% increase for the just-white group over the other group (with a 90% credible interval between 12% and -4%), which is a much less certain result. This is due to the fact that more of the “mass” of the graph hangs around 0%, so our data is less sure of there being any real difference. Anecdotally we know that the effect is real though, so we can conclude within reason that we don’t have enough data points to show the difference, especially since we’re lumping a lot of different ethnicities together, which isn’t ideal.

Asians in management

When asking whether people were managers, we realized management can take different forms. So we checked for both “Strategic Manager” and “People Manager.”

In response to the question “How would you describe your ethnicity?” we only have 16 replies from people who checked “Asian” (including multi-racial,) and only 2 of those have “People Manager” as Yes. However, we can still try to estimate the effect, even with a small sample size:

Model predictions of ethnicity effect on likelihood-of-manager-level

A note about this plot: Please bear in mind that we are discussing probability, so there are no units. (Probability is unit-less, and ranges from 0 to 1.) What’s relevant is that the scale is linear, so a point twice as high is twice as likely. The y-axis represents the likelihood of a given difference between the groups, on a linear scale. We’ve provided the visual representation to help understand the uncertainty in a more nuanced way.

Given we have a very small sample from Asian respondents as a subgroup in the data, the hypothetical model isn’t very sure, but it does back up the statement above – people of Asian ethnicities are around 15% less likely to be at manager level, based on our data.

More on ethnicity

We hoped to provide more insights around more ethnicities. Unfortunately, due to the small sample sizes involved that was not possible in this round. We did the best we could at present, by discussing the two largest groups we have data for. We are acutely aware of the existence of issues facing underrepresented and marginalized ethnicities in the tech industry, and we’re striving to improve this for the next round.

If you’re interested in helping us collect responses from a wider representation of humanity in our next survey, please reach out to team@devrelcollective.fun. We would gratefully welcome the assistance!

SPECIAL THANKS TO OUR DATA SCIENTIST, GREG SUTCLIFFE

Greg Sutcliffe has been working in community management for a decade, and is currently the Principal Data Scientist for the Ansible Community. He's interested in how appropriate use of data can inform the development and governance of communities, especially with regards to open source projects. He also likes cooking.

DRC Survey V: DevRel Salaries and Gender

bffjossy — Mon, 27 Sep 2021 23:12:26 +0000

Gender

Female/Woman
Male/Man
Genderqueer or Gender non-conforming
Non-binary
Intersex
Prefer not to say

As before, let’s start by talking about processing. The data is a small set: ~150 responses total. The question we asked was a multi-select answer, which means people could choose more than one option. However, some things stand out:

Two-thirds of the responses describe themselves as “Male/Man.”
Although multi-select was possible, none of the men selected additional checkboxes
17 people selected more than one option²

We wish we had enough data to treat this with the nuance it deserves, but we simply don’t. The goal is to look at the gender pay gap, while preserving the anonymity of specific respondents. Therefore, we’ve chosen to make the following split: Those who identify as “Male” as one group, and those who picked any other option as the second group. We acknowledge that this leaves a lot to be desired, especially with the erasure and hardships gender-minority people continue to battle. We hope to rectify this in future versions of the survey.

² an example of multiselect would be “Female/Woman” + “Genderqueer/Gender non-conforming”

Modeling salary by gender

As discussed above, there’s a lot of variance in our data, and we want to make sure we can account for it. Notably: Job Bucket; Age; Years Experience; Years in Company; and Country are of the most importance. Having created a variety of models from these columns, we seem to see a robust result of an approximate 7% increase for men over the other group (with a 90% credible interval: 14% to -1%). This is a small difference, equating to around $10,000 at the median:

Monetary values are not reflected in the y-axis
because this prediction is about probability³

³ Imagine you plot the raw data and fit a line to it – you could read off the slope of that line and get a value for the difference between the groups. But there is uncertainty – you could fit many plausible lines. What the model does is fit all of the lines, and tell you how much more likely one slope is vs another, and what the most plausible one is. That’s what this graph is – all the possible slopes and how likely they are. The peak of the graph is the most plausible value of the gap, and the width tells you about how certain it is (high width = lots of plausible lines = low confidence).

Gender and perceived fairness of compensation

As noted previously, we collected about ⅔ of our responses from “male/men”. The data shows that Men and Not-men report being “extremely well-compensated” at the same frequency (14%). If we recall the stacks under the “Sentiment” plot, we see that 65% of men and 50% of not-men believe they’re well compensated to some degree.

	Men-only		Not-men
Extremely undercompensated	2%	2%	6%	6%
Somewhat undercompensated	6%	32%	18%	44%
Mildly undercompensated	18%		16%
Neutral	8%		10%
Mildly well-compensated	14%	65%	8%	50%
Somewhat well-compensated	37%		28%
Extremely well-compensated	14%		14%

Gender and technicality

A question arose: is there any evidence of gender affecting the likelihood of holding a technical role?

In the survey, we asked whether coding was a requirement of the job. We asked in this way because many respondents can code, but don’t have to do so on a regular basis. Since we asked that, we can ask the data “What is the probability of coding, given the gender of the person?” and see what the model thinks.

This is obviously a very simple model, and many things could influence coding indirectly from gender and other sources, however for our data and simple model we see a strong impact here – males are 40% more likely to be coding than others:

Transgender and genderqueer representation

In answer to the question “Do you identify as transgender or genderqueer?” 6% responded Yes, 91% responded No, and the rest declined to answer. Because of the small sample size, it was not possible to draw many conclusions.

Gender and management

US-based manager salaries by gender

For US-based respondents of all career levels who are people managers, responses between those who marked only “Male/men” versus all others combined.

MEDIAN SALARIES
of US-based respondents (in $USD)

	Men-only, US-based	Everyone else, US-based
People manager	$162,000	$165,000
Strategic manager	$160,000	$160,000

This table includes responses from all answers including “mid-career” level with roles involving management duties. An examination of people who are both managers and senior-level would likely show higher figures.

Gender and ageism

The tech industry as a whole is anecdotally ageist, but does gender play a further role in widening that gap? There’s not much evidence of that in our small dataset. If we model salary on age, gender and plot the predictions, we can see that while the range of predictions is large (the model is quite uncertain about the actual salaries), the pattern is consistent.

We see increasing salaries up to ages 45-54, but a drop in salary at the highest age category. However, this is true for both groups (Male and Not Male), and given the uncertainty in this model, we’re not comfortable drawing further conclusions at this time.

Career level and years in DevRel

Career-level is self-reported as entry, mid or senior. For time in the field, we recognize that in some organizations, DevRel has become an official title more recently. Since people who currently hold the title may have performed the duties entailed in the role, we asked them to use their best judgment.

SELF-REPORTED CAREER LEVEL

	Men	Not-men
Entry level	10%	4%
Mid-level	28%	52%
Senior level	62%	44%

Entry level: Ten percent of men self-report as entry-level, with a median of one year of experience in the field. Four percent of not-men describe themselves as entry-level, with a median of 2.3 years in the field.
Mid-level: 28% of men report a median three years’ experience in the field. 52% of not-men identified themselves as mid-level with a median two years’ of experience
Senior level: Both groups, men and not-men, reported a median of five years professional experience in DevRel. Sixty-two percent of men described themselves as senior, and 44% of not-men.

The above table shows all respondents, regardless of location. When career levels are broken out by country, we see the following:

Of US-based respondents, 41% describe themselves as mid-level. Fifty-five percent of respondents described their career level as senior
Of non-US respondents, 28% describe their level as mid-level, with 56% at senior level.

CONTENTS

SPECIAL THANKS TO OUR DATA SCIENTIST, GREG SUTCLIFFE

Greg Sutcliffe has been working in community management for a decade, and is currently the Principal Data Scientist for the Ansible Community. He's interested in how appropriate use of data can inform the development and governance of communities, especially with regards to open source projects. He also likes cooking.

DRC Survey IV: DevRel Job Titles and Salaries

bffjossy — Mon, 27 Sep 2021 23:12:20 +0000

Job Title

One area of interest to most practitioners is job title. We were interested in examining which titles predominate, and whether the titles cluster around differences in pay and/or experience level.

About the buckets

To glean some basic understanding of current norms, we created several rough buckets of titles. These buckets were “DevRel Management,” “Developer Advocate,” “Developer Evangelist,” “Technical Community Manager” and “Developer Experience.” Some responses did not fall into any of these buckets and are therefore not displayed in the chart. An example of one such response is “Event Management,” which was removed because it was so small that it could not be anonymized.

As we have noted previously, the samples are on the small side, and skewed toward the US. This is important to note, because titles can be highly influenced by regional trends and norms.

Job title

In descending order, “Developer Advocate” is the largest group, with almost half of the respondents holding a DA role (47.3 percent). DevRel Management ranks just below at 25.7%; then Developer Evangelist at 10.8%; Technical Community Managers at 5.4%; and Developer Experience at 2.0 percent. Responses that fell into other categories accounted for 8.8% of answers.

A challenge faced by many practitioners is the overlapping and occasionally arbitrary nature of titles in the field. Additionally, since DevRel is currently gaining interest as a specific profession, longtimers frequently find they have been effectively performing the role of (for example) a developer advocate since before the title existed.

Salaries by bucket

Salary ranges for the buckets are below, with the highest and lowest figures for each bucket group. The caveats: because of issues with currencies, this table displays salary information reported in US dollars. As previously stated, salaries in excess of $700,000 USD were removed, as we are unable to determine whether those are typographical errors or accurately reported outliers. The table does not display Developer Experience, because the sample was too small.

The highest values reported came from US-based respondents. The top five values were all in the US, and 90% of the top ten values came from the US)

	Lowest reported USD salary *	Highest reported USD salary*
DevRel Management	$14,000 USD¹	$281,000 USD
Developer Advocate	$38,400 USD	$220,000 USD
Developer Evangelist	$75,000 USD	$275,000 USD
Technical Community Manager	$29,000 USD	$150,000 USD
*_The highest and lowest salaries were not in the same region

¹ While this figure of $14,000 seems low, we decided to display it. Potential caveats to be aware of: the lowest values were located in several non-US countries with small sample sizes. It’s possible that such values are significantly lower because of currency conversion rates; or the number could be a typographical error.

It’s worth noting that the “DevRel Management” also refers to level, while the other categories draw mostly on function. For example, while a community manager falls under the Technical Community Manager category, a Director or Head of Community position is grouped under DevRel Management.

Buckets and tenure/career stage

Survey Questions

“How many years have you been at this company? If it’s a brand new job, please put 0.”

“How many years have you been in Developer Relations? We recognize DevRel is a relatively new title. Feel free to list how long you’ve been in DevRel-like positions.”


	Years at Company			Years in DevRel
	average	median	max	average	median	max
DevRel Management	3.16	2.00	12.00	6.30	5.00	25.00
Developer Advocate	2.01	2.00	8.00	3.40	3.00	11.00
Developer Evangelist	2.09	2.00	7.00	4.00	3.00	15.00
Technical Community Manager	1.66	1.38	5.00	3.40	3.00	10.00

SPECIAL THANKS TO OUR DATA SCIENTIST, GREG SUTCLIFFE

Greg Sutcliffe has been working in community management for a decade, and is currently the Principal Data Scientist for the Ansible Community. He's interested in how appropriate use of data can inform the development and governance of communities, especially with regards to open source projects. He also likes cooking.

DRC Survey III: DevRel Career Stages and Salaries

bffjossy — Mon, 27 Sep 2021 23:12:14 +0000

Career Stages

We’ve plotted career stages and salary below, since we thought it was something people would appreciate. We didn’t find any major surprises in the data, but it’s good practice to check for the unexpected (i.e some other result).

It would appear that once a worker has “enough” compensation, it becomes much harder to know how the package stacks up. All the medians of the people above Neutral stack atop each other. It’s possible this suggests a struggle to understand the difference between “Mildly Well Compensated” up through “Extremely Well Compensated”, but we don’t know for sure.

Sometimes it’s difficult to understand why we’re getting compensated more. If we get more uncertain in our worth as salary rises, then it’s helpful to see what causes those rises. One obvious candidate is career stage:

Career level and salary

This plot is solid, and expected in terms of its progression. It supports the hypothesis that salary rises with experience. However, it doesn’t help us understand the variation in higher salaries we see in “Compensation.”

“Career Level” is a somewhat vague term, so let’s go deeper. We asked about both years of expertise in DevRel work, and also the years in the current company. That suggests a place to look. First, let’s examine years of experience:

Salary, time in field, and time at current company

The right plot is not hugely informative – “Years at Company” alone isn’t telling us much. The left plot suggests a trend of around $6000 increase for each year of DevRel expertise, but there’s a catch!

People tend to gain more money as they gain experience. What we really want to do is compare people who stay in one job to people who move around – both are gaining “years” on the left plot, but one keeps resetting to “0” years on the right plot with each job move.

So, we built a model using both variables. Visualising such models is difficult – we have more than one “y-axis”, and we want to understand the effect of both things, but 3D plots are difficult to read clearly. However, we can compare two purely hypotheticalscenarios – one where a person stays at the same company for twelve years, and one where they move companies every three years:

Modeling job strategy: Salary, while staying or remaining at company

The lower line (all circles) shows the person who stays at Company 1 for twelve years. The upper line (with different shapes) shows the person who hops from company to company.

We’ve omitted the uncertainty (it’s quite large, as the sample is small), but a trend is apparent. In this hypothetical model, while the effect starts off small, there’s a noticeable salary jump after each job hop, and person 2 ends up earning ~$15,000 more after twelve years. In other words, seniority matters, but in this model staying in one place does indeed lead to decreased raises. (This is a well-researched effect, but it’s nice to find it in our data too).

This is a good explanation for some of the variation seen in the Compensation and Perceived Fairness plot – some people will feel well compensated because of how their career shook out, and others will have less satisfaction with their package.

SPECIAL THANKS TO OUR DATA SCIENTIST, GREG SUTCLIFFE

Greg Sutcliffe has been working in community management for a decade, and is currently the Principal Data Scientist for the Ansible Community. He's interested in how appropriate use of data can inform the development and governance of communities, especially with regards to open source projects. He also likes cooking.

DRC Salary II: DevRel Salaries & Fair Compensation

bffjossy — Mon, 27 Sep 2021 23:12:09 +0000

Compensation

The purpose of the survey is to give the DevRel community information about earning potential, parity, and salary fairness. We started by examining how well the community knows that already by asking “Do you feel like you are fairly compensated for your job?”.

Our rationale was that salaries are frequently not discussed among peers, making it difficult to know objectively how one’s compensation fares in comparison. However, most people do know to some degree of certainty how they perceive their own situation, and whether it feels appropriate on a gut level.

Sentiment

Compensation and perceived fairness

We directly asked about compensation in the survey, so that’s an obvious plot to make:

“Do you feel like you are fairly compensated for your job?”

At first glance, there’s a clear divide here between “Neutral and lower,” vs “above Neutral.” From Neutral downwards, there’s a clear trend with Salary – as you would expect. But above Neutral, the medians are all basically the same.

For further context and discussion, please refer to the section “Modeling job strategy.”

Salaries and cost of living

This trend is overall roughly what you’d expect (higher pay in higher cost areas), and the variation is pretty large

US and non-US responses

US-based responses

A question that came up in response to the Perceived Fairness plot was how do people in different locations feel about their compensation? Because we had a majority of responses from the United States, it’s possible to show how respondents in the US answered the question Do you feel like you are fairly compensated for your job?

We thought that it could be interesting to compare this to the non-US based responses. We can see in the responses that US-based respondents report being “Somewhat well-compensated” as their top answer (42.9%). Meanwhile, 25% of Non-US based respondents report they are “Mildly undercompensated”, with “Somewhat well-compensated” in second place at 21.9%.

It’s important to remember that the cost of living in the United States involves localized factors such as the cost of privatized healthcare. Healthcare in the US is commonly accessed as an employee benefit and is therefore a typical part of the compensation package.

SPECIAL THANKS TO OUR DATA SCIENTIST, GREG SUTCLIFFE

Greg Sutcliffe has been working in community management for a decade, and is currently the Principal Data Scientist for the Ansible Community. He's interested in how appropriate use of data can inform the development and governance of communities, especially with regards to open source projects. He also likes cooking.

DRC Survey I: DevRel Salaries and Specializations by Country

bffjossy — Mon, 27 Sep 2021 23:12:02 +0000

DevRel Salaries and Specializations by Country

We think this is a plot many expect to see. It is probably in line with the expectations most people hold. That said, if you’re in a country where there’s at least a small sample, it’s helpful to get a ballpark figure. We have redacted some countries with extremely small samples or in some cases, only a single respondent, in order to preserve anonymity.

The plot demonstrates that the highest reported salaries are in the United States, with the UK in second place. We can also see that the countries with the highest salaries have the widest spreads. A couple reminders: for display purposes, the salaries were converted to US dollars. And the plot should be read while bearing in mind that the infrastructure of countries varies widely, meaning that public healthcare, the cost of education, and other quality-of-life factors can strongly affect the budgetary needs of individuals by region.

Area of Expertise / Type of Community / Job Domain

The top ten answers to the question “What technical communities are you focused on in your current role?” are below, ranked in descending order. The question was multi-select.

This overlapping nature of answers made it challenging to draw conclusions about which specialties are directly responsible for which salary differences on the individual level. We’ve included the top ten answers, ranked from most-frequent to least-frequent however.

Open source
Cloud
API
Developer tooling
DevOps
Web
SaaS
Internal DevRel
Mobile
Data science

Most responses selected multiple options, which was not too surprising. Over half of all respondents self-reported as being currently focused on “Open Source”, regardless of whether the person was based in the US (54%) or outside the US (57%). “Cloud” and “API” came in close behind as seen in the table below

Technical specialties by location

*What technical communities are you focused on in your current role?Please select all applicable options.”

	Aggregate	US-based	non-US based
Open Source	56%	54%	57%
Cloud	54%	54%	55%
API	47%	44%	50%
Developer tooling	44%	44%	44%
DevOps	42%	47%	36%

The specialty rankings (shown here for all respondents, plus segmented by region) were fairly consistent across US and non-US based responses.

SPECIAL THANKS TO OUR DATA SCIENTIST, GREG SUTCLIFFE

Greg Sutcliffe has been working in community management for a decade, and is currently the Principal Data Scientist for the Ansible Community. He's interested in how appropriate use of data can inform the development and governance of communities, especially with regards to open source projects. He also likes cooking.