January 16, 2024 • 35 mins
In this week's episode of Dev Interrupted, co-host Conor Bronsdon is joined by David Yee, VP of Engineering at the New York Times. They dive into the often unseen aspects of organizational structures, discussing how every organization regardless of their defined values and principles, operates with a 'shadow'.
David discusses the role of senior leaders in addressing systemic problems and navigating the tensions between innovation and consistency. He highlights the importance of recognizing and integrating these organizational shadows to foster better decision-making and operational efficiency.
Drawing from his rich experience leading engineering teams in media, this conversation offers a unique perspective on the complexities of engineering leadership and explores the challenges of aligning technology with the century-old tradition of news reporting at the New York Times.
Show Notes:
Support the show:
- Subscribe to our Substack
- Leave us a review
- Subscribe on YouTube
- Follow us on Twitter or LinkedIn
Offers:
Mark as Played
Transcript
Episode Transcript
Available transcripts are automatically generated. Complete accuracy is not guaranteed.
David Yee (00:00):
every organizational leader has sat down at some
point to think about two things.
One, should I reorg?
And two, what are the values andprinciples of my organization?
When we do that, Weoften ask ourselves what
we're optimizing for.
That's a veryreasonable question.
What we don't ask is, whatdo we do with the things
(00:20):
that we're repressing?
How are we going toreconcile with that?
Or
Conor Bronsdon (00:26):
What's the impact of AI generated code.
How can you measure it?
Join linear B andThoughtWorks global head
of AI software delivery.
As we explore the metrics thatmeasure the impact gen AI has on
the software delivery process inis first of its kind workshop.
We'll share new data insightsfrom our gen AI impact report.
(00:47):
Case studies into how teamsare successfully leveraging
gen AI impact measurements,including adoption
benefits and risk metrics.
Plus a live demo of how youcan measure the impact of
your gen AI initiative today.
Register at linear beadat IO slash events.
And join us January 25th,January 30th, or on demand
to start measuring the ROI ofyour gen AI initiative today.
(01:11):
Welcome back to Dev Interrupted.
I'm your host, Conor Bronsdon.
And today I'm joined byDavid Yeat, VP of engineering
at this little companycalled the New York Times.
David Yee (01:18):
We're going to make it someday.
Conor Bronsdon (01:19):
Yeah, I believe in you.
This regional paper is scrappy.
Um, the amazing thing aboutyour career, David, is
you've scaled both smallorganizations and frankly,
enormous ones in the world ofjournalism, music, and art.
And you have an emphasis onproduct oriented engineering
and humane management,as I've heard described.
And that diverse perspectivehas led you to identify
(01:42):
something that you think is thatorganizations all have a shadow.
Can you define what that means?
David Yee (01:48):
Yeah, so what's interesting about talking about
organizations is we sort ofdefine like what they look like.
So if you think about like thevalues that an organization
has or its org chart, theseare all things that you see.
But The thing that I've beenobsessed with is this idea
that every organization isbigger than that, and the
parts that you don't talkabout still operate, and they
are excluded for a reason.
(02:09):
So, it's those bits that areleft on the side that are
still operating that showup in ways that kind of like
can irritate an organization.
. And so the things that I'vereally been thinking about is
how you bring that together.
Conor Bron (02:19):
What are those bits?
David Yee (02:20):
It depends on the organization.
So, these things are incon, either in conflict
with the way you operate.
So you could say like, allright, well we are the kind
of organization that makeslike really crisp decisions.
you know, and everybodyowns their decision,
something like that.
But there's always gonna be aclass of decision where people
are like, oh, well I reallyneed to ask a bunch of people,
or We need to agree on this.
(02:41):
That might not be in yourorg values, it might not be
the way you identify as anorganization, but it's still
something you need to do.
And so if you're goingaround trying to make sharp
decisions about things thatneed consensus, everyone's
going to get just annoyed.
and so that's an example,but it shows up in all
sorts of different ways.
Conor Bronsdon (02:57):
What are the impacts of the shadow?
I mean, you mentioned one, butit seems like there would be
a variety of ways this wouldimpact different organizations.
David Yee (03:05):
Yeah, generally speaking.
the thing it tends todo is cause dysfunction,
slow you down.
And it's not thatthe shadow does that.
It's the fact that you're notengaging with it meaningfully.
You're not really acknowledgingthat this is happening.
The example I was giving atlunch was when you think of an
organization, I had this debatea couple of years ago with my
(03:27):
product partner where I saidwe really have to be We have to
make sound technical decisions.
And my product partner waslike, we need to move fast,
we don't have time to makesound technical decisions.
And so we worked this out,just the two of us were
working this out, debatingback and forth for a long time.
Finally I said, we have tomove quickly while aiming
(03:51):
for technical consistency.
Now the result of that debatebetween two people means that
there's still this question thatthe engineers would ask about.
And this is in a differentorganization about like...
Where do, where do we value,show me where we value technical
consistency, but I do value it.
It's just that when it showsup, it tends to fly in the face
(04:11):
of the world, the org valuessay that we're supposed to move
quickly, why would we do this?
But the answer isyou have to do it.
And so it causes confusion,it can cause delay,
it can cause debate.
And frankly, for people whofind themselves in roles where
that, where the shadow, whereyou're sort of acting on behalf
of the shadow, you can actuallyget really marginalized.
(04:33):
And so there's both anorganizational process problem,
as well as sort of, youreally have to think about
what it means for individualhuman beings who, who really
aren't acting in service ofthe publicly stated goals.
Conor Bronsdon (04:44):
I think it's interesting because it sounds
like this dissonance betweenkind of the formal structure
of the organization and theway the organization actually
functions is likely to causemisallocation of resources,
both if you wanted to findhuman beings as resources,
which is controversial attimes, but we think of them
that way in a lot of ways.
And also I'm sure how moneyis spent in the organization,
(05:05):
how time is spent.
Uh, would that be a goodway to characterize this?
David Yee (05:08):
I mean, that, it's certainly
one aspect of it, right?
So if, if you, I guess I hadn'treally thought about it from
the perspective of like OKRsor metrics, but it's, it's
true that if you say, let's saythat you have an organization
where, uh, the goal is toinvent new things rapidly.
And you're not going tobe able to depend on a
(05:30):
framework to do that.
You're going to have to hackeverything from scratch.
You have a team that happensto be in your organization
that is responsible for aconsistent design system
or something like that.
If those two things are pairedin one organization, you're
always going to prioritizethe resources for invention.
You're never going toprioritize the resources for
(05:51):
consistency, even though thatconsistency is really important.
But at the end of the day,where do your priorities lie?
Where do your valuesand principles lie?
That's where you're goingto be paying attention.
And you have to ask yourself,at what cost are we not paying
attention to this other thing?
And this other thingis just nagging us in
the back of our mind.
And you can ask yourself,why are we wasting so much
(06:13):
time inventing new things?
We should be spending timebuilding a design system.
He says, well, that's notwhat our org is about.
Our org isn't about that.
That's something else.
So there is a larger questionabout how organizations
intersect and how you reallywant to sort of bring those,
you know, bring that shadowinto light is one of the things
that I talk a little bit about.
(06:34):
sometimes you bringthat shadow into light
inside your organization.
You can also find the shadow ofhome in some other organization.
Conor Bronsdon (06:40):
I'd love to, for you to expand on that.
That's a reallyinteresting concept.
David Yee (06:42):
Okay.
So first of all, Ihave to explain that.
of an organization's shadow,and this was sort of part of
my talk, it actually emergesfrom psychoanalysis, and
it emerges from Carl Jung.
A whole section of my talkis just about Carl Jung.
It's a really risky thing toget on stage at an engineering
management conference, butI'm going to talk about
the Swiss psychoanalyst.
(07:04):
But let's just acknowledge theSwiss psychoanalyst Carl Jung
has this notion of shadow,which is about the psyche,
and it's about personality.
If I'm the kind of person whosays, I'm a confident leader,
that's my persona, right,to use also a Jungian term,
like my persona is, like, Idon't know that I am this, but
(07:26):
let's say that I am a strong,confident, convinced leader.
that's how I define myself.
That's how I see myself.
That's how others see me.
That's my conscious self.
The unconscious selfright behind that,
contains a shadow, right?
So that's where wetalk about the shadow.
And the shadow contains allthe parts that are part of
me that I, that don't fit myconception of the person I am.
Conor Bronsdon (07:49):
Does it also contain the parts that do
fit the conception, or do youview it as a separate entity?
David Yee (07:54):
It's a totally separate entity, right?
Conor Bronsdon (07:55):
So, so maybe you're not
confident in the shadow?
David Yee (07:58):
Right.
Maybe you have doubt.
Maybe you're the kindaperson who, who has doubt.
The reason it's in the shadowis because doubt doesn't fit
my notion of being a confidentRight, convinced, strong leader.
Conor Bronsdon (08:07):
That's not how I wanna present to the world.
It's not how I wantto act in the world.
That's, and therefore Ipush it, push it down,
David Yee (08:11):
I push it down, I repress it,
Conor Bronsdon (08:12):
but eventually.
It'll come to the forefront.
Maybe
David Yee (08:15):
it's there.
Right.
Like and you know, any confidentleader, like, something, if
you invert that any leaderwho is devoted to consensus
may have a gut instinct.
That gut instinct.
And so there's nothingwrong with doubt.
Doubt's not bad.
It's, it is a tool, right?
(08:37):
And so ultimately, this kindof gets to the point, it's
like the goal, the way I closethe talk is basically talking
about integrating the shadow.
And both in psychoanalyticterms, but also in
organizational terms.
What you really want to dois bring that to light and
say listen sometimes as aleader I kind of doubt myself.
(08:58):
It serves me in the followingways, and it actually helps
get make sure that I'm makingmore More informed decisions in
moments that are not certain.
Conor Bronsdon (09:08):
Maybe I understand the risks whereas
if I was fully confidentI would ignore the risks.
David Yee (09:12):
Right exactly and that's what happens
to a lot of people so ifwe take that hypothetical
leader That person's gonnarun you off a cliff if
they ignore that shadow.
So it's really importantto integrate the shadow.
The same is trueof organizations.
So if you take that samemodel, and I, I don't know
that necessarily, thisis the best example, but
let's, let's follow it, thatyou've got sort of design
(09:33):
consistency in your shadow.
Okay.
Right?
Yeah.
I have to make consistentdesign, like it makes the
my product easier to use.
But that gets in the wayof innovation, gets in
the way of your publicpersona of innovation.
You have to ask yourself,how am I going to invite this
consistency into the work?
How do I acknowledge,listen, nonetheless,
this is something that myorganization needs to do.
(09:55):
How do I bring it in?
In a complex organization,the answer might be platform
thinking and you wanna movethat into a platform space.
Sure.
Someplace that's gonna beable to have the leverage,
that's gonna be ableprovide the leverage.
Conor Bronsdon (10:06):
So maybe building a platform
team or something.
David Yee (10:08):
Exactly, exactly right.
And you bring that tolight and you say, look,
we need a platform team.
Even though our job is toinnovate, we need a platform
team because ultimately it'sgoing to help us move faster.
one way of thinking aboutthe reintegration of
an organization shadow.
Conor Bronsdon (10:24):
This is fascinating because you've
also just given me a greatargument for why people
need platform teams.
Enable innovation in yourorganization by getting rid of
your organizational shadow andorient it to the platform team
so they can solve your problems.
I love it.
It broadly sounds likeyou think engineering
orgs need to think more indepth about organizational
design and the psyche ofthat organizational design.
, David Ye (10:45):
every organizational leader has sat down at some
point to think about two things.
One, should I reorg?
And two, what are the values andprinciples of my organization?
We do that as anorganizational tick.
We do that as leaders.
And of course thosethings are necessary.
When we do that, Weoften ask ourselves what
we're optimizing for.
That's a veryreasonable question.
(11:06):
What we don't ask is, whatdo we do with the things
that we're repressing?
How are we going toreconcile with that?
Or sometimes on the line, how doyou find yourself in a situation
where something is happeningand you're confused or annoyed
and you're saying like, butI thought our value was this.
(11:29):
How do you look at that,at those, at that moment
of decision when you re orged, when you sort of wrote
your values and principlesand said, Oh yeah, I made
a conscious decision not toinclude that in the values.
Nonetheless, it isimportant right now.
Like, your organizationstill needs those things.
If it didn't need those things,they wouldn't be coming out.
And, I'm not saying you haveto write like a 33 point org
(11:50):
values document, but sometimesyou need to ask yourself, How
do I invite these tensions?
into conscious view sothat I can reconcile them.
Conor Bronsdon (12:01):
Interesting.
So because they're outof view, they continue
to cause extra problems.
Almost like having like, Idon't know, like a, a burr
in your clothing that you,you can't get ahold of.
And once, once it's out, youcan get rid of it, you can
deal with it, you can leveragein whatever way you want to.
Mm-Hmm.
And so you're looking tosay, how can I leverage these
challenges in a way that'spositive for the organization
(12:22):
because we know we actuallydo need to solve them or.
Or use them or else we'regoing to invite more risk.
David Yee (12:29):
Yeah, we're going to invite more risk and
frankly, we're going to wastea lot of time and energy.
Conor Bronsdon (12:32):
And frustrate people.
David Yee (12:33):
This is in, I mean, I'm not the kind of
person who says, well, thisone simple trick to fix
your organization, but I do.
Conor Bronsdon (12:41):
That's what we're going to
title the podcast.
David Yee (12:43):
This one simple trick, the, I think that the,
When I think about operationalefficiency, and this is just my,
this is the way that I think.
When I think about operationalefficiency, I'm not thinking
like, okay, if we could just useAI for this, or if we could just
have engineers work longer orharder, we could get more done.
A lot of times when I look atthe things that slow us down,
(13:06):
it's related to confusion,disagreement, and delay.
And that delay often emergesfrom a sense of uncertainty
that says, I usually knowwhat to do, but in this
moment I dunno what to do.
And I'm really afraid of makinga decision because I'm afraid
I'm gonna be held accountablefor doing something contrary to
the values of my organization.
Or I'm gonna be acting out ofturn, or like, I don't know.
(13:26):
I feel like I should askthe VPE about this, when
actually sort of acknowledgingno, sometimes like.
There are decisions thatneed to get made in ways
that don't mirror thevalues or don't mirror
the hierarchy of your org.
Conor Bronsdon (13:39):
Right, and then that creates additional
communication costs by thatconfusion because now suddenly
you're talking to the VP,you're spending their time,
you're spending your time,you're diving into this problem
that hopefully if the shadowis better integrated into
your organizational You'dbe able to solve because you
understand how to think aboutthis prioritization, how to
think about this context.
David Yee (13:58):
Yeah, because at that point you would have
said, you will have said, andit's really funny actually
because recently we had thisconversation in my part of the
organization where rather thansort of stating the values, we
actually stated the tensions.
So, so one of those tensions,I'm trying to think of like
a really good one of thosetensions was sort of, I need,
on one hand you say, I need theliberty to act independently.
(14:19):
And on the other side,you say, I'm afraid of
stepping on people's toes.
And both those thingsare in play, right?
if I look at the organizationalvalues of the New York Times,
one of them is independence.
Right?
So, the former isreally independence.
I'm going to Act independently,I'm going to like depend on my
research, I'm going to make athorough point of view, and I'm
(14:39):
going to arrive at a decision.
But you'd be surprised howmuch in my organization
there's a fear ofstepping on people's toes.
Now, don't step on people'stoes is not written
in the organizationalvalues of my business.
But it's there, and you haveto be able to talk about it,
and so I, like, one of theways we did that recently was
bringing those up as tensions,and saying, how do you want
to reconcile these tensions?
(15:00):
That's the work of integration,you have to say it out loud.
Conor Bronsdon (15:03):
Do you think that's the result, this
specific tension of not wantingto step on other's toes?
of a way that the New YorkTimes culture functions, or
is it a result of people'sprevious experiences that
they're bringing to the table?
David Yee (15:16):
I tend to think it's all the manifestation
of past experiences.
I don't think, for example,I don't think that the
New York Times would say,would say that in any way.
I don't think it'srepresentative of,
like, what the New YorkTimes would be about.
There's no reason to make thatsort of part of the values.
I think that there'shistorically, in any
organization, there arepast experiences that
an organization has had.
Or that individual people havebrought to that organization
(15:38):
that have caused pain.
And as a result,you repress that.
You say, I'm not goingto do that again.
And it could be one of thethings I think a lot about is
it could be that the, that thepeople involved in those events
and those decisions and the waythey unfolded, they're gone.
Or those, those conditions areno longer true, but you have.
You've made a decision.
Conor Bronsdon (15:59):
You carry that cultural baseline or
foundation that was createdby folks at an organization.
I mean, frankly, we doas a society, right?
Like we're still buildingoff of the culture that
was here beforehand.
David Yee (16:10):
That's absolutely right.
And like frequently thosethings, because, because you,
if, when you look squarelyat them, you, like I said,
you wouldn't say like, yeah,I'm going to go ahead and put
this one in the constitution.
Like it's there that operates.
And so you haveto be able to ask.
Well, what am I goingto do about this?
And you have to be ableto acknowledge that it's
happening, because if you don'tacknowledge that it's happening,
(16:30):
you can't solve for it.
Conor Bronsdon (16:33):
What prompted you to apply this
concept of the shadow fromCarl Jung to org design?
What, what inspired this?
David Yee (16:41):
So, originally, when I was thinking about this...
So, you know, lead dev said,Oh, hey, come give this talk
about influence at scale.
And I said, okay, how am Igoing to do influence at scale?
So I could talk about likecommunication practices, I
could talk about, tailoring amessage, I had all these ideas
and then I said, Why is itthat our org charts don't work?
(17:02):
why is it that...
A manager can't make a decisionin their scope, or the VP of
engineering can't act withoutgetting full buy in from the
product partner of anothervice president of engineering
someplace else in the org.
There's something happening.
It so happened at thesame time I was part of
(17:24):
this coaching program.
That thinks about Jung a lot andwas talking about Jung a lot.
And so, it was surprisinglike as you dive into shadow
work and Jung's concepts ofshadow, it's really hard to
internalize what that is.
and so as I was, I wasthinking about these two
things at the same time.
And so I was telling, my friendNeha, who's presenting also
today, I was like, I thinkof these things as shadows.
(17:46):
there's a book that I citein my talk, about, uh,
about sort of managing theorganization shadow, or
mapping an organization shadow.
and it's, and it, Idiscovered it and it sort
of connects these two bits.
It connects Jungian psychologywith organizational thinking.
Uh, and so actually that leapingoff of that helped me sort
of formulate the connectionbetween these two things that
were going on in my head.
Conor Bronsdon (18:07):
I would be curious if you could share
an example of an organizationthat is either doing a great
job integrating its shadow,or maybe a terrible job.
David Yee (18:16):
I mean, honestly, my assumption is that all
organizations are doing aterrible job of this, right?
If they weren't doing aterrible job of this, then
we wouldn't have our work.
The reason why I'm giving thistalk here today is because as
senior leaders, as directors andvice presidents of engineering,
(18:39):
An engineering manageris having a problem with
another engineering manager,or this process that you
set up isn't working.
and so, the natural processorganizationally is to, pump
this up and you go this up tothe next, up to the next level.
So you're talking aboutdirectors, VPs of engineering,
other kind of senior leaders.
So, by definition, ourjob is to solve problems
that our designed systemsare not equipped to solve.
(19:02):
Which means that, like, you haveto look at the prob you have
to look at the problem and say,Okay, well, what's at play here?
All of us do a really bad job ofthis because we've, we've worked
really hard to create workingsystems, healthy systems, uh,
to be able to serve these needs.
you know, I can, you canimagine sort of any number
(19:23):
of organizations that, thatare well known for sort of
healthy, uh, healthy culture.
So Netflix, for example, andI don't work at Netflix, so I
can't speak for Netflix, butlegendarily from the outside,
everyone says, you know,Netflix has this figured out.
They have this culture document.
It defines everything.
If you follow the culturedocument, it's just gonna work.
It didn't work forthem as they scaled.
(19:45):
Everything that they've donein the last year of their
existence has been to sort ofquestion their culture document.
So in a sense, you mightsay, okay, Netflix is looking
at the shadow and saying,like, , this isn't working.
and you could talk to anybodyat any high functioning
organization and say,wow, it looks like a great
organization to work on this.
And yeah, well, on the groundis a little more complicated.
Right?
Right.
That's real.
So if it's more complicated,what do you do with that?
(20:07):
And so you could imagine thatan organization like Netflix
in making the decision tosort of like question their
culture document and changetheir processes, is trying to
integrate the shadow, whetherthey're doing a good job or not.
I, I don't know, butthat's just one example.
Again, I think mostorganizations of any complexity
and probably the more complexor the more tenured the
organization, the harderthey have, the harder time
(20:29):
they have of doing this.
Find difficulty in reconcilingthe shadow by design.
Conor Bronsdon (20:34):
This is really interesting.
I love to use Netflix as anexample because I spoke with
Carol Barrett at Netflixabout their transition as
they've started to hire morejunior engineers and some of
these changes they've made.
What you're saying resonates.
Netflix, we'respeaking for them.
We don't work in Netflix, but,I think by your definition has
been wrestling with that shadowwhere maybe that shadow, and
(20:56):
again, I'm just guessing itgrew over the years where they
said, this doesn't now workfor us as well as it used to.
We've, we've added thesecultural pieces that maybe we
haven't fully accepted or thatwe, if we need to adjust to.
And now they're saying, okay,how can we integrate parts
of that shadow to improve theorganizational and I would
be curious if, if they thinkthat's what they're doing and
that's how they're improvingthe organization or how
(21:18):
they would think about it.
Maybe an interesting followup for Carol or Catherine
here today is asking themlike, what do you think?
Is that what Netflixhas been doing?
Integrating the shadow?
Does David's talk resonate?
David Yee (21:28):
Yeah, I will ask.
So did I reallychange your mind?
Are you convinced now thatI was right about the way
Netflix ought to work?
Conor Bronsdon (21:35):
Let's get someone from Netflix in here.
Come on.
Producers, can we grab someone?
Gimme a call.
They're shaking my head here.
Yeah.
But maybe they'll, maybethey'll call David after this.
Yeah.
I, I'd love to add somemore context to this.
Sure.
So let's talk a bit about whatyou do at the New York Times
and your organization there.
Mm-Hmm.
'cause I think that'll be a,a great corollary to this,
you know, the, the shadowside of the organization
versus the tensions versus howyour organization functions.
(21:59):
Mm-Hmm.
, how big's your org atthe New York Times?
David Yee (22:01):
So my organization in the New York Times, I
have nine direct reports.
I have about 70engineers all told.
and it's an organizationcalled Subscriber Experiences.
Think of this as theorganization that is
responsible for the homepageof the New York Times.
the News Reader application,if you like, read the New
York Times on your phone.
Like that, that app, thatis an app that we own.
(22:21):
It's in conjunction witha lot of other teams.
Things like personalizationof the experience and the
emails you receive, right?
All of these are likesort of how do you bring,
uh, how do you bring theproduct to subscribers?
So when you describe meas a product oriented
engineering leader, That'smy bread and butter.
That's the kind oforganization that I'm in.
so that's, that'skind of my zone.
and You know, it is a,it is a discreet mission.
(22:44):
We use a mission system.
So we use a discreetmission inside the New
York Times, inside anorganization called XFUN.
and so we haveour own practices.
We have our own principles.
We have our own valuesthat supplement those of,
of the New York Times.
We don't operate exactly likethe organizations around us.
and yeah, like we have, likeyou said, we have our own
(23:04):
tensions, our own problems thatare unique and like we also
reflect on the problems andtensions of the organization
as a whole in our work.
You have to,
Conor Bronsdon (23:11):
would you be open to sharing some about how
your org works and then someof the tensions you identified?
I know you mentionedone earlier.
David Yee (23:16):
Yeah.
So,, this is always theclassic engineering debate.
It's like, do I move fast andbreak things, or do I make the
right decisions up front, right?
Conor Bron (23:22):
Which is hard to do.
David Yee (23:24):
You can't really live by one or the other
without, you know, you haveto eat your hat eventually.
And so one of the thingswe're figuring out now is,
Okay, how do we, how canwe bring products to market
really, really quickly?
while pointing, and so thisis, so if you, let's, the
process of reconciliationlooks a little bit like this.
(23:45):
if we were writing an org valuefor subscriber experiences,
we might say, we need to bringproducts to market as quickly
as possible to evaluate whetherthey serve our readers and serve
our newsroom in the right way.
Right?
And then we would leaveeverything else on the floor,
but by bringing up as a tension.
We can say, we're goingto build things as quickly
as possible to bring themto market and experiment.
(24:07):
But as we do it, we're goingto identify the journey
we're going to take to aconsistent, scalable, resilient,
extensible user experience.
Right, so we're not, we'restill going to be optimizing
for one thing, but we're goingto be doing it in tension
with this thing that wouldotherwise live in our shadow.
Which is making soundtechnical decisions as we work.
(24:28):
We are going to do thistoday, but in six months.
We're going to be liningup this batch of work
that brings it along.
That's not like an uncommongesture in engineering, but
one of the things I appreciatedabout the way we've been
thinking about this is thatyou can say it up front and
you can say, this is, it'snot this, but that it's
more like this in light of.
(24:50):
The long term goal, the shortterm in light of the long term.
Conor Bronsdon (24:54):
How do you measure your success?
I
David Yee (24:56):
mean, in our organization, we
measure our successthrough business success.
That's, that's why we exist.
We're not a platform.
My organization is nota platform organization.
We are defined by whether or notwe're able to engage our readers
in ways that keep them comingback and keep them informed.
I mean, that's the,that's the heart of it.
Conor Bronsdon (25:14):
I love that there's this key core mission
that drives how you do theorg design and also how you
think about the tensionswith your shadow work.
I think that's importantbecause it, I'm sure it makes
it much easier for you toprioritize even when there
is conflict that happens withthe shadow org because you
have a defined clear mission.
David Yee (25:32):
Yeah.
The clear mission doesn't.
It doesn't mean that you can'twrestle with these things.
I would, you know, I'll citethat mission left, right, and
center, no matter where we go.
I mean, that's, that's stilla very reasonable place to be.
It's a good goal to have.
But being able to sayoperationally, philosophically,
as we approach thesethings, we are living with
(25:53):
these things in tension.
And there's no, I'm not, likein all of these tensions,
I'm never going to arriveat a clear statement.
It's going to be acontinual body of work.
It's going to be shiftingbetween that sort of move fast
and like that consistency.
It's going to be shiftingbetween independence and
stepping on people's toes.
it's going to be shiftingbetween flexibility
and rigidity of teams.
(26:14):
Like we're always going tohave to sort of reckon with
that and like figure out whatlives in the light and what
lives in the shadow and howare we going to sort of bring
these things together tohave a healthy conversation.
Conor Bronsdon (26:22):
It's interesting to look at your
career in the context of this.
You've been someone who'sbeen involved in media
engineering, uh, as youput the phrase for a while.
You know, you were a directorof engineering at Vox, you
were a CTO at Editorially andco founder, and you've had
other roles in this space.
Do you think that thereare unique aspects to the
(26:43):
organizational shadows atmedia engineering orgs,
or is this something whereit's fairly common across
the engineering space?
David Yee (26:51):
That's a really good question.
I, yes, this is, thereis something and, but
it's not unique to themedia engineering space.
Let, let me, I'll give you thesort of broad strokes on it.
and then I'll get a little bitcloser because there's a really
interesting contrast betweenVox and the New York Times.
so broadly speaking, I wouldsay any time, especially in
engineering, there's alwaysa, some sort of foil, right?
(27:13):
I remember talking toa friend of mine, at an
organization I won't name.
And, you know, at the NewYork Times at the time, there
was a sort of this questionof like, who's in charge,
product or engineering?
It's going to be productor engineering doing this.
And I was like, and I wastalking to my friend, I
was like, well, what doyou like, you know, what's
your relationship betweenproduct and engineering?
He said, well, it's fine.
(27:33):
I said, well, like, there'snot a tension there?
He's like, no, not really.
I was like, well, you know, uh,how do you, How do you reconcile
with the needs of the businessand like being pushed to do
things that you don't wanna do?
He is like, oh, that'smarketing , . That's marketing.
Oh yeah.
Marketing's you'rethinking about marketing.
Marketing's the team thatscrews us up all the time.
but of course that's justfrom his point of view.
(27:55):
Yeah.
I'm sure if I, if I was friendswith somebody in marketing,
I said, well, what's yourperspective on engineering?
Like, oh, engineering's alwayslike, we wanna move slower.
We gotta make sure wedo this the right way.
, like, you know, and, andI think, so there's al
so first of all, there'salways that dynamic.
Right, and that often theshadow organization, you
could pretty cleanly find somecorollary in the relationship
between two functions orbranches of the organization.
(28:16):
So when I think about usuallywhat you'll hear if you talk to
people at the Washington Postor you talk to people at the
New York Times or Vox Media.
It's this tension between,quote, the newsroom and
the product, business,technology organization.
Not that that is, notthat that's a contentious
relationship, but what oftenhappens is, it is the job of
somebody in the newsroom to beable to report the news and to
(28:38):
be able to present that news aseffectively as possible to their
readers as quickly as they can.
if you think about anyorganization, you think about
CNN, you think about WAPO,you think about the New York
Times, The difference between,reporting on, The kinds of
things that are happening inthe world today in the next
two minutes versus the next20 minutes is everything.
(29:02):
So that need to be able to meetthe news and move with the news
is really, really critical.
And then on the engineeringside, if I just approached and
didn't have any background inthat, didn't understand how
important it was to be ableto say, like I've talked to
partners in the newsroom duringan election year and they've
said like, I was like, what kindof availability do you need?
(29:23):
It's at 99.
9, 99.
99, they're like, how about 100?
And you're like, well,what do you mean?
It's like, no, but what if we'redown for 5 minutes when the
Supreme Court makes its ruling?
Or something like that.
We cannot afford any downtime.
The answer is 100%.
You have to be able tohave, you have to be able
to engage in dialogue there.
Conor Bronsdon (29:39):
And that is tension, as you point out.
David Yee (29:41):
That is a tension, right?
Okay, I get it.
We need to be up when theSupreme Court makes its ruling.
That doesn't mean that oursystems need 100 percent uptime.
It means that we haveto find ways to be able
to meet our readers andconsumers in the ways.
that can follow the news,even accounting for the
possibility that some systemor another might be down.
(30:01):
You can design a systemaccordingly with that,
but you have to be ableto have that conversation
and that tension in play.
when I think about, Theunique challenges of media
engineering, it relates to howan organization came to be.
So the example I alwaysgive is that at Vox, Vox
Media more or less startedas a product organization,
(30:25):
started as a technologyorganization, and it really
started with sports blogs.
So we're going to makeit possible for fans of
the Cardinals to be ableto write about games,
alongside the journalistswho cover those games.
and so it was effectivelya product organization.
You add a bunchof sites like Vox.
com and The Verge andEater and all of them.
(30:47):
and so what you effectively haveis newsrooms that are built on
top of a product organization.
The New York Timesis very different.
You're talking about like acentury and a half of journalism
and of a newsroom that is sortof steeped in tradition and
systems and a newsroom thattalks in systems and you've got
a technology organization builton top of that newsroom and
Conor Bronsdon (31:09):
And to the New York Times is credit.
The New York Times is held upas a paragon of a successful
digital transformation.
Yeah.
Into this era.
But to your point,there's this massive.
Cultural foundationthat's already been
laid over 150 years.
David Yee (31:23):
Yeah.
And that culturalfoundation also is an
operational foundation.
And so at Vox you were able tosay, this is how software works,
this is how technology works.
How would you tell thenews in that context?
And in the New York Times,it's like we have built.
A tradition from over a centuryof how to report the news.
(31:44):
Yeah.
So then how doestechnology meet?
That is very, very differenttension and it creates
a very different shadow.
Conor Bronsdon (31:50):
I'm a big fan of the acquired podcast,
which does like breakdownsof businesses and they did
a really interesting episodefor me as someone who, you
know, I'm, I don't work at theNew York Times, but they did
a whole breakdown of the NewYork Times as transformation
into digital organization.
Yeah.
And it was.
It was fascinating to thinkabout, and particularly in
the context now of rethinkingabout it through, uh, the
Shadoworg piece, because I canonly imagine the amount of...
(32:15):
Institutional thought processesand culture and ideas that
have developed and some ofthat has been calcified and
like, this is how we operate.
I'm sure some of it is, youknow, tribal knowledge too
and like, oh, this is how weadjust that and making that
transition to, okay, now we areextremely digitally transformed.
We have applications, we have,you know, made purchases.
(32:37):
We have a massive digitalsubscriber base, must
bring a host of challenges.
David Yee (32:43):
It does.
I mean, it does bringa lot of challenges.
At the same time, I thinkwhat I've witnessed as
this has unfolded is thewillingness of the organization
to meet those challenges.
Which is not to say thatit's easy, or that people
are comfortable with it, butwe have an executive editor
who is relatively new to hisrole, he was sended to that
role, but before that, like,we were working directly
(33:05):
with this executive editor tofigure out What is the role
of technology in telling thestories of the New York Times?
The fact that we are in theposition we are in right now
as a business, with the digitalsubscription model that we have,
is a testament to how flexiblethe organization has been to
be able to say, you know what,we have to be able to figure
out how to integrate thesetwo different cultures, these
(33:25):
two different businesses, andthese two different conceptions
of the New York Times.
So, you know, you asked meearlier like, who's doing it
well, like I'm not, I'm, I'mnot here saying that it's,
that it's perfect or easy.
We're doing pretty well.
We're, I'm just saying,you know, New York Times
reconciles, its shadowpretty straightforwardly.
There's always gonnabe something that's
gonna piss you off.
There's always gonna besomething that gets in the way.
(33:45):
Those, and, and I say thisin the talk, like there is
no perfect reconciliation.
There is no perfect resolution.
The work is ongoing and thetalk is Titled shadow work.
And so it is anconstantly evolving
Conor Bronsdon (33:58):
process.Thank you for taking the time
to dive into this with me.
I, I've really enjoyed theconversation and it has, I
think, opened up some newthought avenues for me.
I I really appreciate youcoming to Upon podcast
to talk about it with us.
If you are a listener and youwanted to follow along with
David, where can you find him?
David Yee (34:17):
Yeah, so in the roiling hellscape of
social media, I've sort ofpulled sharply back from the
company, the company known asTwitter, but like my handle
there and in a lot of placesis tangentialism, which
is like a handle I came upwith when I was in my 20s.
So that's, that's,that is what it is.
So you find me there, but Iactually post a lot on Mastodon.
(34:39):
And so if you're lookingfor me on Mastodon, you can
look for at David at yee,Y E E dot camp, C A M P.
and that's where I domost of my writing now.
Conor Bronsdon (34:47):
Fantastic.
Uh, David, I've reallyenjoyed this conversation.
Thanks so much forcoming on the podcast.
And, uh, if you want moreinsights from leaders
like David, you can checkout our sub stack at dev
opted dot sub stack.com.
David, thanks for coming on.
David Yee (34:59):
Thanks for having me.
every organizational leader has sat down at some
point to think about two things.
One, should I reorg?
And two, what are the values andprinciples of my organization?
When we do that, Weoften ask ourselves what
we're optimizing for.
That's a veryreasonable question.
What we don't ask is, whatdo we do with the things
(00:20):
that we're repressing?
How are we going toreconcile with that?
Or
Conor Bronsdon (00:26):
What's the impact of AI generated code.
How can you measure it?
Join linear B andThoughtWorks global head
of AI software delivery.
As we explore the metrics thatmeasure the impact gen AI has on
the software delivery process inis first of its kind workshop.
We'll share new data insightsfrom our gen AI impact report.
(00:47):
Case studies into how teamsare successfully leveraging
gen AI impact measurements,including adoption
benefits and risk metrics.
Plus a live demo of how youcan measure the impact of
your gen AI initiative today.
Register at linear beadat IO slash events.
And join us January 25th,January 30th, or on demand
to start measuring the ROI ofyour gen AI initiative today.
(01:11):
Welcome back to Dev Interrupted.
I'm your host, Conor Bronsdon.
And today I'm joined byDavid Yeat, VP of engineering
at this little companycalled the New York Times.
David Yee (01:18):
We're going to make it someday.
Conor Bronsdon (01:19):
Yeah, I believe in you.
This regional paper is scrappy.
Um, the amazing thing aboutyour career, David, is
you've scaled both smallorganizations and frankly,
enormous ones in the world ofjournalism, music, and art.
And you have an emphasis onproduct oriented engineering
and humane management,as I've heard described.
And that diverse perspectivehas led you to identify
(01:42):
something that you think is thatorganizations all have a shadow.
Can you define what that means?
David Yee (01:48):
Yeah, so what's interesting about talking about
organizations is we sort ofdefine like what they look like.
So if you think about like thevalues that an organization
has or its org chart, theseare all things that you see.
But The thing that I've beenobsessed with is this idea
that every organization isbigger than that, and the
parts that you don't talkabout still operate, and they
are excluded for a reason.
(02:09):
So, it's those bits that areleft on the side that are
still operating that showup in ways that kind of like
can irritate an organization.
. And so the things that I'vereally been thinking about is
how you bring that together.
Conor Bron (02:19):
What are those bits?
David Yee (02:20):
It depends on the organization.
So, these things are incon, either in conflict
with the way you operate.
So you could say like, allright, well we are the kind
of organization that makeslike really crisp decisions.
you know, and everybodyowns their decision,
something like that.
But there's always gonna be aclass of decision where people
are like, oh, well I reallyneed to ask a bunch of people,
or We need to agree on this.
(02:41):
That might not be in yourorg values, it might not be
the way you identify as anorganization, but it's still
something you need to do.
And so if you're goingaround trying to make sharp
decisions about things thatneed consensus, everyone's
going to get just annoyed.
and so that's an example,but it shows up in all
sorts of different ways.
Conor Bronsdon (02:57):
What are the impacts of the shadow?
I mean, you mentioned one, butit seems like there would be
a variety of ways this wouldimpact different organizations.
David Yee (03:05):
Yeah, generally speaking.
the thing it tends todo is cause dysfunction,
slow you down.
And it's not thatthe shadow does that.
It's the fact that you're notengaging with it meaningfully.
You're not really acknowledgingthat this is happening.
The example I was giving atlunch was when you think of an
organization, I had this debatea couple of years ago with my
(03:27):
product partner where I saidwe really have to be We have to
make sound technical decisions.
And my product partner waslike, we need to move fast,
we don't have time to makesound technical decisions.
And so we worked this out,just the two of us were
working this out, debatingback and forth for a long time.
Finally I said, we have tomove quickly while aiming
(03:51):
for technical consistency.
Now the result of that debatebetween two people means that
there's still this question thatthe engineers would ask about.
And this is in a differentorganization about like...
Where do, where do we value,show me where we value technical
consistency, but I do value it.
It's just that when it showsup, it tends to fly in the face
(04:11):
of the world, the org valuessay that we're supposed to move
quickly, why would we do this?
But the answer isyou have to do it.
And so it causes confusion,it can cause delay,
it can cause debate.
And frankly, for people whofind themselves in roles where
that, where the shadow, whereyou're sort of acting on behalf
of the shadow, you can actuallyget really marginalized.
(04:33):
And so there's both anorganizational process problem,
as well as sort of, youreally have to think about
what it means for individualhuman beings who, who really
aren't acting in service ofthe publicly stated goals.
Conor Bronsdon (04:44):
I think it's interesting because it sounds
like this dissonance betweenkind of the formal structure
of the organization and theway the organization actually
functions is likely to causemisallocation of resources,
both if you wanted to findhuman beings as resources,
which is controversial attimes, but we think of them
that way in a lot of ways.
And also I'm sure how moneyis spent in the organization,
(05:05):
how time is spent.
Uh, would that be a goodway to characterize this?
David Yee (05:08):
I mean, that, it's certainly
one aspect of it, right?
So if, if you, I guess I hadn'treally thought about it from
the perspective of like OKRsor metrics, but it's, it's
true that if you say, let's saythat you have an organization
where, uh, the goal is toinvent new things rapidly.
And you're not going tobe able to depend on a
(05:30):
framework to do that.
You're going to have to hackeverything from scratch.
You have a team that happensto be in your organization
that is responsible for aconsistent design system
or something like that.
If those two things are pairedin one organization, you're
always going to prioritizethe resources for invention.
You're never going toprioritize the resources for
(05:51):
consistency, even though thatconsistency is really important.
But at the end of the day,where do your priorities lie?
Where do your valuesand principles lie?
That's where you're goingto be paying attention.
And you have to ask yourself,at what cost are we not paying
attention to this other thing?
And this other thingis just nagging us in
the back of our mind.
And you can ask yourself,why are we wasting so much
(06:13):
time inventing new things?
We should be spending timebuilding a design system.
He says, well, that's notwhat our org is about.
Our org isn't about that.
That's something else.
So there is a larger questionabout how organizations
intersect and how you reallywant to sort of bring those,
you know, bring that shadowinto light is one of the things
that I talk a little bit about.
(06:34):
sometimes you bringthat shadow into light
inside your organization.
You can also find the shadow ofhome in some other organization.
Conor Bronsdon (06:40):
I'd love to, for you to expand on that.
That's a reallyinteresting concept.
David Yee (06:42):
Okay.
So first of all, Ihave to explain that.
of an organization's shadow,and this was sort of part of
my talk, it actually emergesfrom psychoanalysis, and
it emerges from Carl Jung.
A whole section of my talkis just about Carl Jung.
It's a really risky thing toget on stage at an engineering
management conference, butI'm going to talk about
the Swiss psychoanalyst.
(07:04):
But let's just acknowledge theSwiss psychoanalyst Carl Jung
has this notion of shadow,which is about the psyche,
and it's about personality.
If I'm the kind of person whosays, I'm a confident leader,
that's my persona, right,to use also a Jungian term,
like my persona is, like, Idon't know that I am this, but
(07:26):
let's say that I am a strong,confident, convinced leader.
that's how I define myself.
That's how I see myself.
That's how others see me.
That's my conscious self.
The unconscious selfright behind that,
contains a shadow, right?
So that's where wetalk about the shadow.
And the shadow contains allthe parts that are part of
me that I, that don't fit myconception of the person I am.
Conor Bronsdon (07:49):
Does it also contain the parts that do
fit the conception, or do youview it as a separate entity?
David Yee (07:54):
It's a totally separate entity, right?
Conor Bronsdon (07:55):
So, so maybe you're not
confident in the shadow?
David Yee (07:58):
Right.
Maybe you have doubt.
Maybe you're the kindaperson who, who has doubt.
The reason it's in the shadowis because doubt doesn't fit
my notion of being a confidentRight, convinced, strong leader.
Conor Bronsdon (08:07):
That's not how I wanna present to the world.
It's not how I wantto act in the world.
That's, and therefore Ipush it, push it down,
David Yee (08:11):
I push it down, I repress it,
Conor Bronsdon (08:12):
but eventually.
It'll come to the forefront.
Maybe
David Yee (08:15):
it's there.
Right.
Like and you know, any confidentleader, like, something, if
you invert that any leaderwho is devoted to consensus
may have a gut instinct.
That gut instinct.
And so there's nothingwrong with doubt.
Doubt's not bad.
It's, it is a tool, right?
(08:37):
And so ultimately, this kindof gets to the point, it's
like the goal, the way I closethe talk is basically talking
about integrating the shadow.
And both in psychoanalyticterms, but also in
organizational terms.
What you really want to dois bring that to light and
say listen sometimes as aleader I kind of doubt myself.
(08:58):
It serves me in the followingways, and it actually helps
get make sure that I'm makingmore More informed decisions in
moments that are not certain.
Conor Bronsdon (09:08):
Maybe I understand the risks whereas
if I was fully confidentI would ignore the risks.
David Yee (09:12):
Right exactly and that's what happens
to a lot of people so ifwe take that hypothetical
leader That person's gonnarun you off a cliff if
they ignore that shadow.
So it's really importantto integrate the shadow.
The same is trueof organizations.
So if you take that samemodel, and I, I don't know
that necessarily, thisis the best example, but
let's, let's follow it, thatyou've got sort of design
(09:33):
consistency in your shadow.
Okay.
Right?
Yeah.
I have to make consistentdesign, like it makes the
my product easier to use.
But that gets in the wayof innovation, gets in
the way of your publicpersona of innovation.
You have to ask yourself,how am I going to invite this
consistency into the work?
How do I acknowledge,listen, nonetheless,
this is something that myorganization needs to do.
(09:55):
How do I bring it in?
In a complex organization,the answer might be platform
thinking and you wanna movethat into a platform space.
Sure.
Someplace that's gonna beable to have the leverage,
that's gonna be ableprovide the leverage.
Conor Bronsdon (10:06):
So maybe building a platform
team or something.
David Yee (10:08):
Exactly, exactly right.
And you bring that tolight and you say, look,
we need a platform team.
Even though our job is toinnovate, we need a platform
team because ultimately it'sgoing to help us move faster.
one way of thinking aboutthe reintegration of
an organization shadow.
Conor Bronsdon (10:24):
This is fascinating because you've
also just given me a greatargument for why people
need platform teams.
Enable innovation in yourorganization by getting rid of
your organizational shadow andorient it to the platform team
so they can solve your problems.
I love it.
It broadly sounds likeyou think engineering
orgs need to think more indepth about organizational
design and the psyche ofthat organizational design.
, David Ye (10:45):
every organizational leader has sat down at some
point to think about two things.
One, should I reorg?
And two, what are the values andprinciples of my organization?
We do that as anorganizational tick.
We do that as leaders.
And of course thosethings are necessary.
When we do that, Weoften ask ourselves what
we're optimizing for.
That's a veryreasonable question.
(11:06):
What we don't ask is, whatdo we do with the things
that we're repressing?
How are we going toreconcile with that?
Or sometimes on the line, how doyou find yourself in a situation
where something is happeningand you're confused or annoyed
and you're saying like, butI thought our value was this.
(11:29):
How do you look at that,at those, at that moment
of decision when you re orged, when you sort of wrote
your values and principlesand said, Oh yeah, I made
a conscious decision not toinclude that in the values.
Nonetheless, it isimportant right now.
Like, your organizationstill needs those things.
If it didn't need those things,they wouldn't be coming out.
And, I'm not saying you haveto write like a 33 point org
(11:50):
values document, but sometimesyou need to ask yourself, How
do I invite these tensions?
into conscious view sothat I can reconcile them.
Conor Bronsdon (12:01):
Interesting.
So because they're outof view, they continue
to cause extra problems.
Almost like having like, Idon't know, like a, a burr
in your clothing that you,you can't get ahold of.
And once, once it's out, youcan get rid of it, you can
deal with it, you can leveragein whatever way you want to.
Mm-Hmm.
And so you're looking tosay, how can I leverage these
challenges in a way that'spositive for the organization
(12:22):
because we know we actuallydo need to solve them or.
Or use them or else we'regoing to invite more risk.
David Yee (12:29):
Yeah, we're going to invite more risk and
frankly, we're going to wastea lot of time and energy.
Conor Bronsdon (12:32):
And frustrate people.
David Yee (12:33):
This is in, I mean, I'm not the kind of
person who says, well, thisone simple trick to fix
your organization, but I do.
Conor Bronsdon (12:41):
That's what we're going to
title the podcast.
David Yee (12:43):
This one simple trick, the, I think that the,
When I think about operationalefficiency, and this is just my,
this is the way that I think.
When I think about operationalefficiency, I'm not thinking
like, okay, if we could just useAI for this, or if we could just
have engineers work longer orharder, we could get more done.
A lot of times when I look atthe things that slow us down,
(13:06):
it's related to confusion,disagreement, and delay.
And that delay often emergesfrom a sense of uncertainty
that says, I usually knowwhat to do, but in this
moment I dunno what to do.
And I'm really afraid of makinga decision because I'm afraid
I'm gonna be held accountablefor doing something contrary to
the values of my organization.
Or I'm gonna be acting out ofturn, or like, I don't know.
(13:26):
I feel like I should askthe VPE about this, when
actually sort of acknowledgingno, sometimes like.
There are decisions thatneed to get made in ways
that don't mirror thevalues or don't mirror
the hierarchy of your org.
Conor Bronsdon (13:39):
Right, and then that creates additional
communication costs by thatconfusion because now suddenly
you're talking to the VP,you're spending their time,
you're spending your time,you're diving into this problem
that hopefully if the shadowis better integrated into
your organizational You'dbe able to solve because you
understand how to think aboutthis prioritization, how to
think about this context.
David Yee (13:58):
Yeah, because at that point you would have
said, you will have said, andit's really funny actually
because recently we had thisconversation in my part of the
organization where rather thansort of stating the values, we
actually stated the tensions.
So, so one of those tensions,I'm trying to think of like
a really good one of thosetensions was sort of, I need,
on one hand you say, I need theliberty to act independently.
(14:19):
And on the other side,you say, I'm afraid of
stepping on people's toes.
And both those thingsare in play, right?
if I look at the organizationalvalues of the New York Times,
one of them is independence.
Right?
So, the former isreally independence.
I'm going to Act independently,I'm going to like depend on my
research, I'm going to make athorough point of view, and I'm
(14:39):
going to arrive at a decision.
But you'd be surprised howmuch in my organization
there's a fear ofstepping on people's toes.
Now, don't step on people'stoes is not written
in the organizationalvalues of my business.
But it's there, and you haveto be able to talk about it,
and so I, like, one of theways we did that recently was
bringing those up as tensions,and saying, how do you want
to reconcile these tensions?
(15:00):
That's the work of integration,you have to say it out loud.
Conor Bronsdon (15:03):
Do you think that's the result, this
specific tension of not wantingto step on other's toes?
of a way that the New YorkTimes culture functions, or
is it a result of people'sprevious experiences that
they're bringing to the table?
David Yee (15:16):
I tend to think it's all the manifestation
of past experiences.
I don't think, for example,I don't think that the
New York Times would say,would say that in any way.
I don't think it'srepresentative of,
like, what the New YorkTimes would be about.
There's no reason to make thatsort of part of the values.
I think that there'shistorically, in any
organization, there arepast experiences that
an organization has had.
Or that individual people havebrought to that organization
(15:38):
that have caused pain.
And as a result,you repress that.
You say, I'm not goingto do that again.
And it could be one of thethings I think a lot about is
it could be that the, that thepeople involved in those events
and those decisions and the waythey unfolded, they're gone.
Or those, those conditions areno longer true, but you have.
You've made a decision.
Conor Bronsdon (15:59):
You carry that cultural baseline or
foundation that was createdby folks at an organization.
I mean, frankly, we doas a society, right?
Like we're still buildingoff of the culture that
was here beforehand.
David Yee (16:10):
That's absolutely right.
And like frequently thosethings, because, because you,
if, when you look squarelyat them, you, like I said,
you wouldn't say like, yeah,I'm going to go ahead and put
this one in the constitution.
Like it's there that operates.
And so you haveto be able to ask.
Well, what am I goingto do about this?
And you have to be ableto acknowledge that it's
happening, because if you don'tacknowledge that it's happening,
(16:30):
you can't solve for it.
Conor Bronsdon (16:33):
What prompted you to apply this
concept of the shadow fromCarl Jung to org design?
What, what inspired this?
David Yee (16:41):
So, originally, when I was thinking about this...
So, you know, lead dev said,Oh, hey, come give this talk
about influence at scale.
And I said, okay, how am Igoing to do influence at scale?
So I could talk about likecommunication practices, I
could talk about, tailoring amessage, I had all these ideas
and then I said, Why is itthat our org charts don't work?
(17:02):
why is it that...
A manager can't make a decisionin their scope, or the VP of
engineering can't act withoutgetting full buy in from the
product partner of anothervice president of engineering
someplace else in the org.
There's something happening.
It so happened at thesame time I was part of
(17:24):
this coaching program.
That thinks about Jung a lot andwas talking about Jung a lot.
And so, it was surprisinglike as you dive into shadow
work and Jung's concepts ofshadow, it's really hard to
internalize what that is.
and so as I was, I wasthinking about these two
things at the same time.
And so I was telling, my friendNeha, who's presenting also
today, I was like, I thinkof these things as shadows.
(17:46):
there's a book that I citein my talk, about, uh,
about sort of managing theorganization shadow, or
mapping an organization shadow.
and it's, and it, Idiscovered it and it sort
of connects these two bits.
It connects Jungian psychologywith organizational thinking.
Uh, and so actually that leapingoff of that helped me sort
of formulate the connectionbetween these two things that
were going on in my head.
Conor Bronsdon (18:07):
I would be curious if you could share
an example of an organizationthat is either doing a great
job integrating its shadow,or maybe a terrible job.
David Yee (18:16):
I mean, honestly, my assumption is that all
organizations are doing aterrible job of this, right?
If they weren't doing aterrible job of this, then
we wouldn't have our work.
The reason why I'm giving thistalk here today is because as
senior leaders, as directors andvice presidents of engineering,
(18:39):
An engineering manageris having a problem with
another engineering manager,or this process that you
set up isn't working.
and so, the natural processorganizationally is to, pump
this up and you go this up tothe next, up to the next level.
So you're talking aboutdirectors, VPs of engineering,
other kind of senior leaders.
So, by definition, ourjob is to solve problems
that our designed systemsare not equipped to solve.
(19:02):
Which means that, like, you haveto look at the prob you have
to look at the problem and say,Okay, well, what's at play here?
All of us do a really bad job ofthis because we've, we've worked
really hard to create workingsystems, healthy systems, uh,
to be able to serve these needs.
you know, I can, you canimagine sort of any number
(19:23):
of organizations that, thatare well known for sort of
healthy, uh, healthy culture.
So Netflix, for example, andI don't work at Netflix, so I
can't speak for Netflix, butlegendarily from the outside,
everyone says, you know,Netflix has this figured out.
They have this culture document.
It defines everything.
If you follow the culturedocument, it's just gonna work.
It didn't work forthem as they scaled.
(19:45):
Everything that they've donein the last year of their
existence has been to sort ofquestion their culture document.
So in a sense, you mightsay, okay, Netflix is looking
at the shadow and saying,like, , this isn't working.
and you could talk to anybodyat any high functioning
organization and say,wow, it looks like a great
organization to work on this.
And yeah, well, on the groundis a little more complicated.
Right?
Right.
That's real.
So if it's more complicated,what do you do with that?
(20:07):
And so you could imagine thatan organization like Netflix
in making the decision tosort of like question their
culture document and changetheir processes, is trying to
integrate the shadow, whetherthey're doing a good job or not.
I, I don't know, butthat's just one example.
Again, I think mostorganizations of any complexity
and probably the more complexor the more tenured the
organization, the harderthey have, the harder time
(20:29):
they have of doing this.
Find difficulty in reconcilingthe shadow by design.
Conor Bronsdon (20:34):
This is really interesting.
I love to use Netflix as anexample because I spoke with
Carol Barrett at Netflixabout their transition as
they've started to hire morejunior engineers and some of
these changes they've made.
What you're saying resonates.
Netflix, we'respeaking for them.
We don't work in Netflix, but,I think by your definition has
been wrestling with that shadowwhere maybe that shadow, and
(20:56):
again, I'm just guessing itgrew over the years where they
said, this doesn't now workfor us as well as it used to.
We've, we've added thesecultural pieces that maybe we
haven't fully accepted or thatwe, if we need to adjust to.
And now they're saying, okay,how can we integrate parts
of that shadow to improve theorganizational and I would
be curious if, if they thinkthat's what they're doing and
that's how they're improvingthe organization or how
(21:18):
they would think about it.
Maybe an interesting followup for Carol or Catherine
here today is asking themlike, what do you think?
Is that what Netflixhas been doing?
Integrating the shadow?
Does David's talk resonate?
David Yee (21:28):
Yeah, I will ask.
So did I reallychange your mind?
Are you convinced now thatI was right about the way
Netflix ought to work?
Conor Bronsdon (21:35):
Let's get someone from Netflix in here.
Come on.
Producers, can we grab someone?
Gimme a call.
They're shaking my head here.
Yeah.
But maybe they'll, maybethey'll call David after this.
Yeah.
I, I'd love to add somemore context to this.
Sure.
So let's talk a bit about whatyou do at the New York Times
and your organization there.
Mm-Hmm.
'cause I think that'll be a,a great corollary to this,
you know, the, the shadowside of the organization
versus the tensions versus howyour organization functions.
(21:59):
Mm-Hmm.
, how big's your org atthe New York Times?
David Yee (22:01):
So my organization in the New York Times, I
have nine direct reports.
I have about 70engineers all told.
and it's an organizationcalled Subscriber Experiences.
Think of this as theorganization that is
responsible for the homepageof the New York Times.
the News Reader application,if you like, read the New
York Times on your phone.
Like that, that app, thatis an app that we own.
(22:21):
It's in conjunction witha lot of other teams.
Things like personalizationof the experience and the
emails you receive, right?
All of these are likesort of how do you bring,
uh, how do you bring theproduct to subscribers?
So when you describe meas a product oriented
engineering leader, That'smy bread and butter.
That's the kind oforganization that I'm in.
so that's, that'skind of my zone.
and You know, it is a,it is a discreet mission.
(22:44):
We use a mission system.
So we use a discreetmission inside the New
York Times, inside anorganization called XFUN.
and so we haveour own practices.
We have our own principles.
We have our own valuesthat supplement those of,
of the New York Times.
We don't operate exactly likethe organizations around us.
and yeah, like we have, likeyou said, we have our own
(23:04):
tensions, our own problems thatare unique and like we also
reflect on the problems andtensions of the organization
as a whole in our work.
You have to,
Conor Bronsdon (23:11):
would you be open to sharing some about how
your org works and then someof the tensions you identified?
I know you mentionedone earlier.
David Yee (23:16):
Yeah.
So,, this is always theclassic engineering debate.
It's like, do I move fast andbreak things, or do I make the
right decisions up front, right?
Conor Bron (23:22):
Which is hard to do.
David Yee (23:24):
You can't really live by one or the other
without, you know, you haveto eat your hat eventually.
And so one of the thingswe're figuring out now is,
Okay, how do we, how canwe bring products to market
really, really quickly?
while pointing, and so thisis, so if you, let's, the
process of reconciliationlooks a little bit like this.
(23:45):
if we were writing an org valuefor subscriber experiences,
we might say, we need to bringproducts to market as quickly
as possible to evaluate whetherthey serve our readers and serve
our newsroom in the right way.
Right?
And then we would leaveeverything else on the floor,
but by bringing up as a tension.
We can say, we're goingto build things as quickly
as possible to bring themto market and experiment.
(24:07):
But as we do it, we're goingto identify the journey
we're going to take to aconsistent, scalable, resilient,
extensible user experience.
Right, so we're not, we'restill going to be optimizing
for one thing, but we're goingto be doing it in tension
with this thing that wouldotherwise live in our shadow.
Which is making soundtechnical decisions as we work.
(24:28):
We are going to do thistoday, but in six months.
We're going to be liningup this batch of work
that brings it along.
That's not like an uncommongesture in engineering, but
one of the things I appreciatedabout the way we've been
thinking about this is thatyou can say it up front and
you can say, this is, it'snot this, but that it's
more like this in light of.
(24:50):
The long term goal, the shortterm in light of the long term.
Conor Bronsdon (24:54):
How do you measure your success?
I
David Yee (24:56):
mean, in our organization, we
measure our successthrough business success.
That's, that's why we exist.
We're not a platform.
My organization is nota platform organization.
We are defined by whether or notwe're able to engage our readers
in ways that keep them comingback and keep them informed.
I mean, that's the,that's the heart of it.
Conor Bronsdon (25:14):
I love that there's this key core mission
that drives how you do theorg design and also how you
think about the tensionswith your shadow work.
I think that's importantbecause it, I'm sure it makes
it much easier for you toprioritize even when there
is conflict that happens withthe shadow org because you
have a defined clear mission.
David Yee (25:32):
Yeah.
The clear mission doesn't.
It doesn't mean that you can'twrestle with these things.
I would, you know, I'll citethat mission left, right, and
center, no matter where we go.
I mean, that's, that's stilla very reasonable place to be.
It's a good goal to have.
But being able to sayoperationally, philosophically,
as we approach thesethings, we are living with
(25:53):
these things in tension.
And there's no, I'm not, likein all of these tensions,
I'm never going to arriveat a clear statement.
It's going to be acontinual body of work.
It's going to be shiftingbetween that sort of move fast
and like that consistency.
It's going to be shiftingbetween independence and
stepping on people's toes.
it's going to be shiftingbetween flexibility
and rigidity of teams.
(26:14):
Like we're always going tohave to sort of reckon with
that and like figure out whatlives in the light and what
lives in the shadow and howare we going to sort of bring
these things together tohave a healthy conversation.
Conor Bronsdon (26:22):
It's interesting to look at your
career in the context of this.
You've been someone who'sbeen involved in media
engineering, uh, as youput the phrase for a while.
You know, you were a directorof engineering at Vox, you
were a CTO at Editorially andco founder, and you've had
other roles in this space.
Do you think that thereare unique aspects to the
(26:43):
organizational shadows atmedia engineering orgs,
or is this something whereit's fairly common across
the engineering space?
David Yee (26:51):
That's a really good question.
I, yes, this is, thereis something and, but
it's not unique to themedia engineering space.
Let, let me, I'll give you thesort of broad strokes on it.
and then I'll get a little bitcloser because there's a really
interesting contrast betweenVox and the New York Times.
so broadly speaking, I wouldsay any time, especially in
engineering, there's alwaysa, some sort of foil, right?
(27:13):
I remember talking toa friend of mine, at an
organization I won't name.
And, you know, at the NewYork Times at the time, there
was a sort of this questionof like, who's in charge,
product or engineering?
It's going to be productor engineering doing this.
And I was like, and I wastalking to my friend, I
was like, well, what doyou like, you know, what's
your relationship betweenproduct and engineering?
He said, well, it's fine.
(27:33):
I said, well, like, there'snot a tension there?
He's like, no, not really.
I was like, well, you know, uh,how do you, How do you reconcile
with the needs of the businessand like being pushed to do
things that you don't wanna do?
He is like, oh, that'smarketing , . That's marketing.
Oh yeah.
Marketing's you'rethinking about marketing.
Marketing's the team thatscrews us up all the time.
but of course that's justfrom his point of view.
(27:55):
Yeah.
I'm sure if I, if I was friendswith somebody in marketing,
I said, well, what's yourperspective on engineering?
Like, oh, engineering's alwayslike, we wanna move slower.
We gotta make sure wedo this the right way.
, like, you know, and, andI think, so there's al
so first of all, there'salways that dynamic.
Right, and that often theshadow organization, you
could pretty cleanly find somecorollary in the relationship
between two functions orbranches of the organization.
(28:16):
So when I think about usuallywhat you'll hear if you talk to
people at the Washington Postor you talk to people at the
New York Times or Vox Media.
It's this tension between,quote, the newsroom and
the product, business,technology organization.
Not that that is, notthat that's a contentious
relationship, but what oftenhappens is, it is the job of
somebody in the newsroom to beable to report the news and to
(28:38):
be able to present that news aseffectively as possible to their
readers as quickly as they can.
if you think about anyorganization, you think about
CNN, you think about WAPO,you think about the New York
Times, The difference between,reporting on, The kinds of
things that are happening inthe world today in the next
two minutes versus the next20 minutes is everything.
(29:02):
So that need to be able to meetthe news and move with the news
is really, really critical.
And then on the engineeringside, if I just approached and
didn't have any background inthat, didn't understand how
important it was to be ableto say, like I've talked to
partners in the newsroom duringan election year and they've
said like, I was like, what kindof availability do you need?
(29:23):
It's at 99.
9, 99.
99, they're like, how about 100?
And you're like, well,what do you mean?
It's like, no, but what if we'redown for 5 minutes when the
Supreme Court makes its ruling?
Or something like that.
We cannot afford any downtime.
The answer is 100%.
You have to be able tohave, you have to be able
to engage in dialogue there.
Conor Bronsdon (29:39):
And that is tension, as you point out.
David Yee (29:41):
That is a tension, right?
Okay, I get it.
We need to be up when theSupreme Court makes its ruling.
That doesn't mean that oursystems need 100 percent uptime.
It means that we haveto find ways to be able
to meet our readers andconsumers in the ways.
that can follow the news,even accounting for the
possibility that some systemor another might be down.
(30:01):
You can design a systemaccordingly with that,
but you have to be ableto have that conversation
and that tension in play.
when I think about, Theunique challenges of media
engineering, it relates to howan organization came to be.
So the example I alwaysgive is that at Vox, Vox
Media more or less startedas a product organization,
(30:25):
started as a technologyorganization, and it really
started with sports blogs.
So we're going to makeit possible for fans of
the Cardinals to be ableto write about games,
alongside the journalistswho cover those games.
and so it was effectivelya product organization.
You add a bunchof sites like Vox.
com and The Verge andEater and all of them.
(30:47):
and so what you effectively haveis newsrooms that are built on
top of a product organization.
The New York Timesis very different.
You're talking about like acentury and a half of journalism
and of a newsroom that is sortof steeped in tradition and
systems and a newsroom thattalks in systems and you've got
a technology organization builton top of that newsroom and
Conor Bronsdon (31:09):
And to the New York Times is credit.
The New York Times is held upas a paragon of a successful
digital transformation.
Yeah.
Into this era.
But to your point,there's this massive.
Cultural foundationthat's already been
laid over 150 years.
David Yee (31:23):
Yeah.
And that culturalfoundation also is an
operational foundation.
And so at Vox you were able tosay, this is how software works,
this is how technology works.
How would you tell thenews in that context?
And in the New York Times,it's like we have built.
A tradition from over a centuryof how to report the news.
(31:44):
Yeah.
So then how doestechnology meet?
That is very, very differenttension and it creates
a very different shadow.
Conor Bronsdon (31:50):
I'm a big fan of the acquired podcast,
which does like breakdownsof businesses and they did
a really interesting episodefor me as someone who, you
know, I'm, I don't work at theNew York Times, but they did
a whole breakdown of the NewYork Times as transformation
into digital organization.
Yeah.
And it was.
It was fascinating to thinkabout, and particularly in
the context now of rethinkingabout it through, uh, the
Shadoworg piece, because I canonly imagine the amount of...
(32:15):
Institutional thought processesand culture and ideas that
have developed and some ofthat has been calcified and
like, this is how we operate.
I'm sure some of it is, youknow, tribal knowledge too
and like, oh, this is how weadjust that and making that
transition to, okay, now we areextremely digitally transformed.
We have applications, we have,you know, made purchases.
(32:37):
We have a massive digitalsubscriber base, must
bring a host of challenges.
David Yee (32:43):
It does.
I mean, it does bringa lot of challenges.
At the same time, I thinkwhat I've witnessed as
this has unfolded is thewillingness of the organization
to meet those challenges.
Which is not to say thatit's easy, or that people
are comfortable with it, butwe have an executive editor
who is relatively new to hisrole, he was sended to that
role, but before that, like,we were working directly
(33:05):
with this executive editor tofigure out What is the role
of technology in telling thestories of the New York Times?
The fact that we are in theposition we are in right now
as a business, with the digitalsubscription model that we have,
is a testament to how flexiblethe organization has been to
be able to say, you know what,we have to be able to figure
out how to integrate thesetwo different cultures, these
(33:25):
two different businesses, andthese two different conceptions
of the New York Times.
So, you know, you asked meearlier like, who's doing it
well, like I'm not, I'm, I'mnot here saying that it's,
that it's perfect or easy.
We're doing pretty well.
We're, I'm just saying,you know, New York Times
reconciles, its shadowpretty straightforwardly.
There's always gonnabe something that's
gonna piss you off.
There's always gonna besomething that gets in the way.
(33:45):
Those, and, and I say thisin the talk, like there is
no perfect reconciliation.
There is no perfect resolution.
The work is ongoing and thetalk is Titled shadow work.
And so it is anconstantly evolving
Conor Bronsdon (33:58):
process.Thank you for taking the time
to dive into this with me.
I, I've really enjoyed theconversation and it has, I
think, opened up some newthought avenues for me.
I I really appreciate youcoming to Upon podcast
to talk about it with us.
If you are a listener and youwanted to follow along with
David, where can you find him?
David Yee (34:17):
Yeah, so in the roiling hellscape of
social media, I've sort ofpulled sharply back from the
company, the company known asTwitter, but like my handle
there and in a lot of placesis tangentialism, which
is like a handle I came upwith when I was in my 20s.
So that's, that's,that is what it is.
So you find me there, but Iactually post a lot on Mastodon.
(34:39):
And so if you're lookingfor me on Mastodon, you can
look for at David at yee,Y E E dot camp, C A M P.
and that's where I domost of my writing now.
Conor Bronsdon (34:47):
Fantastic.
Uh, David, I've reallyenjoyed this conversation.
Thanks so much forcoming on the podcast.
And, uh, if you want moreinsights from leaders
like David, you can checkout our sub stack at dev
opted dot sub stack.com.
David, thanks for coming on.
David Yee (34:59):
Thanks for having me.
Popular Podcasts
United States of Kennedy
United States of Kennedy is a podcast about our cultural fascination with the Kennedy dynasty. Every week, hosts Lyra Smith and George Civeris go into one aspect of the Kennedy story.
Stuff You Should Know
If you've ever wanted to know about champagne, satanism, the Stonewall Uprising, chaos theory, LSD, El Nino, true crime and Rosa Parks, then look no further. Josh and Chuck have you covered.
Dateline NBC
Current and classic episodes, featuring compelling true-crime mysteries, powerful documentaries and in-depth investigations. Follow now to get the latest episodes of Dateline NBC completely free, or subscribe to Dateline Premium for ad-free listening and exclusive bonus content: DatelinePremium.com