00:00welcome to the a 16z podcast
00:03I'm Michael Copeland this segment of the
00:05pod centers on all things logistics how
00:08packages get from point A to point B
00:10though warehousing in between the
00:12security the networking the data the
00:15predictions the whole ball of wax
00:17moderating this discussion is andreessen
00:20horowitz own Jeff Jordan and joining him
00:22are four experts in the space Mike Smith
00:25CEO from stitch fix Laura Duke who runs
00:28enterprise supply chain at Zazzle Mike
00:31Swartz who's a consultant and a longtime
00:34expert in this space and then finally
00:36Dwight Gaston chief delivery officer at
00:38julep Jeff Jordan kicks it off and then
00:41Mike Smith comes back with the first
00:43response for each of you can you spend a
00:47little time describing kind of what your
00:50business does and why your job is hard
00:52you know so well what unique attributes
00:55make it hard and why are you pick
00:57whatever experience you want lob in on
01:00that one yeah okay yeah so our business
01:04is again it's a personalized styling
01:06service for women we use data and
01:07analytics to be able to understand a
01:10woman's style and we send them a box of
01:13five things that should work for them
01:14and they can try it on and would like
01:16highly personal highly personalized
01:18there's not been one fix of all the
01:20fixes that we've sent out that have been
01:21the same so there's always different and
01:24it is it is meant to be able to try it
01:28on the comfort your own home the context
01:30your own closet almost like a fitting
01:32room like a personal fitting room and
01:34you whatever you keep or whatever you
01:38like and you want to buy online you you
01:40buy online and you keep and you send
01:42back what you don't want you get a
01:43prepaid envelope to send back with what
01:46that doesn't work smart operationally no
01:48it's funny that you say that because it
01:51you know I think and it's in its form
01:53sort of sending stuff out getting stuff
01:55back and having it be sort of this whole
01:57reverse logistics thing and a lot of
01:59investors passed on us because they're
02:01like that's a big returns problem and
02:03we're like returns as part of the model
02:04like everyone has returns and we just
02:06sort of operationalize that and it works
02:07for us there's so much of my job today
02:10that seems so less complicated than the
02:13at Walmart because of the range of
02:15product that we had at Walmart from
02:17diamonds to ready-to-assemble furniture
02:20to big-screen TVs two chimneys we sold
02:23chimneys to kayaks I once woke up in the
02:26middle of night thinking we were selling
02:27yachts and I'm glad we never saw that
02:29but there's there is a complexity to the
02:32send-out get back and making sure that
02:34the stuff that comes back is hasn't been
02:37like worn out at a party and so we have
02:40good quality kind of checks as it's
02:43coming back and then the inventory turns
02:45really really fast and so you have to
02:48have a pretty dynamic warehouse and I
02:50think the other complexity of the model
02:52is just managing the growth it's been
02:54crazy growth we've gone from I'll give
02:57you just the warehouse numbers when I
02:58first started the company we did pick
03:01pack and ship out of five hundred five
03:02hundred square feet which was half of
03:05our office then two months later we
03:08moved to a thousand square feet then two
03:11months after that we moved to nine
03:13thousand square feet and now we're in
03:16450 thousand square feet in two
03:19different warehouses so managing that
03:21growth and scale and making sure that
03:24you know you you make really good
03:26decisions about what when you change
03:28your process or think about adding
03:30systems into when you're growing that
03:32much and the business is changing as
03:33much as it has I think has made the job
03:36complex and keeps me up at night
03:40good to hear so and Sue's Azal I work at
03:45dazzle and dazzle is a customized
03:47marketplace so it's a marketplace or a
03:50platform where you can put any of the
03:53world's content on any of the world's
03:55products so you can imagine taking a
03:58picture of your children putting it on
04:00your iPhone case you know taking artwork
04:03and putting it on a print and and really
04:07it's a very dynamic model that has a few
04:11tiers it has it has our makers which are
04:16the outsourced partners as well as our
04:19own internal production which still does
04:21a large percentage of our business and
04:23balanced out by so we have the
04:26and then we also have our contributors
04:28that are contributing art and that can
04:30be anybody from our large partners like
04:32Disney or Marvel to to individual
04:37contributors that are contributing
04:38artwork and they're putting artwork on
04:40on those different products right and
04:43and so so um I think the complicated
04:49part about my job ends up being you know
04:52we have everything from a very large
04:55organization that is shipping that
04:56product to two very small makers that
04:59are potentially only shipping a couple
05:01items a week and we want to create
05:03really an end and customer experience
05:07that is the same for everybody
05:09regardless and so that those small
05:12makers can use on the same logistics
05:16model which really is creating more of
05:19kind of a peer-to-peer a logistics
05:22system that doesn't really exist much
05:23out in the marketplace today so that a
05:26customer or a small maker can ship to a
05:29customer and leverage the contracts and
05:33the technology of the larger platform
05:37and really kind of have this great and
05:40customer experience so it's that part is
05:43pretty complex but my returns is easy
05:45because customizable market doesn't have
05:47much in the way of returns so that just
05:49kind of shows the difference in
05:51different types of operations and
05:53complexities and how different they can
05:56what company you want to pick well you
05:58know I can talk in general I you know I
06:03mean I do a lot of problem solving and I
06:05get called into a lot of situations that
06:08are challenging both culturally and just
06:12the environment so different especially
06:13when I go to different countries and
06:16what's counterintuitive is that it's
06:19really the soft stuff that's hard it's
06:20not the hard stuff that's hard so you
06:23know figuring out the process and
06:25figuring out what needs to get done is
06:26actually the easy part of the job the
06:29hard part is getting the buy-in and
06:31getting people to recognize for
06:33themselves what needs to get done
06:35getting them to acknowledge that there's
06:36a problem and then the execution on that
06:40and I find over and over again that
06:43especially with startup companies that
06:46where they tend to fall short and really
06:47need a lot of support isn't something
06:49that they think they're good at which is
06:51execution so I spend a lot of time if I
06:55look at the total time spent with a
06:58client I spent a fair amount of time
07:01building trust and building rapport a
07:04little bit of time identifying problems
07:07and a lot of time helping them through
07:09the soft skill side now execution by in
07:13the organizational change part of it
07:16that hasn't changed since my Amazon
07:17death is the same same issues there yeah
07:19I bet but that's probably the biggest
07:22challenge that I face right so for me I
07:26think a lot of the current complexity is
07:28just the stage we're at I I joined six
07:31months ago and before that there was
07:32nobody really focused on operations and
07:34so the the the basics just aren't there
07:37so we're everything's breaking so just
07:40trying to figure out how do we build out
07:42our you know call center and get that
07:44you know the the leadership and and
07:46structure in place to handle calls
07:48appropriately how do we deal with
07:50fulfillment we were operating out of a
07:52small fulfillment center about 11,000
07:54square feet in Seattle and no systems it
07:58was just a really really really
08:02struggling so we made the decision on my
08:05second day there to outsource
08:07fulfillment and so now we're shipping
08:09out of a third-party fulfillment yeah
08:11partner in in Columbus and so that's a
08:15big big time suck right now but
08:20longer-term I think what's gonna be
08:22really fun to really figure out is is
08:27the supply chain because we've got we're
08:30generating about 30 products a month
08:33that are brand new and some of those are
08:35just new nail colors but but a lot of
08:36those are new they're cosmetic products
08:39and which by the way is revolutionary in
08:42the beauty business typically Beauty
08:44companies do two sets a year maybe
08:46because they you know they would not
08:48need to manufacture it need to support
08:50it with marketing you know just so done
08:52it's a beautiful target for digital
08:54disruption right so it's so it's it's
08:56it's a huge part of our model but it's
08:58we just need to make sure we're we
09:00continue to get it right and the other
09:02challenge with that is that we have a
09:04hard deadline so every month we put our
09:06new products into our subscription box
09:07and then every month you know by the
09:10middle of the month we've got half that
09:11new product in our warehouse ready to
09:14ship out and and so that is just you
09:18know fraught with you know risk so a lot
09:21of nail polish flew first-class before
09:23you came trying to put it back on the
09:29yeah my favourite jewel is the community
09:32is a rabbit community in there and
09:33they're just wait they've ordered their
09:35new you know their customized their box
09:36at the last minute and then they just
09:38want it three days later and one thing
09:41that happens is nail polish technically
09:43is a hazmat I guess and so they found
09:45out what was using them air to deliver
09:48and they went crazy and and stopped them
09:51you know no no you can't use air you
09:52have to drive it and it says all of a
09:54sudden all the mavens who are waiting
09:56for their product or you know you were
09:58it's not getting there in time because
09:59they you know the Postmaster General
10:01said he could they can't fly it was a
10:04disaster that was day four as a board
10:09member I thought yes yeah so it turns
10:14out first-class mail which is a great
10:16way to get things across the country
10:18from Seattle at a low cost at low cost
10:21suddenly was no longer feasible so we
10:25had to ship our next subscription around
10:27using our other low-cost method which is
10:30DHL Global Mail which takes up to 10
10:33days to get across the country so we had
10:36what this fir this like rabid customer
10:40base was just up in arms they actually
10:42started a petition on change.org
10:46demanding their julep breakup with DHL
10:48so yeah we couldn't get it to get our
10:52business district to Ohio soon enough so
10:56a lot of you came from fairly large
10:59enterprises and joined your enterprises
11:01when they were fairly young talk about
11:05for me it was wasn't that hard I mean
11:08again my exit was a little complicated
11:10but most of the things I'd worked on at
11:12walmart.com again I joined when we were
11:14about a hundred people we had just done
11:16one hundred fifty million dollars in
11:17revenue I left it was a lot bigger but
11:21most of what I worked on were small
11:23company things and I just realized after
11:26nine years they're sort of getting like
11:28Mike talks about getting things done
11:30through influence and change management
11:31that I had had enough I really wanted to
11:35have more direct impact the you know
11:38sort of warehouse staff at walmart.com
11:41versus where I was in my position there
11:44was at least eight layers and so I was
11:46being held accountable to things that
11:48you know what happened in a warehouse
11:50that you know they didn't even know my
11:52name and I didn't really know their name
11:54and that became more challenging so it
11:57wasn't a big adjustment to go small for
12:00me I took a huge pay cut that was
12:03probably the the biggest adjustment that
12:05I had done the math and said I could
12:07make it work and I made it work but boy
12:10the first couple months of paychecks I'm
12:12a big wine person and so not being able
12:16to buy my wine was a little boxes but
12:29but you know sort of sitting behind like
12:32I sat right next to our CTO for the
12:34first six months and so if something was
12:36breaking in the warehouse operation and
12:39the systems I'm like Jeff we got to get
12:41this fixed and that was so empowering
12:43and and much better than you know trying
12:46to get something on the product roadmap
12:48a Walmart comm and waiting for six
12:50months to a year or longer to actually
12:52see it happen so it wasn't a big
12:55adjustment for me it's pretty it was
12:56pretty easy to the big adjustment for
12:58anyone else I think there was there was
13:01adjustments I mean if anybody a an HP or
13:05anybody vid has gone to a start-up I
13:07mean you guys all know it's not all
13:09sunshine and unicorns right I mean
13:11there's challenges and I think I think
13:13ironically the biggest challenge for me
13:15was actually what I went looking for
13:17which was creativity now granted I
13:19I left and I left eBay I came to Zazzle
13:22and there was only 30-something
13:24employees and you know a few million in
13:27revenue and it looked a lot more like
13:29chaos than creativity to me so and I
13:32think that adjustment of being able to
13:36you know even though I had this kind of
13:38mindset in this big company and working
13:44through significant processes that you
13:47know we're well ingrained and a lot of
13:48them were holding us back and slowing us
13:50down but at the same time I could really
13:54had to kind of sit back resist the urge
13:56to fix everything and actually kind of
13:59really recognize the creativity and
14:02build on that now I do think that
14:05there's a fine line between cake a
14:07creativity and chaos and I think kind of
14:10you know managing that line still takes
14:13some finesse but but there's there's a
14:17you know dazzles a lot bigger now but
14:20but we're still kind of heart and soul
14:22startup and I can't say that the chaos
14:24has gone from my life but but I
14:27definitely have found an ability to kind
14:30of eliminate those chaotic elements and
14:32kind of focus on the creative side
14:33legged white when you first show up
14:35small company there's no team so how do
14:39you go about building a team and then
14:41you know particularly how do you balance
14:44the rewarding the people who brought you
14:46to the party you know you're the young
14:48aggressive you know I you know bringing
14:52in talent from the outside you know I
14:54think one of the I was thinking about
14:56some of the challenges as kind of goes
14:57with your first question is the skills
15:01and the people that got you there won't
15:03get you to the next level you know
15:05recognizing that and I think having
15:07conversations with people early is a big
15:10part of that challenge you know that it
15:12takes a certain type of person in the
15:14early stages of startup to succeed but
15:17that person if you're on a really high
15:20trajectory growth curve they're not
15:23going to be necessarily the same people
15:24that take you to the next level and you
15:26can risk losing some really good people
15:28if you don't talk to them early about
15:30how their role is going to change
15:32and how they're gonna have to go maybe
15:34more narrow and deep and not be broad in
15:36their responsibilities all they see is
15:39I'm losing control and losing control
15:41and I feel less impactful so I think
15:44Pete I think people in general is the
15:48biggest shock you know even in the the
15:51time that I spent at Amazon and QVC was
15:54still very entrepreneurial early-stage
15:56kind of companies they were growing fast
15:58and I underestimated how much time you
16:01have to spend recruiting and hiring you
16:04know I would spend 80% of my day
16:07sometimes interviewing and trying to
16:09find the right people and even at Amazon
16:11after my facility with staff we started
16:13we had a 5,000,000 square foot expansion
16:15in 1999 and I was recruiting for other
16:21sites so I would recruit all day long
16:23and then do my job you know from like
16:266:00 to midnight I'd do my regular job
16:29so I think that was that was really a
16:32big stressful part of it you know is
16:35putting the time into hiring and finding
16:37the right people making sure you don't
16:39lose the good people that are there and
16:41helping them understand what their role
16:42is gonna be going forward and then doing
16:45your day job yeah it sounds like you
16:47just walked into a green field I did
16:49yeah yeah and so for for me my I had a
16:53team in each air each of the three areas
16:56and or I have and so it wasn't so much
16:59that I needed to you know build a team
17:01from scratch so I so that was nice
17:03because I've been able to kind of
17:04evaluate people in customer service the
17:09team was so incredibly junior that it
17:11was just there was a clear need to bring
17:12somebody in and the nice part was
17:13everybody loved that the the junior
17:16people who are trying to you know deal
17:18with this you know chaotic situation
17:20suddenly have this you know role model
17:22and mentor so that was that was great
17:23was an easy decision to bring somebody
17:24in because it was clearly necessary and
17:26it wasn't disruptive and for the other
17:30teams what I'm trying to do is figure
17:32out who should stay who can take
17:34different roles and then others in the
17:36organization have have been coming
17:37expressing interest in joining
17:39operations so there's kind of a I'm
17:41trying to not make decisions until I
17:43kind of suss out kind of what what
17:45there are so it's try not to you know
17:49you know change things dramatically by
17:53bringing in a bunch of new people
17:54because that can challenge us talk
17:57damage Ben Horowitz has written a blog
17:59that I really like about how every in
18:02these hyper growth companies that job
18:03requirements change so fundamentally so
18:06quickly that basically everyone in the
18:08organization has to almost reapply for
18:10their job each year you know just it's
18:12that dynamic so that's something that
18:14stay with me for a while how do you how
18:16do you guys could define success in your
18:17operations I mean just what what is what
18:19a success had how are you held
18:21accountable how do you hold your teams
18:22accountable I mean I think an operation
18:29everyone doesn't have to answer every
18:31question this might be oversimplifying
18:34but in operations that I think it all
18:36boils down to either you know cost
18:37control doing things faster or doing
18:41things at higher quality and so it's
18:43determining for what each area what are
18:45the priorities I mean all of those are
18:47good but really deciding which is most
18:51important and then figuring out the
18:53right metrics and and and yeah just a
18:56couple metrics can be really helpful to
18:57focus everybody but for me it's if it's
19:00just setting those and then having
19:02everybody really focus on managing those
19:04and in dealing with it as we grow their
19:07trillion metrics what do you think so
19:10for us I think we've got the call center
19:14we wanted to measure customer
19:15satisfaction and then service level and
19:19those are the two things that are mostly
19:22efficiency and quality right and then in
19:28operations or in fulfillment it would be
19:30on time shipping and then cost per order
19:32to manage that and then we're still
19:36figuring out supply chain I you know I
19:40get asked this question a lot and and
19:42the question usually gets phrases well
19:44what what do they do at Amazon or what's
19:47Amazon's metric or what's best in class
19:50in XY and Z and my response is always
19:53the same it's like why do you care right
19:55so what's really important is I think
19:57the measure of success is
19:59do you do you understand what's driving
20:01your business because what's driving
20:02yours might be different than what's
20:04driving Amazon or another company to is
20:07are you doing better today than you are
20:09than you were yesterday and three how
20:13aligned are the people in the
20:14organization around these two two
20:16fundamental questions one is what's
20:20important if you ask that of everybody
20:22in the organization from the founder all
20:24the way down to the guy picking and
20:26packing the order you should have
20:28alignment and often you don't when you
20:31ask what's important and two is how are
20:34you doing right now and if people can't
20:37answer those two questions you can have
20:38a lot of chaos right so I I define
20:42success as do you have alignment and are
20:45you doing better each day and or you are
20:48you taking steps to understand and
20:50reduce variation in each of those
20:52metrics so it's not the mean you know on
20:54average I can have a hand in a bucket of
20:57ice and one and a bucket of hot coals on
20:59average I'm comfortable you know that's
21:01and even well-established companies I
21:05work with don't understand variation
21:07they do everything off of means the only
21:10other thing that I died from a metrics
21:12perspective and is that it's it's not
21:16only what you measure but when you
21:17measure it so if you're looking at let's
21:20say you're looking at your metrics
21:21monthly and you're a start-up you're
21:24moving so quickly that's like driving
21:26down the road at 90 miles per hour and
21:27only looking in your rear-view mirror
21:29you know you've really got to look at
21:31the metrics as current as possible
21:33and even start to find metrics that are
21:35leading indicators you know so when
21:37you're looking at on-time delivery not
21:39just looking at on-time delivery also
21:40looking at backlog which might tell you
21:42what delivery but what delivery times
21:45are going to be in the future and what
21:46customer enquiries are going to be in
21:47the future so that you can start looking
21:49at metrics that actually start to look a
21:52little forward and as current as
21:54possible rather than just waiting a
21:56month when the damage is already done
21:59what kind of mistakes do you see
22:01startups making on in the operational
22:04front I mean what are you know pattern
22:07recognition says well it do I gave up a
22:13not focus on it fast enough I think I
22:15mean I got lucky in that you know
22:18Katrina the founder and sort of the
22:20investors early stage said this needs to
22:23be a core competencies because we're
22:24sending stuff out and getting stuff back
22:26and I mean it's shocking to me now and
22:29I'm very grateful but they sort of hired
22:31a CEO Oh as employee number four you
22:34know and that's very early to do that
22:37but I think it's not focusing on it fast
22:40enough and then you know there's times I
22:43think when startups and even bigger
22:45companies where you start to lose like
22:47the senior people start to lose
22:48connection with the work that's being
22:51done and I think that's a mistake too
22:53there's got to be ways to that you're
22:55always staying connected so you
22:56understand both on the like what Mike
22:58said on the software side what kind of
23:00culture do I have and are we you know do
23:04they understand sort of a vision the
23:05purpose of what you're trying to
23:06accomplish every day because those jobs
23:08are really hard and keeping them
23:09motivated to do them and have low
23:11turnover is an important part of the job
23:13and I'd say the other part is just like
23:17understanding where you have process
23:18flaws or productivity you know we're not
23:21meeting productivity goals so you need
23:23to stay close to it and I think there
23:25are some leaders that like it's not my
23:27problem like it's my you know supervisor
23:29on the floors problem so I always walk
23:32the floor and say it's close to it it's
23:35not going to come back to that one okay
23:36other no I think it's really important
23:39other other common yeah I think one
23:45thing that I've seen is not negotiating
23:50so so when you have relationships
23:53there's a kind of this with startups
23:54because we're moving so fast is that you
23:56get this kind of grab-and-go philosophy
23:58as far as setting up vendors and
24:00suppliers and big huge partners that
24:02you're gonna try to get something in
24:04place really quickly and put it in place
24:07so you can focus on all the bigger
24:08issues and and there's a lot of times
24:11you think you're small you don't have a
24:13lot of volume there's not a lot of
24:14leverage there for negotiation in the
24:16first place but I think what we're
24:18missing is the fact that even the
24:19biggest partners know that the value of
24:23startups in this economy and they can
24:26years out so if you just take the time
24:28to get the right people in the room and
24:30actually kind of talk through your
24:33vision you can ask for what the big guys
24:35get and you'll be surprised and it
24:37doesn't take much time and you'll get
24:40significant payback in terms of margins
24:42and customer experience today twice a
24:46year meeting with ups and then just talk
24:49to them but how much how big how big
24:51you've gotten since the last time you
24:52were here and it is for a fairly
24:54remarkable yeah I to thing I even in
24:57even in early-stage startup companies
24:59two things that come to mind one is
25:02really understanding and staying focused
25:05on your customer who are they you'd be
25:08surprised how often I ask who the
25:10customer is and I get 50 different
25:12answers so getting clear about your
25:15customers and making sure that
25:16everything you're doing is lined aligned
25:18with their needs and the second thing is
25:20that I find quite often is a poor or
25:26maybe the wrong attitude about defects
25:28about what's you know things that are
25:30wrong people tend to hide them and they
25:33get in trouble if defects are present
25:35and they're always there right I mean
25:37nobody's perfect so I think that's a
25:40fatal mistake early is not having a not
25:44not putting systems in place and metrics
25:47and really embracing problems for what
25:49they are you know when you drive them
25:52down you don't see them you can't do
25:53anything about them so I think
25:55developing a strong culture or attitude
25:57toward making defects public making them
26:00something everyone can Center on is a
26:03very powerful thing yeah and when
26:05defects occur making sure that the root
26:08cause analysis is not attacking the
26:10people involved it's attacking the
26:11process is that it's so important to
26:14make it safe to raise these mistakes and
26:17talk about things if you think the fix
26:18the root cause is about fixing the
26:19process it's not about you know somebody
26:21making us a mistake and if you have that
26:22culture you can really drive much of
26:25that resume it movement going yeah so
26:27putting a mirror up any this any
26:30personal mistakes you've made along the
26:31way in their career that you think might
26:33be illustrative for people to learn from
26:37list yeah I've got a really direct one I
26:41I got on the phone with a customer when
26:43I was not in them but not in a good
26:46place to be talking to a customer I was
26:49to sort of logical rational about
26:51dealing with what they were asking for
26:53and I it wasn't good I mean it almost
26:56turned into sort of a big PR issue for
26:59Walmart and so I was glad that dish fix
27:04the opportunity came along yeah so I
27:10think yeah I was it was it was a
27:12challenging situation it was you know I
27:14just I wasn't trained to handle an
27:17escalated call at that level and I
27:19should have opted out yeah so I think
27:23over delegating is one mistake I made
27:25once and it was pretty painful and it's
27:28it's it relates to like understanding as
27:31you're growing fast who's kind of
27:33keeping up with the needs of the company
27:35and who's falling behind and we were
27:37doing a big expansion in our fulfillment
27:40center Blue Nile and we needed to buy a
27:43second vault we have these big very
27:47secure balls we've built little but we
27:49built and long story short the person I
27:52put in charge of that didn't you know do
27:54a lot of research and the salesperson
27:55didn't do a lot of research and they
27:57bought a vault that would have fallen
27:58through the floor of our warehouse and
28:01and what was the problem with that I
28:05discovered this when it when the 25
28:08trucks were driving from Colorado with
28:10all the pieces and and a really ugly
28:12negotiation with the vault manufacturer
28:14we had it sitting outside our warehouse
28:16for six months while they tried to find
28:18another buyer so it was a nice reminder
28:20of this huge expensive mistake which is
28:23always good in Seattle doesn't rest I
28:29think what am I the one lessons that
28:31come to mind very quickly is you know I
28:34used to have an adversarial
28:36relationships with finance right and to
28:38the point where I'd call them CF knows
28:40instead of CFOs and I learned early on
28:44that it's better to involve them early
28:46and get them to partner in terms of
28:49coming up with what your true call
28:51and and making them part of the solution
28:55rather than somebody that you present to
28:57at the end to debate with you and that
29:01was very helpful on a lot of levels
29:03because they were more involved I think
29:06we got better results yeah absolutely
29:08and it was better for the business some
29:14companies almost take operations for
29:15granted it's a bit out of sight out of
29:17mind you know it just it it happens till
29:22what are great what are some mechanisms
29:23to try to keep you know they have the
29:27entire company keep at that operational
29:30vantage point and you know thoughts in
29:32mind that it's a holistic company that
29:35you know and not a detached you know
29:38work still doesn't work for me two
29:43things come to mind one is just making
29:45sure that at sort of the executive table
29:47that you are seen at the same sort of
29:51light and peer is like the chief
29:53merchant or chief marketing officer a
29:55lot of that is like your own leadership
29:58style and and how you present how
30:01important your group is not in a
30:03political self-promoting kind of way but
30:05more just how this is impacting the
30:07business and why it matters that's one
30:10and then a very tangible one that I've
30:12I've wanted to do in my old job and have
30:14done at stitch fix is having customer
30:17service agents that are always in
30:20headquarters so that people under city
30:22monks the business so they can everyone
30:25can hear defects and exceptions and
30:28things that aren't working so that also
30:29keeps it front and center that feedback
30:32loop is critical yeah I think I think
30:34another thing that has worked for me
30:36personally is is having some discussions
30:39around a shared metric using something
30:41like gross margin or gross margin isn't
30:43owned by one group alone it's owned by
30:45so many and having all the department
30:48heads and key stakeholders in
30:50discussions so that marketing is they're
30:53trying to hit their revenue target they
30:55can have that discussion and say I you
30:57know I want to send out this promotion
30:59with this discount and operations is
31:02right there in the room to say okay well
31:05does in terms of capacity maybe we could
31:07talk about using a different product
31:08that we have more capacity on things
31:10like that so you can have those kind of
31:12cross discussions in the same room and
31:15they can understand the challenges and
31:17then they get to considering those as
31:21they're driving the business even when
31:23they're not in the same room no I'm
31:26gonna echo one of the things Mike said I
31:27think transparency is huge and if you
31:30look at call centers a lot of companies
31:32look at call centers or contact centers
31:34like they're a cost center and really
31:37call centers don't generate calls
31:39they're responding to things that go
31:41wrong somewhere else in the company so I
31:43think creating transparency there is
31:45huge the other thing is finding
31:48opportunities even as you get very large
31:51for everyone to participate in what goes
31:54on in operation so during holiday Peaks
31:56making sure that they be that inviting
32:00people from the front office that
32:01technology people especially invite them
32:05into the warehouse invite them into
32:07operations make sure that your
32:09merchandising team that somebody's there
32:11to physically touch and see the products
32:13and how it's coming in from the vendors
32:15at least one day a week making sure you
32:19just kind of keep the doors to your
32:21operations open is huge I know when I
32:24was at Amazon more code got written in
32:26the fourth quarter when we would get
32:28deployed to the fulfillment centers than
32:30any other time of the year more things
32:32were fixed so very valuable to keep
32:35those I laugh only because be careful
32:38what you ask for and how you implement
32:40because I can tell you our CFO was
32:41really bad at making ties so when we do
32:45that we just have to be very careful
32:47about what what you give them yeah I got
32:51a good quick story on this point too so
32:54it was a Friday afternoon and one of our
32:57chief engineers was in town and he went
32:59down to the warehouse and just wanted to
33:02kind of pick pack and ship and do some
33:04returns he came to me at like 5:30 that
33:07day and was really upset that we had put
33:10you know these associates through a
33:12really bad very manual returns process
33:17had RF gun that he had ordered by Sunday
33:21he had written code that basically could
33:23return process returns in a system y way
33:26Monday we rolled it to the whole
33:28warehouse at that point we only had
33:30probably eight people working returns
33:31but when you get that level of
33:33engagement you actually get to see it
33:34there's huge returns and he just thought
33:37I was the worst operator ever for
33:39letting them go through that for even
33:40you know for six weeks and there's just
33:43huge dividends for take it being
33:45transparent take it to the floor don't
33:47take it out of the conference room
33:48yeah open up for questions after this
33:51last one um security privacy brave new
33:55world out there has that what
33:59complications does that bring so Blue
34:01Nile we had a big we were a big target
34:04for fraud so credit card fraud and is
34:07that kind of yep you're yeah and so we
34:10really had to be world-class at that and
34:13there are a lot of tools out there now
34:14but wanted to steal diamonds they did no
34:17good and so for us I think most
34:23companies have a kind of automated if
34:25they have a fraud prevention process
34:27there's an automated screen where orders
34:28go through and kind of get a score to
34:29see you know how risky they are and we
34:32had that set to be very very sensitive
34:34so that almost all of our orders would
34:36get kicked into a manual review team and
34:38so we had four to five people who would
34:41look at these orders and they would have
34:43they had had some ad hoc tools to kind
34:46of do some further research we would
34:49sometimes call the cardholder sometimes
34:51we would find a different number to call
34:56that person to make sure that we didn't
34:57call the number that the person who's
35:00ordered it gave we so we just we had a
35:05whole very you know kind of hardcore
35:09approach but I would say that there are
35:10a lot of tools out there now third-party
35:12tools that can help you configure
35:16configure the the rules and such to
35:18match your business needs yeah I have
35:22two comments I mean I ran a jewelry
35:24vault for many years and you know always
35:27had you know between 150 and 250 million
35:30Nitori so I used to joke that you know
35:32my job was to walk around everyday
35:34thinking like a thief you know how would
35:36I steal and even though that was my
35:39those are my times at QVC even though
35:41the company invested millions in
35:43surveillance tools and all kinds of seen
35:45and unseen cameras we had metal
35:49detectors that would weigh your both
35:50your ferrous and non-ferrous metal and
35:52your body content and it was programmed
35:54into your badge and even with all those
35:57tools the most effective tools were
35:59getting whoa program in your badge
36:01yeah so when you were when you were site
36:04when you got your access badge your base
36:07metallic weight was programmed into your
36:09card so if you were over or under you
36:11would go through a secondary screen and
36:13it the structure had to be three feet
36:16off the ground it couldn't be near
36:18swinging doors it was crazy but even
36:21with all that money and proximity alarms
36:24on doors you couldn't just hit an exit
36:27there were delays and all sorts of all
36:30sorts of things but the most effective
36:32security systems were people and because
36:35nothing does a better job of pattern
36:38recognition than a human being
36:39so having tools for humans to use and
36:43really relying on people systems is
36:46probably the best way to secure a
36:48facility and the other thing I took away
36:51from from that was not designing your
36:55security systems and your hooks for the
36:5799% that aren't a problem right you
37:01don't want to punish 99% of your
37:03customers for the 1% that's going to get
37:05you and you know kind of keeping that in
37:07mind that you don't want to over
37:10engineer it so that you're creating a
37:11bad experience for your customer right
37:14on the network security side Blue Nile
37:17used to have hackers would hire hackers
37:20to come in and try to you know get
37:22through our systems and that was a real
37:23helpful way to get a good yeah there was
37:27only being started now marketplaces for
37:29white hat hackers I'm gonna come and
37:31pound on your on your infrastructure I
37:33mean smart guys and even PayPal I mean
37:35hundreds of millions of dollars
37:36we're running out the door and so it was
37:39it was a hand-to-hand combat yeah it was
37:42then would that open it up to the floor
37:44questions um so I guess there's one
37:48trend in e-commerce that is about
37:50customization there's another trend
37:51clearly on having your stuff same-day
37:53doesn't really go together we do both
37:55I'm at HP and if we customize it after
37:57we get it to you tomorrow
37:58how do you think about that from a
38:00customer's perspective that's a worried
38:01about the supply chain but how do you
38:03determine customer value I can give you
38:04a choice but having a customer choose
38:06doesn't always necessarily work either
38:09so I think I think that we have so we
38:17have that dilemma all the time and you
38:19know the immediate kind of instant
38:22gratification of this current on-demand
38:23economy and everybody is is an Amazon
38:28Prime customer and is used to getting
38:30things you know next day two days and
38:32sometimes when you're actually making it
38:35it takes longer than that so I think
38:37we've done and from a customer
38:40perspective I think we have found that
38:41it's a lot about expectation setting so
38:45it's about getting out there and really
38:47letting them know hey it's it's um it's
38:50so much storytelling at this point what
38:52we're really getting to is now is saying
38:55this is coming from this maker and you
38:58know it's a small organization in Kenya
39:00that is going to be making this shirt
39:02and that is what you're going to get so
39:05um that storytelling is something that
39:07really you know we've seen this happen
39:09in a lot of industries we've seen this
39:11happen in food from the you know where
39:16people want something that's locally
39:18grown and then moved to their table and
39:20and that kind of storytelling you can
39:23really latch on to and the customer gets
39:26that so if you use that to really
39:28explain here's some of the things that
39:30you're getting and where you're getting
39:31it from you can then explain the
39:34extended the extended time of
39:37customization a little bit better but at
39:40the same time so one is setting
39:43expectations and explaining it but then
39:45the other is shortening those times so
39:47we actively every year and constantly
39:53are working on shortening our lead times
39:55for all of our products
39:56so that we can actually produce things
39:59with most of them within 24 hours a lot
40:02of them same day and get them and there
40:04are a lot of products that we actually
40:06do deliver next day at this point so and
40:09I think it's striving to do a little bit
40:11of both so HP can say you manufacturing
40:15your computers and small villages and I
40:19think I think it's also creating some
40:22visibility into your supply chain and
40:24letting them see how it's progressing
40:26it's you know they don't want to just
40:28wait and have it you know magically show
40:30up on their doorstep if it's going
40:32through different phases at appropriate
40:34you know without overwhelming them yeah
40:37I was gonna say I mean I think that's
40:39the best app and I think sort of setting
40:41expectations and turning your supply
40:43chain out and showing the customer I
40:46know it sounds crazy those is the best
40:48app for sharing it because I have a on
40:54our product roadmap we have a Domino's
40:56you know Pizza Tracker stitch fix go
41:00totally I do think it's about setting
41:03expectations and I do think they are
41:05misaligned and I'm not a huge believer
41:08in same-day I am a very contrary and I
41:11feel like on the same day I just it's
41:14hard for me we did a same-day test in
41:152003 2004 we almost invested in a same
41:20dating company way back then they had
41:22already tested a lot of different
41:23products I don't buy that there's a lot
41:25of product demand for same-day I think
41:28that Amazon will do same-day for grocery
41:30because it makes sense on fresh I don't
41:32I believe they'll do next day on
41:34everything else yeah so or you could do
41:36Apple and just promise four week
41:37delivery yeah yeah I had the question
41:43regarding the leadership style everybody
41:46worked on big companies and then you
41:47guys moved to the startup environment
41:50how did you change your leadership style
41:52in this new environment
41:54if you did and when sorry quick cuz I
42:00was gonna talk about this earlier so
42:01it's the top of mine I had to change a
42:04lot because I was managing to folks
42:07initially were really small company that
42:1022 or 23 years old and I think they when
42:14they heard the CEO of Walmart calm was
42:16joining the company they thought that I
42:18knew everything and I have spent a lot
42:21of time telling them and it's hard
42:23because it's like you know I'm telling
42:24you that I don't know everything I'm
42:26telling you also I'm not an idiot so
42:28like there's a balance but I want you to
42:31challenge me I want you to sort of ask
42:34good questions and make sure we're
42:36making good decisions but when we make a
42:38decision you need to trust that I know
42:39enough that this is the right decision
42:41at Walmart I was used to managing VP
42:45level people that you know they
42:46challenged me and that was just the way
42:48it was I had to train and teach really
42:51young people how to be challenging in
42:54the right way so is it was it definitely
42:56a style to my experience when you're
43:00running big functions of big companies
43:01you actually forget how to be proactive
43:03you are reacting constantly and the
43:06level amount of stimuli you're getting
43:08is just keeping the plate spinning is a
43:10major accomplishment you show up at
43:12startup and you know it's all about
43:13proactive you know and so when I went
43:16from PayPal to open table I was like hey
43:18someone should run that you know do the
43:19financial planning team should run a
43:21model on this crickets you're like oh
43:24I'm back on Excel I mean I did that no
43:27that's a tough that's a tough enjoy you
43:30know at the risk of sound like a
43:31contrarian I I would say my leadership
43:33style didn't change that but you know I
43:37get this question a lot from guys should
43:39I be should I be a directive or should I
43:43be a collaborative kind of leader and
43:47you know I think it's situational and
43:49you have and maybe in a start-up
43:51environment you have to be more
43:52directive sometimes in collaborative or
43:55I think it's just recognizing when which
43:57approach is most effective you may have
44:01a bias toward one way or another but you
44:04know if the house is burning you don't
44:05have time to call a meeting and say what
44:07does everybody think which exit should
44:08we use you know you got to get out of
44:10the burning building so you know I would
44:14say personally from my own perspective
44:16you know I've always been very hands-on
44:18I've always gone to where the problems
44:20are I've always tried to communicate
44:23and I'm always direct you know if I took
44:26it I've played a lot of athletics and I
44:28think I learned a lot from the coaches I
44:30had I never loved my coaches but I
44:33respected them because they always told
44:35you if you're doing well
44:36quickly they told you for doing badly
44:38quickly and if you did something that
44:40was going to hurt the team you heard
44:42right away right too often I think
44:44leaders lose that perspective and they
44:48only give that kind of feedback when
44:50it's time for the annual appraisal or
44:51time for the annual performance review
44:54or semiannual I think it's important to
44:57do that every single day when you do
44:59this this is what happens these are the
45:01consequences don't do this do more of
45:03that you know it's pretty basic
45:05I would agree it's situational for me it
45:08was it was my concern coming into a
45:10company after being somewhere for 15
45:12years was like how do i balance the need
45:15to learn before you know deciding or
45:17recommending things versus the need to
45:20get things done and for me I quickly
45:22realized I was - been there done that
45:24guy for customer service and fulfillment
45:26so that was easy to just go and then
45:29I've been more in the question-asking
45:31mode on the on the other areas so trying
45:33to kind of adjust that way yeah if you
45:37look 10 years out when it comes to
45:39supply chain in ops what do you think is
45:41going to be the biggest change that will
45:43hit your functional area biggest change
45:50that's I think everything is changing
45:52right now that's what's so exciting
45:53about supply chain in operations right
45:55now is that in this kind of on-demand
45:57economy it's a complete blank canvas for
46:00new operations and creative operations
46:02solutions and I mean if you think about
46:05a lot of the operation solutions a lot
46:08of the methodology was built back in you
46:12know their 80s and 90s of the mass
46:14production retail era and we've been
46:18kind of working off of that and so I
46:20think that the changes that are
46:21happening right now are amazing you know
46:24from peer to peer and logistics - to
46:28personal shopping to same-day delivery
46:30to drones you know there's just there's
46:33a lot of things that are changing so I
46:35don't know that I can say what I think
46:37to be I just think that what's happening
46:40is that from an Operations perspective
46:41rather than seeing operations kind of
46:44behind the scenes doing the optimization
46:46we're starting to see operations come
46:48out and really drive revenue and be
46:51entire business models and I think it's
46:56gearing up to be a major disruption like
46:58kind of the next major disruption I
47:00think in the commerce space that I don't
47:02know that I can nail it down to what the
47:04biggest thing is gonna be that I think I
47:06think there's going to be a lot of them
47:07I think that's why maybe somebody else
47:09has a has a nail I think if I knew I
47:11wouldn't be sitting here I'd be getting
47:15funding for the next big thing I do
47:18think what's what a trend that you see
47:20happening is that smaller startups are
47:24starting to are enabled or being enabled
47:27to provide some of the same levels of
47:29service that larger more established
47:31companies I've only been privy to in the
47:34past so you know everything from the
47:36technologies and material handling
47:38systems having you know cloud-based
47:43telephone systems you know used to have
47:45to have a big box in your warehouse now
47:48you can just subscribe to the service
47:49and I think all those kinds of
47:51disruptive technologies are making it
47:53easier to provide first-class service
47:56without having to have really deep
47:59pockets so I think that's a trend which
48:02i think is kind of exciting it's a you
48:04know I think the Amazon's of the world
48:07are going to have a lot of competition
48:09from smaller startups on the service
48:11front again physical goods lead times
48:14have to get shorter so in jewelry and in
48:17in beauty products we're just dealing
48:19with this old-school industry that
48:22operates on the model of you know it's
48:25going to take six months and the the
48:28retailers are all you know just just
48:30trying to act as fast as possible it
48:33just doesn't work like we're trying to
48:34disrupt our space and and we're gonna be
48:37pushing our suppliers to figure out how
48:38to get faster and faster and I imagine
48:40others are too and I just feel like you
48:42know that the world of a you know
48:4518-month product launch cycle just can't
48:49can't maintain itself if something's got
48:51a got a break I think it's kind of
48:52forced companies in through that entire
48:54supply chain to get more flexible to
48:56getting more nimble and then they'll
48:58know they'll be able to win I mean all
49:01the three companies up here are all
49:03personalized yeah they're sending out a
49:05personalized product Shutterfly's in the
49:07house sending out a personalized product
49:09we're seeing a trend towards not you
49:11know it or not not towards mask ops
49:13customization but the ends of one you
49:15know more and more business models are
49:17being pitched to us are you know
49:18personalized service to every consumer
49:20and so that has massive operational
49:24shout it out patent troll if you leave
49:39will still address the patent trolls are
49:45sort of a pain point that blue nile we
49:48dealt with you know IP you know
49:53litigation threats and at julep we got
49:57our first letter so that's top of mind
49:59because it's current but that just seems
50:01like a you know an ongoing problem sorry
50:11the love the people marketplace things
50:13we work with our running into a big
50:15labor law you know Daniel visit a
50:18contractor as an employee you what
50:19you're competing with the union labor
50:21force so that is seeing a lot more
50:22though yeah more on the infrastructure
50:27said what systems are you using that
50:29maybe other companies don't know about
50:30you know something beyond your ERP or
50:32what are you doing to enable your
50:33operations pieces that you think we
50:35ought to be looking at I don't have
50:42we're building at all ourselves and so
50:45we use Zendesk for customer service but
50:48I'd say that the tie-ins to Zendesk are
50:51very proprietary in terms of like how
50:53tickets how we think about tickets
50:56related to customer segmentation or how
50:58we think about tickets related to
51:02there's more accountable to better
51:03operations but it's not that we're so
51:06you know it has to be us that builds it
51:10it's just that when we have researched
51:12outside parties to talk to us about how
51:16they can fit into our model our model is
51:18funky enough and different enough that
51:20it doesn't fit and so we've just sort of
51:22said we'll listen but likely have to
51:25build our own stuff to run the business
51:27the way we want to run it and that's
51:30really empowering to engineers I don't
51:34like one of the challenges like we've
51:35had in our businesses is getting in
51:37front of engineers because it's a
51:40women's fashion company and that's just
51:41not where we're getting as much kind of
51:44in the front end of the funnel people
51:46coming through but once we show
51:49engineers what we're doing there's a lot
51:52of excitement because of the systems
51:54that we're building and also just the
51:56data and analytics and how engineers can
51:58work with the data science seem in a
52:00very proprietary and cool way so it's
52:04for us a competitive advantage once we
52:06get in front of an engineer building our
52:08own stuff I'm seeing a lot of that in my
52:11codes there's a lot more is built and
52:14bought in and just because the models
52:16are different you know it's a big
52:20commitment though once you decide to go
52:22that route now you're now you're saddled
52:24with maintaining it and that's that's a
52:27global issue you know finding platforms
52:31that handle most of your needs and
52:34resisting either resisting or going both
52:36feet in on owning the code is a big
52:40decision one more question
52:43can you talk a little bit about sort of
52:45in season demand planning and sort of
52:48who should own that where you've seen it
52:50owned successfully and how that sort of
52:53what the interactions are Thanks
52:57both Blue Nile and julep that demand
53:00plan is is with the merchandising team
53:03the there they're trying to drive the
53:05revenue with the product assortment and
53:07the promotions and and that makes sense
53:09to me and then we couple very tightly
53:12with that though on the operation side
53:15that and do our own translation into you
53:18know orders units calls and that sort of
53:21thing so you know they're closer to what
53:24we're going to sell and then we take it
53:28well and it's also being responsive I
53:31mean because the plan is just the plan
53:33and you know that's that's the that's a
53:39long and probably a never-ending battle
53:42between the people who do the demand
53:44planning and the operators right so what
53:47you describe is very common you you get
53:49a sales plan that's expressed in dollars
53:51and then somebody tries to convert that
53:54to orders and then they make extrapolate
53:57that to units and they may take a
53:59monthly sale plan and try to water it
54:01down to a weekly or a daily or sometimes
54:03by shift plan it's always wrong by
54:05definition it's always wrong so the
54:08biggest challenge from an operation
54:09standpoint is how do you build enough
54:13flexibility into your capacities to
54:16handle not just the plan but both sides
54:21of that plan you know how do you respond
54:23downward so you don't throw a lot of
54:25money away and how do you flex upwards
54:28so you don't blow customer service
54:29during those periods yeah that's got to
54:31be part of it ya know so what we found
54:33is that when we've had like for a while
54:35we have we asked the you know planning
54:38team on the merchandising side to
54:40generate the units and orders and what
54:44we found is they they have less sort of
54:46skin in the game and in terms of that
54:48accuracy and and so we we wanted to own
54:51that because we could make the better
54:53you know better estimates and also
54:54understand the the kind of margin of
54:57error better so that we could understand
54:59what variability to build in yeah and
55:01and depending on the business here and
55:03you really do have to build a church for
55:05Easter Sunday because you have to be
55:07able to handle the worst case scenarios
55:09or or blow sales that you'll never get
55:11back possibly lose customers that you'll
55:14never recover so I think from an
55:17operation standpoint staying ahead of
55:18that power curve was always a big
55:20challenge couple of things that's a
55:23demand planning at stitch fix reports
55:29but we do it a little bit differently
55:31like it feels right to me but it's
55:34because I went from church for Easter
55:36Sunday and really big peaky season like
55:39holiday season - its stitch fix we try
55:43to have quality be the foundation and
55:45the three constraints being quality of
55:47how many stylists can pick the five
55:49things after the algorithm runs and ops
55:51capacity and that drives what we
55:54actually do so essentially it's a little
55:57bit of a tail wagging the dog where we
55:59say by day by size and kind of a really
56:02cool algorithm this is how much we can
56:04do and still keep quality across all of
56:07those dimensions and the customers come
56:09in to that now I just left a meeting
56:10where you know marketing saying add more
56:14capacity we got more demand why can't
56:16you do more and we have really tough
56:18conversations about sort of demand
56:20planning as it relates to keeping
56:21quality because what's great about from
56:24an operator's perspective is I tell our
56:26two warehouses many weeks out this is
56:29how many fixes you're doing and they can
56:32operate that very well and that's when
56:35when you run it like this
56:36that's where I can see where we have
56:38process issues or cost issues it's
56:40harder from my perspective to see how to
56:43get a really good process when one day
56:46or shipping 6000 the next day you ship
56:4812,000 the next day ship 2,000 I can see
56:51it when I get the ship 6,000 three days
56:53in a row because then I can see where we
56:55need to improve so we manage that your
56:57models tends towards linear versus it
57:00has we don't have a seat yeah we that we
57:01are been if we benefit a lot by having
57:03an on pretty much a non seasonal
57:05business like shut the Shutterfly's not
57:08seasonal right yeah I'm used to peaky -
57:11and it's heart I think it's hard to
57:13manage peak and be held accountable to
57:16costs and quality because like Mike said
57:19you have to build the you either have to
57:22invest in technology or have a flexible
57:25work force love to thank our panelists
57:30for a very stimulating conversation
57:32thanks everybody thank you