00:00we need to try to get back to like
00:01fundamentals and how do we actually
00:03iterate and ship fast we have to build
00:06products that are constantly evolving
00:08and that are constantly learning and
00:09evolving from those learnings the only
00:11way really to stay ahead of that curve
00:13on that consumer side is get the
00:16Lightnings keep hitting I mean that
00:18sounds hard be real that's incredibly
00:20hard second product hitting so rare
00:22third product hitting you're a
00:24generational company it's like there's
00:25this big wave coming and it might hit
00:28your ship is there a way to avoid the
00:30wave I don't think so I think it's kind
00:32of a tsunami if you sit there it's just
00:33going to wash you away people will want
00:35to talk to people so like Community
00:37becomes much more important referral
00:39becomes much more important where the
00:40math becomes much more important the new
00:42limiting factor is possibly creativity
00:44that's a very very fun place to be
00:46despite it being easy ever to build a
00:51to grow one especially sustainably with
00:55many tried and true channels no longer
00:57working as they once did paid platforms
01:00were hit with app transparency tracking
01:01SEO is not threatened by AI chat social
01:05graphs have moved to interest graphs
01:06product-led growth is easier than ever
01:08to copy and virality is just as
01:11transient if not more transient than
01:14before and those are just a few of the
01:16examples that Founders are facing so
01:19what growth Green Space still exists
01:21especially in an era of austerity how do
01:24you balance efficiency and profitability
01:26while still growing and what channels
01:28matter and how do you master them in
01:302023 these are just a few of the
01:33questions we address in this episode
01:34together with three people that have
01:36successfully grown companies through
01:38similarly confusing waves of the past
01:40first off we have Gina got Health now
01:43leading latitude but perhaps most well
01:45known for her impressive reign as the VP
01:48of growth at Duolingo growing them from
01:503 to 300 million users with no paid
01:54budget we also have Kieran Flanagan join
01:56the conversation long time SVP of
01:59marketing at HubSpot where he was a
02:00founding member of their international
02:02business all the way up to HubSpot
02:04hitting over 1 billion dollars in ARR
02:07and finally we have our very own Brian
02:09Kim partner at a16c who also held
02:13several leadership roles at snap from
02:15their hyper growth days all the way
02:17through IPO alright I won't keep you
02:19waiting any longer let's dive in as a
02:23reminder the content here is for
02:25informational purposes only should not
02:27be taken as legal business tax or
02:29investment advice or be used to evaluate
02:31any investment or security and is not
02:33directed at any investors or potential
02:35investors in any a16z fund please note
02:38that a16z and its Affiliates may also
02:40maintain investments in the companies
02:42discussed in this podcast
02:44for more details including a link to our
02:46investments please see a16c.com
02:59what are you seeing in terms of growth
03:01that still works in 2023 are there white
03:04spaces that companies can can tackle
03:06what's happening is kind of really
03:07interesting because there may be like AI
03:10is somewhat of a Tailwind for like maybe
03:1312 months or 18 months for the best
03:15people and then who knows after that
03:17maybe a headwind across everything so I
03:19think certainly in search I've really
03:21dug in there and there are lots of
03:24examples of the click-through rate
03:26decreasing and the amount of traffic you
03:27get from Google search Pages Pages
03:29decreasing but I still think there's
03:30like a lot of value in that paid ads
03:32very similar I still think there's a lot
03:33of value in that what really happened
03:35with with the popularity and growth of
03:37the internet is like marketing growth
03:38became much more measurable and they
03:40could have a seat of the revenue table
03:41and they could say yeah I put a dollar
03:42in here and I get three dollars out here
03:44and the growth that's starting to happen
03:46today or where things are starting to
03:48shift is a little bit back to like this
03:50indirect causation these channels like
03:52if you look at the platforms that that's
03:54growing it's like short short from video
03:55Tick Tock YouTube like these kind of
03:57media channels if you look at where I
03:59think things will go with the evolution
04:01of AI people will want to talk to people
04:03so like Community becomes much more
04:05important referral becomes much more
04:07important where the math becomes much
04:08more important all of these things are
04:11are going to shift us back into like the
04:13place where we have to be okay doing
04:16things where you can't measure
04:17everything directly all the time and I
04:20actually don't think that's a bad thing
04:21because like because I think the over
04:22index of measurements forces everyone
04:24into the middle and everything becomes
04:26like mundane and lacks creativity and so
04:30I think that there's like some pros and
04:33to actually this happening but there's
04:35certainly a change in terms of how you
04:38are able to do your work and actually
04:40relay the value of that work to like the
04:43founder who the founder is for the most
04:44part still going to be stuck in the
04:45binary give me three dollars for the
04:48dollar that I give you we're almost like
04:50going back to the origins of advertising
04:52isn't that like there's that saying that
04:54says you need to hit someone seven times
04:55before they convert and Facebook maybe
04:57changed that maybe it was like two times
04:59and then it feels like we're sliding
05:01back in the other direction I think
05:04it has always been the case that people
05:05need repetition to really notice
05:07something and yes more and more there's
05:10more media there's more for us to watch
05:12and listen to and and we're more
05:14distracted so I think that that becomes
05:16even more the case but I I I may be uh
05:19you know simplistic but I don't think
05:21that much is changing because to me
05:23thinking about growth is about the
05:25framework for understanding what makes
05:28people tick what people care about how
05:30to reach one person or one media Outlet
05:33or one channel whatever it is that will
05:36reach a lot more people in the most
05:37efficient way possible what are the the
05:39moments in your product that people
05:41really really care about meaning like
05:43number of people who actually encounter
05:45that moment and if you actually make a
05:47modification like make some sort of like
05:50um you'll actually see significance as a
05:52school results on the other end because
05:53you know people really care about that
05:55one moment it could be a notification it
05:57could be a screen you know we we keep
05:58seeing these Trends and and things grow
06:01and ebb uh SEO rules are constantly
06:03changing because we're at the whim of
06:05algorithms and so we're just constantly
06:07trying to adapt places become saturated
06:09like at one point it was really trendy
06:11and cool to do ads inside of the subway
06:15in New York City like that was on like
06:17someone discovered that in 2000 112
06:18which is funny because it feels pretty
06:20obvious then suddenly everyone was doing
06:21it everyone's testing with notifications
06:23and everyone's saying notifications are
06:25just there's too many and it doesn't
06:26matter I think all of those channels are
06:28still very relevant it's about finding
06:30how to actually impact how people feel
06:32and finding you know new ways of
06:35impacting people that are cheaper
06:36because less people are using it over
06:38and over so I think that that will
06:40continue to be the case and you know it
06:43just depends on on what's the trendy
06:45Channel and what are people paying
06:46attention to but if you think a priori
06:48like that like how do I reach one that
06:50can then reach a lot or what are the
06:52things that I can modify that actually
06:53make a difference then I don't think
06:55that much has changed I I have like a
06:57slightly heretical take so this is going
06:59to be fun but um you know when we talk
07:01about gross especially I I know a16d
07:04we're all we're a Silicon Valley and the
07:05stalwart what have you but when I think
07:07about gross it's like gross tactic gross
07:10uh you know channels and uh you know
07:12even the word gross hack uh became very
07:14popular in that you know there's this
07:16prevalent Theory and and sort of Kieran
07:19touched on it where you can put in a
07:21dollar and get back why and like I put
07:23an X dollar get get why back and I think
07:25that certainly became true and you know
07:28there became so many channels and there
07:31became so many wisdoms and tried and
07:33true tactics whether it's SEO or
07:35ambassador program or community building
07:37where everyone's trying to learn from
07:39each other I'm like how do I hack this
07:40thing Gina said something interesting
07:42where this is sort of how I think about
07:44it where like moments in the product
07:45when people go aha like this is great at
07:48the end of the day I think gross we're
07:49sort of conflating two things one is the
07:52actual product is is a sticky does it
07:54does it enable you to continue to come
07:56back because growth at the end of the
07:57day is like how many people continue to
07:59use your product it's like not a
08:01singular metric of like on day x how
08:04many people showed up well that's useful
08:06for today but actually for a company
08:07that is ongoing concern what you want to
08:09know is like how many people
08:10consistently show up and how does that
08:12grow right it's a cumulative concept
08:14which means that you know the fact that
08:16people come back day in day out is
08:18probably one of the most important
08:20factor in in assessing the gross
08:22potential and growth of it which to me
08:25goes back to a little bit of product
08:26where when people first come on like we
08:29all notice but you know the biggest drop
08:31off typically day one or day Zero when
08:33you know people come on they do
08:35onboarding and maybe they didn't get to
08:37the aha moment quickly
08:38so why would they come back if they
08:41don't come back then your game as it
08:42grows is like how do I get more
08:44acquisition right because gross to me is
08:46also sort of a you know numerically
08:48almost an equation there's like you know
08:50people who showed up yesterday plus new
08:52people that you acquired minus the term
08:53people plus the resurrected people
08:55that's sort of like an equation
08:57and when you think about you know how to
08:59get the retention up I think it all goes
09:01back to product why as Founders and
09:06why is there growth in your product like
09:08why is it useful for more people and it
09:11is is your product you know constructed
09:13in a way as Gina said like gets you you
09:16know helps people understand what your
09:18product is about when they first use it
09:20can you deliver that product value again
09:22and again consistently I think now we
09:24call it product like gross but it's
09:25honestly like a core product value on
09:27how you actually Define that is one and
09:30a lot of things that we're also talking
09:31about here is like distribution so
09:33there's like a product in the
09:34distribution side of it and if we think
09:36of like distribution I think you know
09:37the panel is exactly right they're you
09:40know they're uh you know areas where
09:42it's less good now and they're areas
09:44that are continuing to be good you know
09:46you talked about advertising there's
09:48that that saying that you know you never
09:50know where half of it goes and you know
09:53maybe we're going back to that but you
09:55know I think there are many other areas
09:57that continue to do work work well
09:59whether a Subway ad or Remnant TV as a
10:02radio it depends on your product is and
10:04when your intended audiences if you're
10:07you know marketing or or appealing to
10:10Young Generation of you know Gamers or
10:12communities or Niche areas and
10:14Reddit and Discord are actually pretty
10:16good places I also think you know
10:19ATT did kneecap a lot of the sort of
10:22advertising paid advertising uh side of
10:24things but you know the AI like Kieran
10:27also talked about can also be used for
10:29that purpose right you know Facebook or
10:31meta is investing heavily on you know AI
10:35side of it so that all sometimes all you
10:37need to do is just say hey like I have
10:39this idea here's like the bass creative
10:42and now they're able to like change what
10:44how it looks and add different you know
10:46copy like change the thing to see what
10:48works automatically at scale with AI and
10:52deliver higher value and that's also you
10:55know very promising I think direction
10:56that we're seeing where the efficacy of
10:59the paid side it may also be restrunged
11:01yeah I mean I like the breakdown between
11:03there is product and there is
11:04distribution those ladder up to growth
11:07but maybe if we just attack this idea of
11:10building a good product that's not a new
11:12idea but it does feel like at least to
11:15Me Maybe we are in a different era where
11:17it's easier to copy products and so is
11:20it just about being a first mover in in
11:23an era where things are easier than ever
11:25to build and maybe Gina you can speak to
11:26that because Duolingo kind of famously
11:29grew with no marketing budget is that
11:31correct or very little marketing budget
11:33and it was it was exactly product-led
11:36growth that resulted in in that company
11:38becoming so big yes so what I think has
11:41really changed with tech is that people
11:44could say oh my product was really good
11:46before but like it wasn't as easy to
11:48know whether a product is good or not
11:50good as it is now because you can
11:51actually see that in metrics in terms of
11:53as Ryan was saying just in terms of
11:55retention and people coming back again
11:56so now that defines what a good product
11:58is but I it's not just about building a
12:01product that's good but also the fact
12:03that with tech we have to build products
12:07that are constantly evolving and that
12:08are constantly learning and evolving
12:10from those learnings but the main thing
12:12I think that like when you talk about
12:14um what we're trying to do at Latitude
12:16now which is my new company the
12:18importance of not just building a great
12:19product but understanding really clearly
12:20what your mode is and your mode needs to
12:23be something that makes it really hard
12:24to replicate your product
12:26um and it can be anything right like so
12:28for example if Kim Kardashian launches a
12:30shoe that's exactly just like every
12:32other shoe that exists out there the
12:34mote is Kim Kardashian because her brand
12:36and her audience and like the fact that
12:38she is an influencer she's now telling
12:40people to buy the shoe that's a most it
12:42means that like it's the thing that
12:43makes it really hard for you to go and
12:45just do that again with Duolingo the
12:47product is the moat and that's the most
12:50incredible thing because of what I just
12:52said before about the iteration and the
12:55anyone at any particular moment in time
12:57and many did try could actually go and
12:59copy exactly what Duolingo was doing
13:01they could they in fact use the same
13:03font in many cases they use the same
13:04design colors like the same form and you
13:06can do that but you don't have the data
13:09and the team behind it that knows how to
13:11use that data that is constantly
13:13learning from every single day of you
13:15know all of the usage that we're getting
13:17and understanding what's leading to
13:19every single decision and why we're
13:20modifying with things in a particular
13:22way that's impossible to catch up to so
13:25at every moment in time even if you
13:27mimic exactly what Duolingo is there's
13:29no way you're going to get to the next
13:31day of Duolingo because you don't have
13:33all of that history or the knowledge of
13:34how to use that history so you could
13:36just spend all your time trying to like
13:37copy Duolingo every single day but that
13:39doesn't actually get you to solve a
13:41problem for a consumer and understanding
13:43how what you're doing is actually
13:44solving a real problem which is at the
13:47core of as you were saying a retentive
13:49products that people actually like and
13:51that defines a good product
13:53um so it's not as easy as it sounds
13:55it's definitely not as easy as it sounds
13:57Karen I want you to tap in here you now
14:01are the CMO at zapier which I feel like
14:03actually is is a great parallel to what
14:05Gina just said I mean zapier has so many
14:07you could say clones or competitors
14:09right like I've used many of them many
14:12of them are cheaper how do you think
14:13about that differentiation and building
14:15a product and to Gina's Point like
14:17getting the right data and involving it
14:19in a way that you can stay ahead you can
14:21have emote yeah I think first remover
14:23Advantage is useful if you can build a
14:27healthy retention curve and that
14:28retention curve helps you make that
14:29product much better over time and so
14:31you're just like making it much harder
14:32to compete with you and that can be in
14:33all manner of things like zapier's mode
14:35for some time was just they had much
14:37more expansive coverage of apps and the
14:40amount of things you can automate with
14:41and then like more recently it's like
14:43how do we integrate AI how do we make
14:45automation much much easier for
14:46companies how can they actually do
14:48natural language within their apps like
14:50one of the things that's super
14:51interesting about the the the AI Parts
14:53is coming back like the distribution and
14:55the product part and the product life
14:56growth and like AI is a phenomenal
14:58unlock for product like growth because
14:59product like growth will get naturally
15:01built into every single product because
15:02every single product will have a natural
15:04language interface and so like you will
15:06actually be able to onboard the products
15:08and use products without actually any
15:09human intervention intervention and so
15:12and I think that's an incredible like
15:13augmentation of like how how apps work
15:16in general but I think the the mode part
15:21going to be much more complex like I'm
15:24talking from through the lands of B2B
15:25not through consumer apps but certainly
15:27for B2B it's going to be somewhat more
15:28complicated to figure out over time
15:30because AI does commoditize things and
15:32make things much easier to replicate and
15:34we've seen like AI maybe one of the few
15:36platform shifts that actually skews
15:37towards the monopolies and conglomerates
15:40versus like these individual companies
15:42who just get sucked up and there are
15:43features of these existing platforms
15:45that get kind of skews towards if you
15:46have distribution you might win because
15:48I think AI features might be more of a
15:50retention play than an actually new
15:52acquisition play which is like if you
15:53build the AI features into a platform no
15:56one will leave to you to actually use
15:58all these different like Point Solutions
16:00uh so I think the moat thing is I think
16:03modes will exist in similar formats
16:05there may be just companies with less of
16:07them and I do think distribution as a
16:08mode retention is a mode Network effects
16:10will still be emote Community is a mode
16:11so I think I think of all of those
16:12things in relation to zapier like how
16:14how are we touching on all of those
16:15things as we can continue the Growth
16:17Company one thing I want to respond to
16:18what Kieran said um you know at Latitude
16:20I would say the community is our mode
16:22but the reason why the community is
16:24remote in addition to the distribution
16:26aspect of it is the trust building so
16:28trust can be a mult because that takes a
16:31while to build so for B to be product I
16:33think that's also super relevant because
16:34if someone can come into the market and
16:37just like offer the same product but
16:38you've built a brand on a zapier side
16:41that is super trustworthy and you know
16:44and so people will will continue to to
16:46trust new features or new solutions that
16:49you put out there instead of trying to
16:51you know maybe try new products I think
16:53that that's another aspect worth
16:55considering by the way Karen how many
16:57times a day do you get asked if it's
16:58zapier or zapier I don't even know if I
17:01say it correctly I said it's out here
17:02but it may be safe here
17:04I heard one of the founders say it's
17:07it's zapier it rhymes with happier
17:09because it makes oh there you go okay
17:11great I have something about it either
17:15that's how I thought about it yeah it's
17:17zaps yeah yeah the main reason I join is
17:19just like saying the word zaps like it's
17:21just a cool world I know we talked a
17:23little bit about B2B just talking about
17:26more of the consumer side
17:28gross and sort of thinking about
17:30distribution it it has changed I think
17:34you know a long time ago already Mark
17:37Zuckerberg at meta already said you know
17:39it's it's a battle for the you know home
17:42and more and more even recently people
17:45are downloading less and less apps over
17:48part one part two is
17:50is harder and harder to get discovered
17:53in the App Store you know App Store
17:55ranking is uh is hard to sort of uh get
17:57Discovery now so now you have this issue
18:00of one people are less willing to
18:02download brand new things and two even
18:04if they sort of kind of heard about you
18:06it's like harder to get discovered so I
18:08think part one of the consumer side the
18:13uh for lack of better word maybe I can
18:15call it ossified where there are two
18:18distribution platform Google Play and
18:19Apple Store great and they just
18:22distribute apps and people re-download
18:24ton of apps they retry a lot of things
18:26with brand new things
18:29or certainly less than before so now you
18:32have this conundrum of like okay like
18:34it's just harder to break through the
18:36noises and we talked a little bit about
18:38especially Kieran talked about you know
18:40maybe AI is a is an accelerant for
18:43retention maybe less of an acquisition I
18:45I think that's that's correct and that's
18:47interesting where also we have a lot of
18:49these like incumbents right like
18:51especially in the consumer side
18:53incumbents come with Incredible modes
18:56they already have the network effect
18:58they already are synced to everything
18:59they already have login for everything
19:01they already have all your data they
19:03know what you bought they they they know
19:07you talked about how to do how do you
19:11it's very very very difficult actually
19:15even if you have a clever hack all these
19:18companies come up with Incredible new
19:22and they enter the zeitgeist
19:24and through entering the Zeitgeist
19:28solve the you know first order question
19:30how do I break through the noise and get
19:32distributed initially and have more
19:34people to try it out
19:35that alone is a feat that's incredible
19:38if you have highly retentive product on
19:40top of that legendary that's amazing
19:43fantastic and all these all these
19:44products that we talked about did have
19:46that now the question is okay copying
19:49Gina you talked about people copying
19:51down the fonts I I understand that so
19:54deeply of that Snapchat for four or four
19:56four or five years and
19:57yeah you know down to the font and every
20:01single like UI level yes they they copy
20:03everything and um how do you actually
20:06fight that as a new entrant
20:09especially when incumbents come uh
20:12you're burying a lot of these already
20:14like weaponaries if you will that have
20:17retention and data and everything it's
20:19hard and I and I think actually and this
20:22is like a sort of a difficult answer the
20:25only way really to stay ahead of that
20:27curve on that consumer side is
20:29get the Lightnings keep hitting which is
20:33so so so hard I think getting you know
20:36the first hit is so hard second product
20:40so rare third product hitting you're a
20:43generational company so you know it's
20:46sort of that Innovation curve that I
20:48think is like very interesting so for
20:50the companies that are newly entering
20:52and bringing building new product and
20:54we're on the side of the builders right
20:55incumbents may have like all the
20:57advantage but we want to see new
20:59products we want to see new you know uh
21:02hungry founder that are building for the
21:03next generation and they're not happy
21:05with the current set of solutions so the
21:08only way potentially is you know their
21:10series of things but like step one boss
21:12game is like how do you get through the
21:13distribution noise step two boss game is
21:15you know how do you get retentive
21:17product right step three and this is
21:19like a Holy Grail in my mind is like how
21:21do you keep hitting that
21:23lightning like get get that lightning to
21:25hit continue to enter Zeitgeist and for
21:27that I you know I think it's
21:29thoughtfulness I think it's a structure
21:31of the company I think the culture of
21:33the founders and the team of wanting to
21:35continue to put out the best and
21:37brightest work that they have
21:38continuously versus potentially resting
21:41on laurels or optimizing for things that
21:43have worked in the past and and thinking
21:45that their first iteration product is is
21:47what is going to bring them to the next
21:49next level because they may not be
21:51because the distribution advantage of of
21:54the incumbents as well as the difficulty
21:56of getting further discovery
21:58what's an example of that that sounds
22:00really hard is there an example of a
22:02company that you've seen execute on that
22:04hard mode really well today
22:06I mean you know one of the last examples
22:09uh I've seen in you know one of our
22:11portfolio companies be real has done
22:13that too with with certain features but
22:16I'll talk about snaps is that's the name
22:18I know well and you know it's been
22:20copied so many times what have you you
22:22know the first hit one may uh think is
22:24the disappearing images that was
22:26revolutionary that people thought of
22:27images as something to keep on your
22:29phone it's like for recording and for
22:31keep keepsakes you turn that into
22:34disappearing thing which made the medium
22:35itself a message communication method
22:38that's revolutionary very interesting
22:40you know sexting was the first use case
22:42but it like catapulted the company into
22:45great so that was going really well
22:48now then the kicker is they then started
22:53right stories of you know these
22:55disappearing photos like linked together
22:57that shows you a day and time in your
23:00life in a passive way but also goes away
23:03revolutionary in that in that sense then
23:06that to me is a second Zeitgeist hit
23:08that enabled by the way what a great
23:10example of something being copied to
23:12right everyone has so many different
23:14everybody has it like like LinkedIn has
23:17stories like no no dig on LinkedIn But
23:19like everyone has stories now so you
23:22know it's a medium and like a the thing
23:24that worked really really well and
23:25potentially even birth like the short
23:27video uh type thing uh the trend so to
23:30think about the second Innovation that's
23:32like a huge innovation
23:34third also is you know they did a rest
23:37area it was like okay we we are seeing
23:39more and more people abandon the front
23:41result of the front facing camera what
23:43is the problem well they don't like what
23:44they look like well let's add dog ears
23:48on people's face let's let's uh like
23:50rainbow pukes on people's face and
23:52honestly and again that's another thing
23:55uh Steph where everyone copied like
23:57everyone has lens now like it's it's
23:58very very beautiful dog ears now Tick
24:00Tock uh you know puking rainbows blue
24:03eyes what have you but these augmented
24:06reality effects are are now prevalent in
24:08because you're a product so this is one
24:10example where I think each subsequent
24:13hits have essentially catapulted the
24:16company the new height and that I know
24:21very difficult very long term with
24:24several bets going at the same time
24:27uh and become being very thoughtful of
24:29what to kill and what to put out there
24:31so it's that system of innovation it's
24:34that hunger and continues continuous
24:37past that I think is is very important
24:39that's feed and hunger and being able to
24:41like Implement all that at a fairly
24:43large and established company I think
24:45it's a very difficult thing so I think
24:47there's a cultural aspect I think
24:48they're the founders mentality of always
24:51wanting to be ahead of the curve always
24:53implementing and sort of uh looping into
24:56new products that the users are looking
24:58for how do you Foster that culture I
25:01think especially in 2023 where maybe
25:03budgets aren't as big maybe some of the
25:06talent you had before is not used to
25:08that austerity or that that idea of
25:11having to ship really quickly maybe
25:13spend less money and just figure it out
25:15I've talked to a lot of companies this
25:16year that were CD public
25:20I feel like a lot of Founders are
25:23actually excited to get back to shipping
25:25like I like I think we went through this
25:27ballooning of companies during covid and
25:31everyone's like wow like I don't really
25:33know how this company even works anymore
25:35yes and we need to try to get back to
25:38like fundamentals uh and how do we
25:40actually iterate and ship fast but I
25:42think what really is exciting in Tech is
25:45like getting back to the reason that
25:47people were excited to get into Tech
25:48like certainly I was as a failed
25:50software engineer is like to do work
25:52like like not to like figure out how to
25:55like coordinate hundreds of people but I
25:56actually like to get back and do work
25:58and ship things and I think the faster
25:59you ship the more you learn so I think
26:01it comes from the founder like I guess
26:02it's a long-winded way to say like I
26:04think you'll have some Founders that
26:05Embrace this and actually at the heart
26:07they actually thrive in the kind of
26:09chaos they actually enjoy this period of
26:12time especially like a lot of Founders
26:13who are building with AI I've met Lots
26:15who are like this is the best time to be
26:17a Founder because we have no idea how
26:19this is going to go and they love that
26:20they don't they love like not knowing
26:22what the end result is rather than just
26:23tweaking the dials and trying to eke out
26:25little bits of growth each and every day
26:26and that trickles down from them right
26:28they they set the tempo and it's up to
26:30the rest of us to up level to meet that
26:32bar this comes from two places one as as
26:34Karen said from the founder so from the
26:36top and I think that what that the
26:37founder can do is come up with
26:39artificial deadlines for example we have
26:41to launch in September because we're
26:42doing an event for 4 000 people and
26:44that's when we need to launch and that's
26:45it or like a Duolingo would be like oh
26:47you know we told Apple that we're
26:49actually going to make this Sprint and
26:51now that's that's there so we just have
26:53to do it and so suddenly it's not like
26:54something that like can be pushed
26:56because there's this external deadline a
26:58lot of times it was artificial because
26:59you kind of created but now it's set in
27:01stone so I think you can help from that
27:02perspective and the other one I think is
27:04actually creating a culture from the
27:06very practical point of view of like a
27:08process like for example Sprints you
27:09know so you have like two weeks prints
27:11and that's sort of the Cadence at which
27:12things happen and maybe that's the
27:13Cadence at which you're doing a b
27:15testing making sure that you are
27:17prioritizing as part of that Sprint and
27:20that you have actually have that you're
27:21not just shipping to ship because then
27:23you can just like spray and pray but you
27:25need to like really be learning from
27:26everything that you're shipping and
27:27iterating so that has to be part of the
27:29process whether it be retrospectives or
27:32whether it be having accountability like
27:34our product team at Latitude has a show
27:37and tell every two weeks and where and
27:38sort of the whole company is there
27:39expecting to see what's evolved at
27:42Duolingo we had a practice of making
27:44sure that every experiment had a proper
27:45write-up in the beginning with like all
27:47of the hypotheses and what we actually
27:48expected and then like two weeks later
27:50to be able to come back and then look at
27:52all of those because otherwise we'd be
27:54running around thinking about the next
27:55test and we'd forget to go analyze think
27:57about what we learned and apply so I
27:59think that all of those you can sort of
28:01have little levers that creates the
28:03culture from a practical standpoint so
28:05that you can set and forget and not have
28:06to as a Founder keep watching and seeing
28:09if people are are moving fast and you
28:12know and iterating so funny you
28:13mentioned the artificial deadlines
28:15because I was uh I was thinking about
28:17this recently about how you know when
28:18you go through college you have exam
28:20sometimes you have projects that have
28:22due dates and then when you enter the
28:24workforce especially over the last few
28:25years you have those but they feel very
28:28very gray and so I think it's a just a
28:30funny acknowledgment about how sometimes
28:32that works in the workforce but um you
28:34know you spoke to experimentation
28:36how do you how do you think about that
28:38in you know in 2023 should companies
28:41especially when things are tighter
28:42should they be leaning more into
28:44experimentation because it's like okay
28:46we got to figure this out like Karen
28:47said we don't know where ai's going or
28:49should they be really drilling down on
28:51you know maybe the one or two channels
28:53that they know work I always think of
28:55like your growth initiative as a
28:58portfolio of bets and your portfolio is
29:00like very similar to like how you think
29:01about investing you want to have a
29:02spread across things that you have a
29:04high degree of confidence in and the
29:05thing that you have a high degree of
29:06confidence in is where you have
29:08historical context to say we've done
29:09something similar before and it will
29:10likely work and I can measure that
29:12against like what's a potential upside
29:13but you definitely want to take some
29:15things that are like more my speculative
29:17guests now all my speculative guests in
29:19terms of investing from 2020 to 2022 are
29:22all down by 80 so they definitely do
29:24lose I have a lot of I've lost a lot of
29:25money but in terms of your growth
29:27portfolio you want to take those big
29:29swings and I think one of the things
29:30that happened to your point staff like
29:32that phenomenon of like this is kind of
29:35a deadline but it's not really a
29:36deadline that definitely is like that
29:38had that definitely like I've been
29:40attacked for a while that definitely was
29:42not always the case and became the case
29:44when Tech was ballooning and I think
29:46it's good to get back to actually like
29:47this stuff truly matters and hidden
29:49these days truly matter but I think the
29:51other thing that we've got away from is
29:52like having people in the company who
29:54will take risks who have deep conviction
29:56about things who will take risks who
29:58will like say hey we think this is the
30:00right thing to do give me a couple of
30:01people and I'll either prove this at
30:02weekends or nights or evenings or and
30:04I'll figure out this is like worth doing
30:06or not and I don't have any historical
30:07context to say this is a good idea but
30:09like I've talked to customers and I
30:10think this is the thing that's going to
30:11work and so I do think that we need to
30:13have experimentation across the entire
30:15Spectrum like from like high degree of
30:16confidence high impact to like high like
30:18really Game Changer thing not a lot of
30:21content uh strong context to say this
30:23works but like it's a real thing worth
30:24doing and if you are at the point where
30:26you can have statistical significance
30:27because you have a lot of users then I
30:30think you know you have to everything
30:31and it needs to be an experiment even
30:33the things that you already know like
30:34the channels that are already working
30:35because if it's not an experiment then
30:37you are not looking at it from a very
30:39scientific perspective and gaining data
30:41that will actually tell you whether you
30:43can replicate this when when you launch
30:44it to a lot of users so I don't think
30:46that but I agree 100 with with Kieran in
30:49the portfolio approach it's not either
30:53um but I I do believe in being as data
30:55driven as possible of course there are
30:57exceptions and there's an art to things
30:59but to be as data as driven as possible
31:02and that's where you'll find
31:03experimentation to be super important
31:05and if you are an early stage startup
31:07you need to take big swings because if
31:09you get in incremental improvements and
31:11things you're just not going to get
31:12anywhere you can like improve things two
31:14percent over and over and no one will
31:16notice you and the way that I would
31:17approach it from a very tactical
31:19perspective is I would list all of the
31:21ideas of things that I would like to
31:22work on whether it be like tried and
31:24true channels that I want to iterate on
31:25or big swings and I would just try to
31:29rank them based on Roi like how hard is
31:31it going to be to do this what do I
31:33think is the return gonna be and then
31:35just try and that's how I would use like
31:37that that's the information I would use
31:38to make a decision about the portfolio
31:40of things we'd be working on as a team
31:42yeah just to give one example on the big
31:43bets just uh because I think sometimes
31:47we think of these things as like the
31:48kind of a B test like for HubSpot when
31:51we acquired like acquiring the hustle
31:52for us was a BET right we had deep
31:54conviction about the the future of B2B
31:56Mark or like content needing to be much
31:59more engaged and need to be part of the
32:00everyday presence of how people you know
32:03interact with content wanting to engage
32:04with content but we had never bought a
32:06media company before we had never
32:08actually had yeah we had like never had
32:10like media talent that we had to like
32:13figure ahead into it to the company
32:15um and we use that talent to like build
32:16a podcast Network you built the Creator
32:18program for our staff we built all of
32:20that all of this stuff are a really
32:21incredible presence on YouTube and we
32:24got incredible talent that I think is a
32:26real mote against most B2B companies
32:27because they do not have this Talent
32:28within that company that's like an
32:30example of like a sizable bet mapped to
32:34something that we had deep conviction
32:36um so I think that I think those kind of
32:38things do matter it was like a fun
32:40Matrix of like important and Urgent and
32:42important non-urgent like the two by two
32:44type thing and you know some of the bets
32:46are you know both in important and
32:47non-urgent and important and urgent but
32:50then you have these other you know ideas
32:52where you have this inkling especially
32:54in the consumer side it's nearly
32:56impossible to predict how something will
32:58do especially if it's like a fairly
33:00large bet so there's a little bit of an
33:02intuition where you know you sort of
33:04rank a couple of options and like look
33:06like I think this is a big product bet
33:07and that that could be one of the
33:08experiments or or sort of projects that
33:11that you work on Karen you said
33:13something earlier where uh I really
33:17at the end of the day where we are the
33:19founders and operators do what they do
33:21because they want to ship ship things
33:22and you know ship things that people use
33:24and uh back to the fundamentals I keep
33:27thinking it's one line or like the comsy
33:30never made good Sailors and so you know
33:33we're not at a like a common scene and
33:35then you know what what are good Sailors
33:37made of and they're they're probably
33:38made of like a lot of experience but
33:41also knowing how to do the basics really
33:42well really really well and so going
33:45back to that fundamentals to me is
33:47really interesting because there are two
33:50different ways to deal with this which
33:51is like how do I get out of this and
33:52what do I do and there's all these like
33:54tactics and hacks that you might be able
33:56to do but there's another point of like
33:58no actually the the way we become like
34:01an important generational enduring
34:04get back to the foundation and
34:06fundamental then build stuff that people
34:09want to use and genuinely as helpful to
34:11them and I and I think that constitutes
34:13some of the ways as how people can think
34:15about making big bets totally
34:18um maybe one topic we could just address
34:20directly is AI because on one hand it
34:23you know you're all kind of speaking to
34:25this idea of the fundamentals right
34:28focusing on what really matters but then
34:30you have this big wave like you know to
34:32use your sailor analogy Brian it's like
34:34there's this big wave coming and it
34:37might hit your ship and you got to
34:39figure out how to how to integrate it
34:42how are you seeing companies really kind
34:45of intelligently creatively on more so
34:47the distribution side or thinking about
34:49how companies can grow but you know look
34:51this is very new and I I I'm very new to
34:53understanding AI so I you know I'll just
34:56put that there but I think that from a
34:58distribution perspective there's two
35:00ways to approach AI one what are the
35:03ways in which we can use AI to reduce
35:05the work that we're putting into things
35:06so like when we're talking about Roi
35:08like that's the investment it's just
35:09like reduce amount of time spent on
35:11things like for example copywriting or
35:13image generation I am a purist when it
35:15comes to copywriting so like I think
35:17that everything needs to be like a you
35:19know Pulitzer Prize winning copy so I
35:21don't actually like copywritten by Chad
35:23GPT but it's actually like 75 to 80 of
35:26the way there and so I think that you
35:28can probably use that you know to
35:30benefit your company in many ways either
35:32by you know helping organize ideas or
35:34coming up with a V1 of copy and I think
35:36you may be able to do that with image
35:38and potentially video as well so I would
35:40I would think about how to Leverage
35:42AI to reduce work to automate things and
35:45just make things cheaper from the time
35:47end and spend perspective Gina by the
35:50way you gotta incorporate that into your
35:51prompt you gotta say YouTube YouTube
35:54title written by Pulitzer Prize winning
35:57writer yes I'm sure I think you know
36:04motion and the ability that humans have
36:06to understand each other and how to make
36:08each other feel is really going to be
36:10really hard for AI to pick up on and so
36:12we are able to come up with just that
36:13one word or just that one punctuation
36:15that really like drives a message and
36:18that's going to be hard to replicate
36:19unless it becomes a little bit of a
36:21recipe at which point it stops impacting
36:23people because people become a little
36:25bit numb to that but then that drives me
36:27to the second point in which I think we
36:28can use AI interestingly which is people
36:30are very impressed by AI right now which
36:32means they're more likely to pay
36:33attention which means they're likely to
36:35think it's funny which means they're
36:36likely to think it's moving or shocking
36:39and all those are all things that help
36:41your brand become remembered and
36:43ultimately that's what marketing and
36:44brand work is and so I would leverage AI
36:47in as much as I can to accomplish that
36:49what I think of AI for distribution that
36:51kind of split it into two categories and
36:52Gina covered one of them which is
36:54there's a lot of people that are just
36:55using AI to replicate the things they do
36:57and so that really is just like a cost
36:58saving you know marketing I think might
37:00move from like more of a channel domain
37:03expertise to more a journalist like we
37:04have have generalists and early startups
37:06I think that will actually be true of
37:08late stage companies because these
37:09generalists will be able to stitch
37:10together like these kind of micro apps
37:12to create customer experiences by just
37:14using data sets and using AI modeling
37:16and that will suit like operators and
37:18technologists that's great news by the
37:20way because like so many of us are
37:22generous and I'm like oh great I'll have
37:24a job yeah yeah I think I think if
37:26people are like enjoy data enjoy
37:28learning how to use apps and are able to
37:29like think like product Builders because
37:30I think marketers and growth people need
37:32to think like product Builders like how
37:33can I use this to stitch together some
37:35sort of build an experience for someone
37:37and I actually how do I creatively do
37:39that with these apps and data the other
37:41the so I think there's like the
37:43replicating what we already do and you
37:45see a lot of this with like outbound
37:46sales right like that's the number one
37:47use case you see right now hey I can
37:48drag in all this content and use chat
37:50TBT to create very very cool like iPad
37:52and sales the stuff that I'm interested
37:54in is like maybe maybe that's 90 the 10
37:56that is like what are things you can do
37:58with AI that is not possible today that
38:00Hudson distribution I'll give you like
38:01one quick example so like if you if you
38:03come into zapier because it's like six
38:05thousand apps and there's I think if you
38:07actually all the different like triggers
38:08and things like millions of different
38:09variations of what you could actually
38:11automate so how would I ever know to
38:14figure out like what I can onboard you
38:16to or like what you need to do when you
38:17actually sign ups user and the best way
38:19I could do that is like build an
38:21automation guide just for you right and
38:23today the way I would do do that is I
38:25would get someone to reach out talk to
38:26you about your Tech Sac figure out what
38:28role you are and then contextualize
38:29workflows and the language you
38:31understand that are like contextual to
38:33your job right hey Steph you're a
38:35marketer you want to generate more leads
38:36you want to send leads to the sales team
38:37you want to enrich them you want to make
38:39sure they can close and the customers
38:40more like you understand that language
38:41and you understand why that's valuable
38:42again today human would need to reach
38:44out write the automation guide for you
38:46so not economical not scalable we like
38:49one thing I've been testing is like
38:50we've been testing like chat gbt being
38:51able to do this like grab data from like
38:53something like clearbet or something
38:54like that text stack be able to feed in
38:56your role be able to feed in contextual
38:58information about your Tech stack
39:00company and then have that write an
39:02automation guide for you and then
39:03actually be able to service that
39:04automation guide whether through sales
39:05whether through chat whether through all
39:07these different mediums and it's really
39:09good like surprisingly good like the
39:11first time I did I was like boy like
39:12this is like wow like this is something
39:15we cannot do at scale today that's the
39:1710 that I think if you do as a fast
39:19mover you get this like little window of
39:22time before everything gets commoditized
39:24and everyone's the same stuff that again
39:25if you have good retention you actually
39:27get meaningful growth there because you
39:29have that retention curve you can get
39:30those users to stick upsell across sell
39:32and actually that's really really
39:33valuable so there's the 90 that I think
39:35most people are trying to replicate the
39:37things that we already do but that 10
39:39like is super interesting which is what
39:41are the things we can do at scale today
39:42that we could not do previous to
39:44actually have in this technology I'm
39:48I heard this quote once uh which is that
39:50the best growth hacks are never shared
39:53um so you got to be careful Kieran
39:54sharing that 10 on the podcast yeah I
39:57need to uh keep the rest of ourselves
39:59that's true I'm smiling because this is
40:02such a fun conversation and and Gina and
40:04Kieran you votes hit on like sort of the
40:06non-replicability of like human and I'm
40:08thinking about this Nikon ad that just
40:10like came out a few days ago which is
40:12like don't give up on the real world and
40:15unbelievable image of real places and
40:18like a fake prompt that like uh that
40:20would generate such real image and it
40:23was a very powerful message of like look
40:25like yes you know uh AI can do a lot of
40:28things but don't don't give up on the
40:29real world and it was very powerful and
40:33you know can AI come up with that I
40:35don't know but so like it it's it's very
40:38the on the consumer side a couple things
40:41uh going back to your question staff uh
40:43it was like what can the companies do to
40:45use it to your advantage one is the
40:48willingness to try as Gina said is is
40:52meaning uh at least for the consumers
40:54for the first time in a while they're
40:56willing to give it a shot they're
40:58willing to actually download and use the
40:59product and try it out and go through
41:01the app download and all the things
41:03which they weren't before so for the
41:06first time actually you have this window
41:07of opportunity where people are willing
41:09to give it a shot so now acquisition
41:11side uh you have this sort of potential
41:14window where you can appeal uh as like
41:17either first mover or a great product to
41:19to get a lot of sort of usage um you
41:22know Karen shared a zapier example where
41:24you could also use it uh very
41:26strategically to deepen your core
41:28product value like why are people using
41:30the product that you have out there and
41:31how do you actually use this new
41:34capability uh New Wave if you will too
41:38further strengthen and deepen the reason
41:40why people use your product
41:42and sure you can brand at Ai and make it
41:45sexy and like more people will be
41:47excited what have you but
41:49you know I learned one thing from snap
41:50where the the luxury the sort of lens
41:53product was a state-of-the-art AR
41:55product like augmented reality product
41:57we did not call it augmented reality for
42:00four years and we actually implicitly
42:02banned it in inside a company to call it
42:04an augmented reality because that's
42:07technology that's not the point the
42:09point is for users to giggle and use it
42:12and have that emotion so
42:14as you think about like these new
42:15technologies that are being enabled and
42:17Karen talked about the 90 and 10 net new
42:19things but there are these like ways to
42:22deepen the core product value of why
42:24your product exists that I think you
42:26know you can call it AI you don't have
42:27to call it AI it could be abstracted
42:28away it doesn't matter if it makes the
42:31job that you do much easier in two ways
42:34one either take care of the rot thing
42:36that you're doing in a mechanical turkey
42:37way great or you know all of a sudden
42:40deliver value that was impossible before
42:42because they were cost prohibitive or
42:44net new I think that's incredible so you
42:47know people you talked about you know is
42:50there a way to avoid the wave I don't
42:51think so I think it's kind of a tsunami
42:53if you sit there it's just going to wash
42:55so you do need to respond and how you
42:58respond I think is important where
43:00rather than saying oh there's like a new
43:02shiny thing we're gonna bolt it on I
43:04actually think it's going back to the
43:07hey like why do our product exist and
43:10why do customers use our product every
43:12day in a day in and day out how do I use
43:14this new capability to actually augment
43:16that I make that even easier so I get
43:20very very excited about uh you know
43:24existing companies as well as net new
43:26companies because they both now have a
43:29new thing why right like for existing
43:32companies there are ways to actually
43:33deepen your mode actually and then use
43:35this product abstract it away to make it
43:37so much more useful than before
43:40new companies all of a sudden you have
43:42this moment in time where people are
43:43willing to give it a shot and in the
43:45consumer side what I also get really
43:47excited about is there are these net new
43:49behaviors right I know you know we
43:52talked about how we'll never sort of
43:54fully rely on you know computers for
43:56creativity but you know the cost of
43:59creation has nearly gone to zero now
44:02the the post-production of this podcast
44:04the actual creation the ums and Oz and
44:07the the eye move and everything you can
44:09actually pretty easily do that through
44:11AI potentially with a lot of new
44:12products out there and then and all of a
44:15sudden you know maybe you don't need a
44:17huge cereal maybe you don't need all
44:18these Amazing Fantasy gadgets to to do
44:20what we do and the bar is lower the
44:23entry entry point is lower meaning that
44:25you know all of a sudden the marginal
44:27cost of creation overall whether it's
44:29image or video or prompt or Pi what have
44:32you it may be quite low which then means
44:35the new limiting factor is possibly
44:39and that's that's a very very fun place
44:42to be speaking of creativity I wanted to
44:44end this with a little bit of a
44:46challenge a prompt for the three of you
44:49um you know we talked about a lot of
44:52um we've talked about product and
44:54distribution and how those come together
44:56and you know one product that I work on
44:58is the a16z podcast and if we think
45:01about austerity and trying to grow in
45:04turbulent times when you know maybe
45:05budgets aren't as high if we had ten
45:07thousand dollars to grow the a16z
45:09podcast given all the channels that you
45:12could invest in how would you think
45:14about that and please don't just tell me
45:15build a better product because we're
45:17working on that but on the distribution
45:20side how would you use that money well
45:22from someone who has built to try to
45:25build two podcasts my favorite word for
45:27podcast growth is like grind it is yeah
45:29grind you know because there's not a lot
45:31of meaningful things that work for
45:33podcast promo so like one idea I'll give
45:36you I'll just give you a wacky idea
45:37instead okay that may or may not be
45:39usable so one of the things I would do
45:41in this has kind of been done I guess
45:43before but it might be a more novel way
45:44to do it is like get your audience like
45:46Pick 10 entrepreneurs from way back in
45:48the past that are no longer with us like
45:50all the way back to Henry Ford and all
45:52these incredible entrepreneurs get your
45:54audience to vote on like what's the
45:55number one person you would like to hear
45:56interviewed in this pod uh get them to
45:58vote whatever the top one is use a 10K
46:00to give to someone to create an AI
46:02version of that like person you know
46:03it's pretty easy to create a deep fake
46:05version and then you should just
46:06interview them for half an hour and like
46:08put that out into the interweb and
46:10that's probably a good way to like get
46:13interest in the podcast through
46:14something that will hopefully get a lot
46:17of social traction I like that Gina
46:19Brian what would you do I have an
46:20equally wacky idea uh being of consumer
46:23mindset I I think one of the distributor
46:25distribution channel that works these
46:27days is Tick Tock so 10K not not a lot
46:30but uh you know maybe maybe the way to
46:33you know fight uh find a intentionally
46:35quite controversial topic we talk about
46:37it put a use AI to splice it up so that
46:41it's like entertaining and like very
46:42easy like to the point and then
46:44essentially the call out is fight us
46:46stitching coming fight us and like a hot
46:49take find us and you know actually use
46:53that remix culture of the internet to
46:55let that sort of do what it does and you
46:58know the reason why the dollar would be
46:59interesting is sometimes it's
47:01interesting to seed these things
47:03so have like one or two you know
47:05influence or what have you or like over
47:06Founders who like we can you know invite
47:08over and like here's a piece of content
47:10now go debate that and actually be
47:13pretty brutal because like people love a
47:15good fight and put it out there and see
47:18how that does I actually kind of love
47:20this idea even just the tag fight us is
47:23so strong that's cool
47:25yes I thought about it and I have a
47:27boring idea but this is what I would do
47:30I think that your bet here would be to
47:32try to increase your Audience by looking
47:34at places where your audience isn't but
47:37you have a huge potential audience and
47:39for me that would be
47:42people in older demographics that are
47:44very business savvy and have been very
47:46active their whole lives and kind of get
47:48technology but are a little bit out of
47:49the loop and don't know what a16z is and
47:51I've never heard of anything related to
47:53it and I have had conversations with
47:55brilliant people in finance who are
47:57maybe like in their 60s let's say maybe
48:00even 50s who have no idea what a16z is
48:02which in my world is crazy and those
48:04people read The Wall Street Journal and
48:06those people read the New York Times so
48:08I would go and do you know something
48:11with one of those massive Publications
48:13to reach those audiences the thing about
48:15growth is that it's not just about
48:17growing but you have to grow in terms of
48:18like the right people so one way to game
48:21your question which is something that
48:22girls teams do all the time is I'm just
48:24gonna go to a country where buying a
48:26user or buying a listener is super cheap
48:28and I'm gonna go and like you know do
48:31some sort of massive campaign because
48:32ten thousand dollars go so far in like
48:34southeast Asia or like Latin American
48:36countries and then just get a bunch of
48:38people to listen to Oasis but it won't
48:40actually list matter because they're not
48:42but that's just you know that would be a
48:44gameable response which I just want to
48:46leave here because I think
48:48um it's really core to growth in order
48:50for you to actually come up with a good
48:51growth strategy you have to think about
48:53how can you game this problem and if you
48:55can easily game it then you're looking
48:59thank you all this was so fun
49:03thanks for listening to the a16c podcast
49:05if you like this episode don't forget to
49:07subscribe here on YouTube to get our
49:08exclusive video content we'll see you
![How To Succeed As A NEW & YOUNG Realtor [Deals Every Month + Luxury Listings]](/_next/image?url=https%3A%2F%2Fi.ytimg.com%2Fvi%2FrtrRghgTayc%2Fhqdefault.jpg&w=3840&q=75)
