0:00 I think the way that you find Alpha is
0:02 either by doing things that no one else
0:04 knows about I think another aspect is
0:05 probably doing things that everyone is
0:07 convinced will not work I think a good
0:09 leader needs to know how to do everyone
0:10 on their team's job but poorly and I
0:13 think the poorly part is important I
0:14 would always skew more Junior I think
0:16 hiring for potential especially early on
0:18 in the business is far more important
0:20 ready to go
0:21 [Music]
0:32 George I'm so excited for this dude ramp
0:34 is one of my favorite companies to
0:36 feature I've heard so many great things
0:38 about you so thank you so much for
0:39 joining me today no thank you so much
0:41 for having me now listen as we just said
0:43 I know a show is going to be great when
0:45 I actually have to do very little work
0:47 because you give me such great
0:48 suggestions you said to me before growth
0:51 is just science and most marketers are
0:54 bad at science I was always terrible at
0:57 science so you it doesn't bode well for
0:58 me what did you do you mean by that
1:00 statement the goal of growth is to
1:02 figure out how to grow the business and
1:04 usually like early on that's very top of
1:06 funnel Focus how do you figure out how
1:07 to get more leads how do you figure out
1:09 a channel that works and is repeatable
1:11 and has uh predictable outputs given
1:14 some inputs and the honest answer is
1:15 like no one really knows every business
1:17 is different so even if you understand
1:19 from past experience something that's
1:21 worked usually just taking like that one
1:23 Playbook or that one tactic and copy and
1:25 pasting it to a new company like
1:26 generally doesn't work and I think
1:28 that's the tendency of of a lot of
1:30 marketers so when I say that like growth
1:31 is mostly just science I mean that you
1:34 kind of have to come in with a blank
1:35 slate form a hypothesis and then run a
1:37 bunch of experiments and you'll be
1:39 surprised about what works you'll be
1:41 surprised at what doesn't work um but
1:43 ultimately if you if you run enough of
1:45 those experiments you'll you'll find
1:46 something uh and I think that's usually
1:48 lost on a lot of marketers because they
1:50 don't think in terms of experiments they
1:51 think in terms of like what do I know
1:53 and how can I apply it here is that
1:55 because of the profile of person that
1:57 they are that they don't think in that
1:58 way partially it's probably partly the
2:01 profile but it's also partly just like
2:02 how you think like I think the type of
2:04 person that would go and become a
2:05 chemist like me or an engineer it's
2:08 probably very different than someone
2:09 that's like hey I want to become a
2:11 writer or I want to get into comms or PR
2:14 like very valuable skills but very very
2:16 different from the way in thinking or
2:18 way of thinking that would make someone
2:19 successful like scoping an experiment
2:21 and thinking about like what my
2:23 hypothesis is and how I'm going to
2:24 measure results it's just like a
2:25 different part of your brain like i'
2:26 would be a terrible writer for example
2:28 you said about running experiments there
2:30 and running a number of them can I ask
2:32 in your mind is growth about increasing
2:35 performance by 1 or 2% in many different
2:37 areas or is it about needle moving
2:41 chapters of a company and being much
2:43 more pivotal in that respect the short
2:46 answer is it has to be both and what I
2:48 mean by that is if you're doing a good
2:51 job and it depends on the stage of the
2:52 company but in general if you're doing a
2:54 good job you're thinking in terms of
2:56 different time Horizons so you have to
2:58 have some like bucket of bets or
3:00 experiments that are going to be those
3:02 like big swing huge step changes and
3:04 impact but those are generally highrisk
3:07 High rewards you can't just do that
3:08 otherwise you're going to fail and miss
3:09 your number this quarter um at the same
3:11 time like you need to have some bets
3:13 where you have high confidence but
3:14 probably not going to move the needle a
3:16 ton but it'll help you get the 2 3 4 5%
3:19 Improvement like this quarter and then
3:21 you have everything in between so like
3:22 ideally you're being intentional with
3:25 how you're allocating your resources
3:26 across like everything from like the
3:28 very long term to the very short term
3:30 and how you make those bets and how you
3:31 allocate those resources is actually
3:32 like a conversation you should probably
3:34 have with Finance with leadership and
3:36 align it to the goals of the company but
3:38 that that's my way of thinking about it
3:39 you said there about bets and I often
3:42 think about growth very much like
3:43 Venture which is you place a number of
3:45 Investments SL bets and You observe and
3:48 then you wait to see what works and
3:50 double down do you agree with that
3:52 analogy how do you think about what is
3:55 enough bets and failure rate Associated
3:58 oh yes so it's absolutely a portfolio
4:01 but I think like an Investment Portfolio
4:04 it depends what you're optimizing for
4:05 and it depends on the stage of the
4:07 company so your risk tolerance is going
4:09 to depend on the stage of the company
4:11 it's going to depend on if you're
4:12 optimizing for growth or profitability
4:14 or whatever it might be so it is a
4:16 portfolio um I think that you should
4:19 assume if you're doing things right you
4:20 should assume the majority of your bets
4:22 are going to fail which is why I always
4:24 believe that velocity is probably more
4:26 important than getting things perfect uh
4:28 but it is a spectrum like you could do
4:30 some really well-controlled rigorous
4:32 experiments and it's going to be
4:33 incredibly academic and you're going to
4:35 learn something but it might take you
4:36 like a year to like say something
4:38 conclusive um or you can just run a
4:40 bunch of experiments at incredibly High
4:41 Velocity it's kind of sloppy and
4:43 similarly you might actually you might
4:44 not learn anything uh because you made a
4:46 bunch of mistakes or didn't fully think
4:48 through how you'd measure something so
4:50 you kind of have to balance it but in
4:51 general i' probably bias to run thing
4:54 run more things faster than run
4:56 something perfectly does philosopy
4:58 impact quality of conversion and what I
5:01 mean by that is if we're a little bit
5:02 sloppy but with very high velocity we
5:05 won't spend so much time on the visuals
5:07 for that campaign the graphics for this
5:10 it's velocity but it's not as good how
5:14 do you think about that tradeoff I think
5:16 if you had to choose velocity is more
5:18 important but there's there's a other
5:21 side of that coin which is if you're
5:22 just doing a bunch of sloppy things it
5:23 doesn't matter how many things you run
5:25 you're not going to actually learn
5:26 anything what did you do that was sloppy
5:29 that you wish wasn't
5:31 sloppy um I'm reminded of this one time
5:34 I won't say which company uh but we had
5:36 a web page and the web pages conversion
5:38 rate was trending down and it was
5:40 trending down for a long time and it was
5:42 our number one channel and it was like
5:44 hey what are we what are we going to do
5:45 to solve this and there were kind of two
5:46 paths there was one where we could run a
5:48 bunch of individual well-controlled
5:50 experiments and understand hey changing
5:52 this button or changing this H1 is like
5:54 what's going to improve the page and it
5:55 worked or it didn't or we could just run
5:58 everything at once kind of use our gut
6:00 kind of use our past experience and like
6:02 hope that it worked and we ended up
6:03 going that that lad route uh which was
6:05 definitely the sloppier route the less
6:07 rigorous experiment um it had ended up
6:10 work it did end up working uh so we did
6:12 end up like 3x in the webpage conversion
6:14 rate like over the course of a couple
6:16 weeks and helped us hit our number that
6:17 quarter I say that that was like a
6:19 mistake because we never actually knew
6:21 what did or didn't work and we had to go
6:22 back and undo a lot of the um changes
6:24 that we made once we did end up AB
6:26 testing them but I think that goes back
6:28 to like the portfolio that's like we
6:30 were operating on a very short time
6:31 Horizon so we had to like optimize for
6:33 velocity versus like rigor and then once
6:34 we were no longer under the gun we went
6:36 back and optimized for riger to
6:37 understand what worked and what actually
6:38 didn't how do you think about giving
6:41 something enough time to know if it
6:43 works we want high velocity and we need
6:45 to move on but also sometimes it takes a
6:48 little bit of time content in particular
6:51 honestly I would say 12 to 18 months
6:54 yeah how do you think about enough time
6:56 but not being slow it goes back to the
6:58 portfolio like if if you're going to
7:00 allocate 20 30% of your time to like
7:02 longer term bets that's fine be okay
7:04 with the fact that you're not going to
7:05 get results anytime soon um having said
7:08 that it would be great if you could
7:09 figure out what some leading indicators
7:10 are or scope down the experiment so you
7:12 could get some signal that like hey we
7:14 have confidence this Willer won't work I
7:16 think in general I've always tried to
7:18 prioritize experiments based on impact
7:20 and effort I think those are the obvious
7:21 ones but also confidence and time to
7:23 results like if you're really confident
7:24 something's going to work you should
7:25 just do it um if you're really
7:27 unconfident that something is going to
7:29 work but the time to results is really
7:30 fast you should do that as well and I
7:33 think usually those two dimensions are
7:34 lost when people are are prioritizing
7:35 they tend to just go for impact and
7:37 effort and not think about confidence or
7:39 time to results when you think about
7:41 impact you also said there the word
7:43 indicator I think it's really important
7:44 to understand what you're actually
7:46 trying to move what the core objective
7:48 is what's your biggest advice to
7:50 Founders and growth teams on how to set
7:53 the right metric that you want to move
7:56 that's where I think the experimental
7:58 design and the rigor of experiments
8:00 comes in uh you'll you'll have a
8:02 hypothesis and I think most marketers
8:04 tend to jump to let's go do something
8:07 let's go launch something but the
8:08 experimental design like understanding
8:10 what you're actually able to measure
8:12 what you will measure how long it'll
8:13 take to get that result if it's
8:14 statistically significant or not all of
8:17 that honestly like you either need to be
8:18 able to do that math yourself or go find
8:20 someone on a data team and go work with
8:22 them or acknowledge the fact that hey we
8:24 don't actually have a good way to think
8:25 about this or measure this and that's
8:27 okay we're going to go collect some
8:28 qualitative data and it means we might
8:30 be wrong I think we just need to be hon
8:32 folks need to be honest about that
8:34 upfront and Define that upfront is
8:35 probably how I would think about it so
8:37 we have this portfolio of bets okay and
8:40 then we see one that starts to work do
8:43 we immediately double down on it then
8:45 how do we know how much to double down
8:46 on do we set a benchmark of what good is
8:49 versus great how do we think about that
8:51 yeah yeah absolutely you should like
8:53 triple down on that uh I think that
8:55 another mistake most startups make is
8:57 they see something that works and
8:58 they're like okay great let's increase
9:00 our spend let's double it let's triple
9:01 it ideally if it's working you take that
9:04 channel to saturation as quickly as
9:06 possible like if you if you graph out
9:08 what the results are um over a long
9:10 period of time you're probably going to
9:11 see it approach an ASM toote you're
9:13 going to see the incrementality of those
9:15 results start to like Decay if you just
9:17 burn a shitload of cash though on a
9:19 channel super quickly say you're
9:20 spending I'm 10K and you're like it
9:23 works let's put 200k on it the 200k will
9:27 not be nearly as efficient as that 10K
9:30 would it not have been better to do 30
9:33 50 70 and gradually get up there than
9:36 just whack it as hard as possible I'm
9:39 naive so so yes but it depends so if
9:42 you're able to graph that response curve
9:45 going from 10K to 200k obviously you're
9:47 probably not going to be able to go to
9:48 10K to 200k overnight um if you were
9:50 that actually would probably be a good
9:51 problem to have uh but yes if you're
9:53 able to like graph that response curve
9:55 and see when something goes from linear
9:56 to start to decay in terms of response
9:59 that tells you it's like okay we're no
10:00 longer getting an expected output given
10:03 the input and then we have a
10:04 conversation on if the returns are worth
10:06 it so figuring out the ASM toote where
10:08 things start to actually Plateau is is
10:10 the most important thing and I think
10:12 most companies or most startups at least
10:14 they get to that ASM ASM toote too
10:16 slowly and they should really be scaling
10:19 much much faster if they find something
10:20 that works um having said that I think
10:23 like most things saturate probably more
10:27 slowly Than People expect um and so
10:29 going from 10K to 200k might not be as
10:31 insane as it sounds uh but it depends
10:33 you've got to watch it do CS get cheaper
10:37 over time as brand becomes better known
10:40 you become more branded in an ecosystem
10:43 or do they get more expensive as you
10:44 saturate the core target market and you
10:47 have to expand into maybe less directly
10:50 relevant icps the only right answer is
10:53 that like yes CS become more expensive
10:55 like as you get more and more market
10:56 share it makes sense that it's every
10:59 incremental acquisition is going to cost
11:01 more than previous um having said that I
11:03 think the reality is that's usually not
11:05 true usually you figure out new products
11:07 to sell that actually make the ltvs
11:09 improve you figure out new geographies
11:11 to break into you figure out new
11:12 channels that work you figure out maybe
11:14 that hey actually combining different
11:16 channels has a halo effect and you're
11:17 not fully capturing that and what the
11:19 customer acquisition cost is so the
11:21 reality is that no there's it takes a
11:24 really long time to get to the point
11:25 where like the macro effect of
11:27 saturation is hitting your CAC um but
11:30 that's how I think about it do you think
11:32 early stage Founders should look at LTV
11:34 often C to LTV is kind of the hailed
11:36 metric George you know as well as I do
11:38 it's so difficult to calculate LTV in
11:41 any product let alone early stage
11:43 products how do you think about LTV and
11:45 its utility value to early stage
11:47 Founders I think it's a reasonable
11:49 framework I think you have to have some
11:51 threshold that you agree on as a
11:53 business like this is what we're willing
11:54 to spend and this is what we think a
11:56 customer is worth but the reality is
11:57 it's like false Precision like you're
11:59 not going to know where your LTV is if
12:00 you've been in business for a year um or
12:03 6 months or whatever it might be so I
12:05 wouldn't over index on that false
12:06 Precision I would instead just
12:08 acknowledge that there's some threshold
12:10 we're going to be okay with uh in terms
12:11 of spending to acquire a customer and
12:13 what that customer is worth and that
12:14 should change over time as you learn
12:16 things and run experiments when you're
12:18 running experiments do you have culture
12:20 challenges in terms of maintaining moral
12:23 if you shut off someone's baby and what
12:26 I mean by that is someone's really been
12:28 working hard on the YouTube the
12:30 Instagram the SEO and you say George
12:34 it's 3 months we're not seeing the
12:35 returns needed cut it but they are very
12:38 attached to it how do you think about
12:40 that no one should be that attached to
12:43 the experiment that they're running like
12:44 I think that is a cultural problem I
12:46 think that they should acknowledge that
12:48 the vast majority of what they work on
12:50 is going to fail and that if they're not
12:52 failing like honestly they're probably
12:53 not doing their job well uh you need to
12:55 be running a bunch of stuff it needs to
12:56 be unique it needs to be creative and
12:58 that means that most is not going to
12:59 work if you're that attached to
13:01 something that's working you should be
13:03 very very high on the confidence aspect
13:05 of how we prioritized it and in that
13:07 case like maybe we were wrong we should
13:08 have that conversation but you
13:09 definitely no one should be that
13:10 attached to any experiment how do you
13:13 use post-mortems to effectively analyze
13:16 the success rate of different
13:18 experiments programs um I would say like
13:21 you should be doing premortem and
13:22 postmortem I think that's part of like
13:24 good experimental design like
13:25 understanding what are the different
13:26 failure modes for this and and
13:28 acknowledging like what the probability
13:29 of those failure modes are if when and
13:31 so you do a sorry just so I understand
13:33 because so many Founders get really
13:34 granular and like get notebooks out for
13:36 like pre morms this is right before
13:38 we're about to start we plot out the
13:40 three main things that could likely kill
13:42 this
13:43 project that's one way to do it uh I
13:45 would actually like get more specific
13:47 than that when you're planning the
13:48 experiments like why would this fail
13:49 it's like oh we're not going to have a
13:51 large enough sample size as we predicted
13:53 or like there's a million things you
13:54 should probably write out those million
13:56 things and then I think what's more
13:57 interesting is when you do a postm
13:59 if the experiment failed for something
14:01 that you didn't actually anticipate
14:02 something that you didn't like factor
14:03 into your your experimental design that
14:06 I think is an interesting conversation
14:07 but if it's something that you had kind
14:08 of anticipated and maybe the probability
14:10 was wrong or whatever it might be that
14:12 is less interesting and I probably
14:13 wouldn't even do a postmortem for that
14:15 how often is the premortem the reason
14:19 why something didn't work how often do
14:21 you get it right I would say for the
14:24 high probability high confidence bets
14:26 the things that you're doing to hit a
14:28 number this or even this year I I think
14:30 pretty high it's probably like 90% plus
14:33 um for the big swings there's always
14:35 like some Black Swan there's always
14:36 something that like you cannot
14:38 anticipate you had no idea and that's
14:40 where the postmortem actually I think is
14:41 valuable and trying to figure out what
14:43 you learned and how it might apply to
14:44 either other big swings or um maybe even
14:47 like higher confidence bets like that I
14:49 think is actually the much more
14:50 interesting and more valuable
14:51 conversation I think every growth team
14:53 is going well I couldn't predict
14:55 Co yeah that is very true that was a
14:58 tough one for us all um okay so we have
15:00 that as the premortem for the postmortem
15:03 how do we structure that who's invited
15:06 what does that look like so I think the
15:07 person that was ultimately like the drri
15:09 for an experiment the person that also
15:11 hopefully Scopes the experiment they
15:12 should write the postmortem and I think
15:15 they really need to be clear on number
15:17 one did we run the experiment well um or
15:19 did it fail for some reason that we
15:21 could have avoided uh I think like
15:24 outside of did we exper uh did we did we
15:26 scope the experiment well it's did learn
15:29 something that could be generalized to
15:31 other aspects of the business I think
15:32 that's an important aspect to have and
15:34 then I think like the third most
15:35 important component is just making sure
15:37 that the right cross functional
15:39 stakeholders are there so the folks that
15:40 can learn from it the folks that can
15:42 prevent the failure in the future um I
15:44 think that's maybe like the third most
15:45 important aspect of of a postmortem but
15:47 beyond that like the structure I don't
15:49 think actually matters that much it's
15:50 going to be very company dependent what
15:52 do you like to do is it a notion is it a
15:54 Google doc when do you send it out
15:56 because you want people to have time
15:58 with it but not too much time how do you
16:00 think about that yeah I'm a Google doc
16:02 person uh I think the per the drri
16:03 should write it up they should send it
16:05 out in advance at least like 24 hours in
16:07 advance so people have time to like
16:08 think about it um but ideally like I'm
16:10 more in favor of of a live conversation
16:12 rather than add some comments have a
16:14 conversation in the comments in the doc
16:15 and then like have a maybe more boring
16:17 live conversation so I'm I'm I'm big on
16:19 live I I really like one of the
16:21 suggestions which is like are there any
16:22 learnings that could be generalized to
16:24 other aspects of the business is there
16:26 an example of one in the past that you
16:29 could share just to illustrate that a
16:31 little bit I'll give you a really
16:33 tactical example um a really tactical
16:35 example was we were doing some very
16:38 basic AB testing on our homepage and we
16:41 saw that a red button by far
16:43 outperformed anything else now red is
16:45 like as a button is generally like a bad
16:47 idea it has a negative connotation it's
16:49 like something is wrong don't hit the
16:51 red button but we couldn't ever find
16:53 anything that outperformed on a pure AB
16:54 testing basis like that red button now
16:56 it turned out when you actually went and
16:58 looked at the data by segment although
17:00 it outperformed in general it's severely
17:02 underperformed for like the Enterprise
17:04 segment um it's like a great example of
17:06 like simpsons's Paradox when we took
17:08 that information and like removed that
17:10 from the web team and applied it to the
17:12 content team and applied to everyone
17:13 else that was using the red button as a
17:15 best practice that was like incredibly
17:16 valuable because most of the content we
17:18 were generating was for Enterprise most
17:19 of the webinars we were generating were
17:20 for Enterprise like most of the direct
17:21 mailers we were sending were for
17:23 Enterprise uh so what we were actually
17:24 doing by saying hey red is the best
17:26 button or or sorry the best practice was
17:28 actually like decreasing the performance
17:30 of all of these things we were doing
17:31 that were specific for the Enterprise
17:32 segment can I ask you does grow sit in
17:35 its own team or there are you in the
17:38 product team are you in the marketing
17:40 team where do you sit I think my
17:42 personal opinion is I think growth
17:43 should be independent I think like the
17:45 growth team's mandate should be to
17:47 figure out how to grow the business and
17:49 that should be more than just marketing
17:50 it should be more than just product it
17:52 should mean that like you have the
17:53 mandate to do whatever is the highest
17:55 leverage and to me that means you
17:57 probably should report report to I
17:59 honestly that's one of the reasons why I
18:01 love ramp is the growth team reports to
18:02 one of the co-founders um I think some
18:04 companies have Chief growth officers I
18:06 think other folks have growth
18:08 organizations that are more like SWAT
18:09 teams and kind of roam between different
18:11 parts of the business ultimately it does
18:13 depend on on the business but I think
18:14 they should be as independent as
18:15 possible youve said to me before about
18:18 Seeking Alpha in growth it was a
18:21 cliffhanger because I had no idea what
18:23 you quite meant by it what did you mean
18:25 by Seeking Alpha in growth yeah uh I was
18:28 that in terms of for I was I was
18:30 stealing Alpha from from investing
18:33 terminology like what's your unfair
18:34 Advantage is maybe a better way to put
18:35 it and I think like that is the key to
18:38 being a good growth team like you got to
18:39 figure out something that is not
18:41 saturated and that other people ideally
18:42 like are not doing and I think the way
18:45 that you find Alpha is either by doing
18:47 things that no one else knows about
18:49 either most likely because it's so new
18:52 uh like think about when Tik Tok first
18:53 came out I don't think any B2B Brands
18:55 were thinking about advertising on Tik
18:56 Tok I don't think there even was
18:57 advertising on Tik Tok when it first
18:58 came out but the folks that were really
18:59 ahead of the game they're like
19:00 eventually they will be advertising on
19:02 Tik Tok and it's going to be totally
19:03 saturated with consumer Brands to begin
19:05 with but eventually there will be a B2B
19:07 opportunity and when they were the first
19:08 ones to run those experiments they
19:10 probably had huge returns um so anyway
19:12 doing things that I think folks don't
19:14 know about is like one aspect I think
19:16 another aspect is probably doing things
19:17 that everyone is convinced will not work
19:19 uh I remember first suggesting we try
19:22 Direct Mail like years and years ago and
19:24 it was like why would Direct Mail work
19:26 like junk mail to people's homes like
19:27 that absolutely won't work and it became
19:29 like one of our most successful like
19:30 biggest channels after like a few
19:31 iterations so um I'm happy to chat about
19:34 where you find these things but you've
19:37 can we can we actually because it's it's
19:39 like it totally makes sense like do
19:41 things that people don't know do things
19:43 that people don't believe how do you
19:45 find them um I think this actually is is
19:48 a good segue into like how do you just
19:50 learn in general uh as a a member of the
19:51 growth team or someone at in a company
19:54 in general I would say that there's
19:56 three areas like you can learn
19:57 academically you can read a book and
19:59 learn what's worked for other companies
20:01 or other growth folks or uh companies in
20:03 the past even are great like aspects of
20:06 of being able to learn so learning
20:07 academically is one aspect um I think
20:09 you can also learn from your peers so
20:11 you can go and learn from Folks at other
20:13 companies growth people at other
20:15 companies I think that's a great like
20:16 area of of uh of finding Alpha um
20:19 granted probably not the the best unfair
20:22 advantage since someone else knows about
20:23 it but I actually think the most
20:24 interesting is to go and learn from
20:26 other niches so like other verticals
20:28 other geog Ries especially if they're
20:29 tangential to what you're doing uh a
20:32 great example is like WhatsApp like
20:33 WhatsApp as a marketing channel is not
20:35 huge for most brands uh or most
20:37 companies in the United States but it's
20:38 massive uh in a lot of like
20:40 International regions and I think like
20:42 great could we go try like WhatsApp
20:44 instead of sending emails um it'll
20:47 probably fail but again if you're doing
20:48 enough of these experiments you'll find
20:50 something that works that no one else is
20:51 doing my question you on the number one
20:53 academically is that not just learning
20:55 playbooks and do you worry that then
20:57 you'll speak to to your friend George
20:59 who will tell you what worked at Sam
21:01 Sara but ramps a totally different
21:03 business and so respectfully to your
21:06 point on playbooks earlier they're not
21:07 applicable often does academic learning
21:10 really work I think it does but it
21:13 depends how you look at it so yes there
21:15 are playbooks that you can just take and
21:16 copy and paste there are also playbooks
21:18 that you can look at and think
21:19 critically and adapt and then there are
21:20 playbooks you kind of come up with from
21:21 scratch I think a lot of academic
21:24 learning is taking a past Playbook and
21:26 adapting it and like the example of
21:28 always comes to mind for me is in in the
21:30 world of attribution in marketing so
21:32 like media mix marketing oh sorry media
21:34 mix modeling or marketing mix modeling
21:36 depending on who you talk to MMM like
21:38 everyone's talking about it now is the
21:39 best way to do attribution blah blah
21:42 blah uh that's a really old concept
21:44 that's like from the old like Mad Men
21:46 style advertising days in like the 1950s
21:48 and 60s so I think that's a good example
21:50 where hey you could actually go and look
21:51 at how did they measure the impact of
21:53 ads and literal newspapers before
21:55 anything was digital uh back in like the
21:58 1950 s or 60s and how do you actually
22:00 adapt that Playbook or that methodology
22:02 to modern day like digital advertising
22:05 or Modern Day digital marketing as a
22:06 whole um there's probably a million
22:08 examples like that actually Direct Mail
22:09 is probably another example of that
22:10 there are so much what did you see in
22:13 direct mail that no one else saw because
22:16 so many of your friends were like George
22:17 crushed it on direct mail and no one
22:19 thought this was a good idea so what did
22:21 you see that no one else so no one else
22:23 was doing it and it was incredibly
22:25 scalable like that was pretty much it
22:26 like the fact that no one else was doing
22:27 it it's like okay that's interesting we
22:29 should try it but the fact that if it
22:30 works it' be incredibly scalable and the
22:32 fact that you can run it with very large
22:33 sample sizes meant that we could run a
22:35 lot of experiments and parallel and
22:36 learn really really quickly like there's
22:38 very few channels where you can go like
22:40 hey tomorrow let's go reach like 200,000
22:42 people uh direct mail email those are
22:44 like kind of some of the only channels
22:45 that that would allow you to do
22:46 something at that scale and th run a lot
22:48 of experiments what is no one trying
22:50 today that you think is
22:53 interesting um I don't know if I can
22:55 give away my secrets but I think the
22:56 honest answer is like on the B b side
22:59 and I don't think this is super Cutting
23:00 Edge but influencer marketing um maybe
23:03 like a year or two ago this would have
23:05 been a little bit more Cutting Edge but
23:06 everyone's always thought of influencers
23:08 and and user generated content maybe uh
23:10 to get more specific as more of a b Toc
23:13 tactic that worked really well I would
23:15 argue that works just as well and maybe
23:17 even better in the B2B world like
23:18 especially if you're moving at Market
23:20 moving to Enterprise yeah there's
23:21 definitely unfair Advantage there but
23:23 it's a lot of work it's a lot of work um
23:26 how do you think about attribution and
23:28 challenge of capturing conversion you
23:31 know I'm interviewed Nick at revolute
23:33 and Antoine who's the head of growth at
23:35 revolute and their biggest mental shift
23:37 for both of them was the power of brand
23:39 marketing but both being aware of the
23:41 challenge of having no freaking idea
23:43 what it does to your revenues actually
23:45 how do you think about the importance of
23:47 knowing source of Revenue versus brand
23:51 marketing uh goes back to the portfolio
23:53 I I think that you have to be okay with
23:55 the fact that most brand marketing is a
23:57 long-term bet it's it's high risk it's
23:59 high reward uh at least that's how a
24:01 growth person would probably think about
24:02 is there a stage of company where it
24:03 becomes interesting I think if you if
24:05 you're seeing all of your core direct
24:07 response channels start to saturate then
24:10 it's something that you have to start
24:11 doing and the assumption is that like
24:12 investing in brand is going to do one of
24:14 a few things it's either going to make
24:16 folks that have a problem and are aware
24:18 they have a problem aware of you and I
24:20 think that's how probably most companies
24:21 start thinking about it but the more
24:23 interesting aspect is like true demand
24:25 generation it's like hey there are
24:27 people out there that don't realize they
24:28 have a problem and brand marketing like
24:30 instead of being like Hey try XYZ
24:32 product or try XYZ company it's like
24:34 this problem exists like there's
24:36 actually a better way and of course like
24:37 we're the better way but that I think is
24:39 like a much more interesting type of
24:41 brand marketing to open up like demand
24:44 for a new part of the market that that
24:45 you aren't able to reach with
24:47 traditional channels can I ask just on
24:49 the bets that we continuously go back to
24:51 how much is too much concentration when
24:53 you think about you know portfolio it's
24:55 like hey traditionally Adventure you
24:57 don't want to be
24:59 definitely not more than 10% in one
25:01 single investment how do you think about
25:03 too much concentration in a a growth
25:06 portfolio if you're doing things right
25:07 the concentration will change um it's
25:10 almost like when you find something that
25:11 works if you do a really good job of
25:13 saturating it as quickly as possible you
25:15 will be by almost definition like really
25:16 concentrated there at least for a period
25:18 of time it's really how quickly you can
25:20 diversify and stack different bets and
25:23 different wins so that you're not
25:25 concentrated for too long of a period of
25:27 time um but I I think concentration in
25:29 of itself is not a bad thing it means
25:30 that you're doing a good job maximizing
25:32 an area how do you think about
25:34 communicating growth goals to other
25:37 elements of the organization you're
25:39 sitting there in your independent growth
25:41 team say and you also have to work with
25:43 product and you also have to work with
25:45 marketing how do you work together most
25:47 efficiently to communicate this is what
25:49 I'm going after I think communicating
25:52 the way growth thinks about things is
25:54 probably like more important than like
25:56 what specifically they're doing as long
25:57 as everyone aligned that like the growth
25:59 team's job is actually aligned with
26:01 everyone else's in the company which is
26:02 like we need to be successful like
26:03 growth team's job is not to make anyone
26:05 happy it's to make the business
26:06 successful and that might mean like for
26:08 a period of time working really closely
26:10 with product or really working really
26:11 closely with product marketing or some
26:13 element other element of the business
26:15 but really they should recognize and
26:17 growth should be able to communicate
26:19 that we're bringing a unique skill set
26:20 and a unique point of view that
26:22 hopefully it's complimentary to their
26:23 skill set and their point of view But
26:24 ultimately the same goal of of making
26:26 the business successful um
26:28 so I know that's a little bit meta but
26:30 that's that's how I would describe it
26:31 you're an angel in my company George and
26:33 we're sitting down for a coffee and
26:35 we're we're going for some advice and
26:37 I'm at about a million in Revenue um
26:40 I've just raised a series a is now the
26:43 time to bring in a head of growth and
26:44 how do you advise me on when is the
26:46 right time to bring in someone for
26:48 growth and what type of person is it
26:50 senior or Junior I would always skew
26:52 more Junior I think hiring for potential
26:54 especially early on in the business is
26:56 far far more important I actually think
26:58 it would be a mistake to hire someone
26:59 more senior like if you're series a
27:01 hiring a really smart generalist that
27:03 can think in terms of first principles
27:05 and that can think logically in terms of
27:07 solving problems is probably more
27:08 important than anything what background
27:11 like solving problems and that kind of
27:12 generalist what background do you find
27:15 is best unless you're trying to solve a
27:17 very specific problem and you have a
27:19 really high confidence that this is the
27:20 right problem to solve I would not hire
27:22 a specialist or someone with a
27:24 traditional markeing background I think
27:25 hiring someone that has demonstrated
27:29 they can think logically and they can do
27:31 do math um is really really important
27:33 early in the early days so that might be
27:35 Engineers that might be X Finance folks
27:37 uh that might be an EXC consultant that
27:38 hated Consulting and like really wanted
27:40 to stick with something longterm and get
27:42 their hands dirty I think those make
27:43 generally the best first uh first growth
27:46 hires but really you should just be
27:48 trying to assess for potential which I
27:50 think is a combination of like can they
27:52 have Vision can they kind of see areas
27:54 where they can go like can they acquire
27:55 new skills really really rapidly um and
27:58 are they like internally motivated is
28:00 there any profile before we actually dig
28:02 into skills and skill detection is there
28:04 any profile where you're
28:06 like I don't love that background It's
28:08 just tough to be good at growth with
28:10 that background yeah if you've been at a
28:12 company especially a company that's an
28:14 order of magnitude or larger or bigger
28:16 for more than a few years it's it's just
28:18 really hard to then change your mindset
28:20 and go to a smaller startup and think
28:22 about like how hard you have to like get
28:24 your or how much you have to get your
28:26 hands dirty and how differently you need
28:27 to think about things like I think if
28:29 you've been at a much much larger
28:30 company you're going to tend to think
28:32 more in terms of playbooks and and rely
28:33 more on your past experience than on
28:35 what you can learn from others or what
28:37 you can learn from from maybe like other
28:38 geographies or other companies or other
28:40 peers uh so going from Playbook to first
28:43 principal thinking is generally
28:44 difficult to do so I would I would I've
28:46 personally like stayed away from that
28:47 type of profile h no I I totally
28:49 understand that perspective on the skill
28:51 detection how do you run an interview
28:54 process then again where you're advising
28:56 me I I have a number of applic and we
28:58 have a candidate pool how should I spend
29:02 the first meeting what questions should
29:04 I ask what should I try and uncover I
29:06 think if you're able to do some sort of
29:09 back Channel and some sort of test in
29:11 that first interview that is going to be
29:14 like the most valuable highest signal
29:15 thing you can do like when you say back
29:17 Channel or test what do you mean do you
29:18 mean like references or like physical
29:20 take-home test test I I both so exactly
29:23 both uh if you're like a series a
29:25 startup it's probably like finding the
29:27 best and hiring convincing the best
29:29 talent to join your company is generally
29:30 going to be really hard I assume most
29:32 people have probably never heard of most
29:33 series a startups um and in that case
29:36 yeah the best folks you're going to be
29:37 able to hire are going to be like
29:38 referrals warm intros that sort of thing
29:40 so getting some sort of back Channel
29:42 information from the person that is
29:44 hopefully connecting you is is I think
29:45 like actually the first step of the
29:46 interview process um I would argue the
29:48 second should be some sort of case study
29:50 or some sort of take-home test um you
29:53 could do the case study live in the
29:54 conversation with them just to see how
29:55 they think about things but I would not
29:57 be opposed or I don't think it would be
29:59 a bad idea to go from warm conversation
30:02 get to know you kind of selling you on
30:03 the opportunity in the business straight
30:04 to a take-home test so the first call
30:06 should probably be about selling uh and
30:09 then the second and hopefully you're
30:10 selling based on the referral that you
30:11 got just got the very strong referral
30:13 you just got and then the second stage
30:14 would be okay let's assess okay let's
30:16 assess take-home assignment what do you
30:20 like to give as a take-home assignment
30:21 and what advice would you give for me
30:23 yes uh make it as real as possible and
30:25 make it quantitative uh and actually
30:27 maybe the third thing I'd say is
30:29 understand what good looks like before
30:30 you send the test out so a great example
30:33 is I like to just take a Salesforce dump
30:36 or a dump of data and be like hey what's
30:37 the best campaign what worked best here
30:40 um what were the best leads whatever it
30:41 might be and if it's real world data
30:44 it's going to be really messy like
30:45 there's going to be a bunch of Tricks
30:46 they have to catch there's going to be a
30:47 bunch of mistakes they have to figure
30:49 out duplicate leads or um the dates
30:51 don't align or missing data whatever it
30:53 might be so I think like that itself is
30:55 like can they work with real world data
30:56 is I think a really interesting question
30:57 to answer then the other is like how
30:59 much of this did they think through what
31:00 was their thought process like how
31:02 quickly were they able to adapt how
31:03 quickly were they able even to like do
31:05 the tests did they ask questions about
31:07 it all of those I think are signals
31:08 before you even see the output uh to
31:11 understand like are they at least
31:11 thinking about things in the right way
31:13 but the real world is messy they should
31:15 they should work with real world data so
31:16 they work with real world data they come
31:18 back and you're impressed what happens
31:21 next do we go straight to an offer do we
31:23 have a hiring panel do we do any other
31:24 steps they should they should at least
31:26 meet some other members of the team like
31:28 ideally you want this to be a mutual fit
31:30 like both culturally but also they want
31:31 to work with the people on the team the
31:33 team wants to work with them so I think
31:34 there's still value in doing a panel but
31:36 it's almost more um culture fit and more
31:40 team fit more than anything but ideally
31:43 you are scoping that test that you send
31:44 them so you know if they can do the job
31:46 when you've got growth Talent wrong what
31:49 did you not see that you wish you had
31:51 seen when you hi them it's generally
31:54 been hiring a generalist that ultimately
31:58 hey this isn't for me I don't want to do
31:59 this after all um I've never done this
32:01 problem or I've never done this job
32:02 before now that I'm doing it I really
32:04 don't enjoy it I want to go and get back
32:06 into investing or I want to do like
32:08 XYZ honestly I I think it's difficult to
32:11 assess for that uh and you just have to
32:13 acknowledge that like no one even the
32:14 best person hiring is still going to
32:16 make a lot of mistakes and that's that's
32:17 okay I think probably what's most
32:19 important is just being as transparent
32:21 as possible both of in terms of the
32:22 opportunity and then in terms of how
32:23 you're being going to be assessed and
32:25 also just being transparent like if this
32:27 doesn't work that's F like we've all had
32:28 jobs before we're all going to have jobs
32:30 again like that's that's okay do you
32:32 agree that people are destined SL much
32:35 better suited for certain phases of a
32:39 company life I wouldn't say they were
32:41 destined but I think you can be suited
32:43 for it like if you spend the last five
32:46 jobs and 15 years in a certain stage of
32:48 company I think it's going to be
32:50 difficult to adapt your way of operating
32:51 way of thinking to a different stage of
32:53 company um having said that I mean I
32:55 don't think anyone in particular is like
32:57 destined for a certain stage I think
32:58 like if you're very very hungry and
33:01 you're a selfstarter and you love just
33:03 doing a little bit of everything it's
33:04 like okay you're probably going to do
33:06 better at like a seed or series a Stage
33:08 Company than like a 100,000 person
33:10 wellestablished business uh but that
33:12 might change over time uh that might
33:14 change as they gain more more experience
33:16 or do different things you need to
33:17 invest in your people right I mean
33:19 that's helps them to grow and develop
33:21 with scale and you said to me before
33:23 management is a skill that needs to be
33:25 invested in do you think it's currently
33:28 invested in by most companies by most
33:31 companies no I I don't I think that like
33:34 philosophically they're like yes we want
33:35 to invest in people's Learning and
33:37 Development I think and I think they
33:38 even believe that that's true but like
33:40 actually doing it is really difficult
33:42 like it takes time it takes resources
33:44 you have to be intentional it has to be
33:45 topped down um go how how how do you do
33:48 it then help me I I think Sam SAR
33:51 actually had like a great model for this
33:52 the CEO and founder like he was very big
33:54 on learning and very big on reading and
33:56 I remember like one time
33:58 uh he was hanging out in like the
33:59 cafeteria and like what do you like to
34:00 do in your free time like I like to read
34:01 books uh and so as a result you saw how
34:04 that permeated throughout the culture so
34:05 I remember they launched something when
34:07 I was there called leadership principles
34:09 and everyone that was a leader within
34:10 the company got a big box shipped to
34:13 their home and it had literally like 15
34:15 books in it and they're all business
34:17 books and the expectation was that you
34:19 read one of these books every single
34:20 month and then rejoin a conversation
34:22 with other other peers and this is all
34:25 organized to discuss like one of the
34:27 principles in these books that samsara
34:29 wanted you to embody uh and then there
34:31 was like an element where you actually
34:32 then have to go put it into practice and
34:33 demonstrate you're putting into practice
34:35 like to me that is a great example being
34:36 very intentional about hey we value
34:39 Learning and Development here and we're
34:40 going to add a structure and we're going
34:42 to add accountability to it to make sure
34:44 that you are not just learning things
34:45 but putting it into practice that took
34:47 time that took a lot of investment uh
34:49 that took really it coming from the
34:50 founders um and I think like a lot of
34:54 companies maybe sometimes get caught up
34:55 in the day-to-day and they say hey we
34:57 value learning but go figure it out on
34:58 your own rather than coming up with like
35:00 a holistic program or holistic mechanism
35:02 like that how do you think about the
35:04 effectiveness of that I read Howard
35:05 schulz's book on Starbucks and
35:08 Leadership Lessons From Starbucks I run
35:10 a Media company in a world of
35:12 Tik Tok and deep seek I mean it's
35:14 completely different to the point on
35:16 playbooks the leadership principles
35:19 probably don't align most leadership
35:22 books written 20 years ago do not take
35:24 account for a postco millennial
35:26 generation how do we think think about
35:28 applicability and actually the lessons
35:30 we teach being wrong I don't think
35:31 business has actually changed that much
35:33 tactically sure maybe it has the
35:35 channels Tik Tok didn't exist 30 years
35:37 ago but I don't actually think business
35:39 and being a good manager in particular
35:40 has changed that much or the skills
35:42 necessary to be a good manager have
35:43 Chang that much a great example is like
35:45 the book The Goal like I I'm a big fan
35:47 of that book I have some recency bias
35:49 because we reread it recently but one of
35:51 the key Concepts in that is like the
35:52 theory of constraints and now this this
35:54 doesn't have anything to do with
35:55 management but the concept of the theory
35:57 of constraints existed 40 years ago and
35:59 it still exists today like every team is
36:01 operating with some bottleneck in their
36:03 process and being able to identify that
36:05 bottleneck and remove it is a very
36:07 valuable skill for anyone but it's
36:08 especially valuable for a manager uh
36:10 that's how you get the most out of
36:11 people that's one of the core jobs of
36:12 being a manager so that's just one
36:14 example like I would argue that most
36:16 business books as long as you're vetting
36:18 them well and as long as like you are
36:20 being intentional on what is the
36:21 principle you want to pull out of this
36:23 and have your team put into practice
36:24 there's probably something that can be
36:25 learned from from almost any book regard
36:27 L of how old it is what is the biggest
36:29 bottleneck in your role today at ramp
36:32 that if removed would be the biggest
36:34 game changer the the biggest bottleneck
36:36 is probably the velocity of
36:37 experimentation like there are so many
36:39 opportunities we don't have enough time
36:41 we don't have enough resources that's
36:42 probably like at the if you zoom out far
36:44 enough that's probably the constraint
36:45 for most businesses okay what would what
36:48 would you like to do but because of time
36:50 or lack of resources you're not able to
36:52 do number one there are a lot of things
36:55 I would just like to get off the ground
36:56 faster when you're on external parties
36:59 either like external vendors or external
37:01 lead sources or other legal teams to
37:03 review your terms and services things
37:05 just go much more much more slowly um I
37:08 wish that there was a way to speed all
37:10 of that up um I think that ramp actually
37:12 does a really good job of saying like
37:14 let's optimize for Speed just accept
37:15 those legal terms maybe not necessarily
37:17 but let's let's figure out a way to move
37:19 as quickly as possible and just like
37:21 launch it and test it and then we can
37:22 figure out the details if we're going to
37:24 actually scale it up we spoke about
37:25 investing in in people becoming manages
37:27 and developing I spoke to so many of
37:30 your former you colleagues and they all
37:33 said one of your biggest strengths is
37:34 the willingness to roll up your seves
37:36 and do the work yourself my question to
37:38 you is ironically how willing should a
37:41 manager be to do IC work versus that
37:44 actually just being a plaster to not
37:47 having great ic's it's a great question
37:50 I think a good a good leader I would't
37:52 even say just manager I think a good
37:53 leader needs to know how to do everyone
37:55 on their team's job but poorly and I
37:58 think the poorly part is important they
37:59 should know how to do the job so they
38:01 can step in if they need to or so that
38:02 they're dangerous enough to to ask the
38:04 right questions but if they know how to
38:06 do the job better than the people
38:07 they've hired then they didn't hire the
38:08 right people and it's going to lead to
38:09 micromanagement it's going to lead to to
38:11 your point filling in gaps that are not
38:13 the most effective use of their time um
38:16 so being able to number one either learn
38:18 enough of what your team does so that
38:19 you can do it poorly or number two hire
38:21 people that can do what you need them to
38:23 do better than you know uh I think is is
38:25 kind of the key okay so you hired me
38:28 I've joined your team I'm I'm a junior
38:29 you you hired very Junior with this one
38:31 George sorry um but uh what is that
38:35 onboarding look like what do you expect
38:37 from your new growth hires in the first
38:40 30 days first 30 days it's learn the
38:43 business it's learn the team it's learn
38:45 like your your area of of domain I I
38:48 would expect in the first three days you
38:49 know how to do your job uh and that's
38:50 pretty much it and you understand how
38:51 the company operates and how the company
38:53 makes money like I think a strong
38:54 Foundation is actually understanding how
38:56 the business works what do I do as what
38:58 do I do as a leader to give you those
39:00 best 30 days do I just say just Shadow
39:02 the out of me do I say go sit in
39:04 support do where do you go great
39:06 question I think this again goes back to
39:08 to the investing in learning I think the
39:09 leader needs to be incredibly detailed
39:11 in what those first 30 60 90-day plans
39:14 look like who you're meeting what you're
39:15 doing with your time like actually this
39:17 was something I learned from from
39:18 samsara as well like I remember my first
39:20 30 days were written out in excruciating
39:22 detail I think the first two weeks were
39:24 completely scheduled out like to the
39:25 minute for me uh and as a result it was
39:28 also very clear to understand if someone
39:29 was learning and picking things up and
39:31 it was very easy to compare folks if
39:32 they were given the same like structured
39:34 onboarding um so I firmly believe like
39:37 the first 30 days is like you learn the
39:38 business and you learn the job now
39:40 Beyond those 30 days like you should be
39:41 able to start showing like some sort of
39:43 Step change impact start having some
39:45 ideas start making things your own and
39:47 then by 90 days you should be starting
39:48 to actually like show the fruits of
39:50 those new ideas and the fruits of like
39:52 that new experience that you're bringing
39:53 or New Perspective that you're bringing
39:54 should you always go for early wins just
39:57 to get points on the board if you see
39:59 the opportunity yes and there should be
40:01 the opportunity um if you were you you
40:03 were ideally hired because you have some
40:05 skill set or some um unique perspective
40:08 uh or there's some Gap in the business
40:10 that you're feeling so there should be
40:11 some early points on the board you can
40:12 put up having said that if someone is
40:14 not able to put up points on the board
40:16 you should understand why uh and if they
40:18 were set up for success and if they were
40:20 not you should move that as quickly as
40:21 possible but there are some roles and
40:23 there are some problems that do just
40:25 take time to solve and that's where the
40:27 okay how can we scope this down and run
40:28 an experiment to validate uh are we at
40:31 least on the right track is important
40:33 but there are some aspects or some times
40:34 where it is difficult to put early
40:36 points on the board George how fast do
40:38 you know if someone you hir isn't good
40:41 enough I think you generally have
40:43 inklings in the first couple weeks but
40:44 you should give some people the benefit
40:45 of the doubt like you should run some
40:47 sort of Str and I guess this is why
40:48 structured onboarding is so important um
40:51 and this is why making sure people ship
40:52 things in their first week is so
40:53 important you actually can assess them
40:55 based on facts versus just Vibes or
40:57 feelings
40:58 um but you generally have some sort of
41:00 Viber feeling in the first week or two
41:01 do you know what I find really hard what
41:03 it's hard to get rid of someone after a
41:05 week it's like you haven't given it
41:06 enough time do you know what I mean it
41:08 looks like you're just cutting it
41:09 prematurely short so you know but you
41:11 keep them for three months because
41:13 you're like well at least now it looks
41:14 like I can say I've given them Fair try
41:17 yeah I think I think that's true I I
41:19 don't actually think that's wrong so
41:20 good what are the most common reasons
41:23 growth hires don't work I think like the
41:25 number one reason is it was it was on
41:28 the manager it was a mish hire uh they
41:30 didn't scope the role effectively or the
41:31 role changed the needs of the business
41:33 changed and they hired the wrong person
41:35 out the wrong profile I think generally
41:37 managers try to come up with like a
41:39 laundry list of skills that they want to
41:41 hire for and they try to find someone
41:42 that checks most of those boxes but
41:45 that's almost the easy way to do things
41:46 I think the much harder way is to get
41:48 really really C crystal clear on what is
41:50 the one thing that we need like what is
41:52 the one skill or trait or piece of
41:54 experience one problem we need to solve
41:56 then hiring someone that you have high
41:57 confidence can do that or fill that Gap
41:59 versus check a bunch of boxes so not
42:01 getting clear on that I think is the
42:02 number one reason why you would Mish
42:03 hire and that's mostly on the manager if
42:05 I'm being honest can ask how does AI
42:08 change the role of growth you know we
42:10 see some pretty futuristic things in
42:12 terms of here's my budget go spend it
42:14 across five channels and it will do it
42:17 in the optimal fashion for you is that
42:20 the future of growth and how do you
42:21 think about how your role changes with
42:23 an increasing prominence of AI it's in
42:26 that example I think it's hard to find
42:27 Alpha if you're just having a ai go find
42:29 the opportunities for you having said
42:31 that I think AI is tremendously changing
42:33 like the role of growth like it's is it
42:35 hard to find Alpha in that respect
42:37 because they'd have all of your
42:38 connected data history and so they could
42:40 look at all prior performance conversion
42:42 on every channel compare it like for
42:45 like Benchmark it against standards and
42:47 then do what's best for you not a
42:50 anonymized data set this is what you
42:52 should do it could be incredibly
42:54 personalized it could be personalized
42:56 but almost by definition that's just
42:57 going to give you incremental gains on
42:59 things you're already doing and things
43:00 that has data on uh I think like to get
43:03 real Advantage like how maybe we'll get
43:05 to this point at some point soon but I I
43:07 think it's hard have the AI go talk to
43:09 like 50 people in the industry and like
43:11 synthesize what they're doing and apply
43:12 what they think is going to be relevant
43:14 to your business model and your ICP and
43:16 like go apply that that sounds really
43:18 difficult to do at least at the state of
43:19 AI today but but maybe not in a year who
43:22 knows tooling wise does it change much
43:25 uh using AI yeah
43:27 yes 100% uh I remember like a few years
43:30 ago I was such a strong Advocate that
43:32 like everyone on the growth team needs
43:34 to be technical they need to like learn
43:35 SQL they need to Learn Python they need
43:38 to learn how to work with like a growth
43:39 engineering team like very very closely
43:41 I think like AI upended that like you
43:43 don't need to be nearly as technical as
43:45 as you used to like AI can help you
43:46 write the code it can tell you other
43:48 ways of doing things like it's even just
43:50 like in that element of like who I used
43:52 to like have a bias for hiring it's
43:54 totally changed in my mind you growth
43:56 people bluntly are either
43:58 performance-driven scientific or they
44:01 are creative driven artistic does a
44:04 world of AI make one likely to be more
44:08 successful than the other I think it is
44:10 a good it's a good co-pilot for lack of
44:13 a better term for for both of those
44:14 roles I think that like on the creative
44:16 side it can help you think about new
44:17 ideas and help you brainstorm on the
44:19 performance side it can help you analyze
44:21 the data and maybe be more efficient or
44:23 move more quickly you've got to choose
44:25 which one it helps more
44:28 so here here's my bias I am not super
44:30 creative at the end of the day I'm not a
44:31 creative-minded person so I think it
44:33 helps me a lot more I think you are
44:36 don't discredit yourself so you think it
44:38 helps creative people more I I think so
44:40 like or maybe it helps uncreative people
44:42 more is maybe what i' would say um the
44:45 number of times that I've had chat GPT
44:47 just be like or any type of AI really
44:49 I'm like how do I make this better how
44:51 do I add a joke in here what are some
44:53 other like visuals I can do what's a
44:54 good analogy for XYZ like I wouldn't
44:57 have been able to do that on my own I
44:58 would have had to go to someone on
44:59 product marketing or content be like hey
45:01 I need some help can I ask you how do
45:02 you advise Founders competing incredibly
45:05 intense competitive markets you know
45:08 ramp is in a very competitive market
45:10 across a number of different products
45:13 what's your biggest advice on competing
45:16 in very proliferated markets I think
45:19 it's like most things you just have to
45:21 have a better product and you have to
45:22 have some plan for distribution I think
45:24 that having a better product probably
45:26 solves like 80 90% of the battle uh but
45:29 you also need to think about how you're
45:30 going to distribute that that product
45:32 and like what your unfair Advantage is
45:33 in distribution and hopefully that's
45:35 where the growth team comes in and where
45:36 they can provide value but I do think
45:38 it's mostly on the product do you agree
45:39 with the suggestion build it and they
45:41 will come absolutely not no like I'm I'm
45:44 on I'm a marketer at the end of the day
45:45 and I'm a growth person like that is
45:47 that is antithetical to everything we
45:48 stand for there are some products sure
45:50 where like yes they it is so amazing and
45:53 the word of mouth was so strong that
45:54 like you could they did build it and
45:56 people did come I think that is so out
45:58 of the ordinary and leaves your fate so
46:01 much to chance that that's not a good
46:03 approach listen George I've Loved this
46:05 you are unbelievably concise with
46:08 answers I call it kind of word economy
46:10 which is like value per word and like
46:13 most people are fluffy and don't say
46:15 much but it takes a long time you're
46:17 unbelievably concise with very dense
46:19 value it it makes my life a joy to be
46:21 quite honest um I want to do a quick
46:23 fire with you so I say a short statement
46:25 you give me your immediate thoughts does
46:26 that sound okay sure let's do it okay so
46:29 number one what's the most common
46:31 expensive deadly irreversible mistake
46:34 you see Founders make hiring for
46:37 experience because they don't know any
46:40 better yeah like we don't know how to do
46:42 this we'll hire someone that
46:44 does and it should be we should hire
46:46 someone who doesn't it should be like we
46:48 should actually assess for hey how do we
46:51 determine what good looks like and how
46:52 do we test them for it and like could
46:54 someone figure this out could someone on
46:55 the team figure this out rather hiring
46:57 you think most Founders know what good
46:59 looks like that's the hard thing they
47:00 they got to figure it out like I think
47:02 that's where the learning comes in like
47:03 go talk to other go determine let's I'm
47:05 going to make this up like you're trying
47:06 to figure out how to do cold calling
47:08 maybe this will work for us maybe it
47:10 won't go identify 10 companies that do
47:12 cold calling really well go talk to the
47:14 people that built that infrastructure
47:15 and built that system etc etc do you
47:18 know what dude do you know why I started
47:19 these vertical shows because one of my
47:22 companies hired someone for growth that
47:24 was just obviously terrible to me and I
47:27 said this is clearly a mistake and they
47:28 said I don't know what good growth looks
47:30 like and I was like well if I spoke to
47:34 the 10 that I know and publish them
47:37 you'd have an idea that George is top
47:39 tier and that's what I should look for
47:42 in someone that I work with and so this
47:44 was meant to be a benchmarking of world
47:47 class quality in each vertical and
47:49 that's why we started them I love it I
47:51 think that's why I'm such a fan was very
47:53 kind um tell me what's the most
47:56 underappreciated a growth Channel today
47:59 uh it is probably honestly something on
48:00 the influencer side of things uh I don't
48:02 think most people think of it as a
48:04 growth Channel but influencers
48:05 absolutely is at least on the B2B side
48:07 can you do influencer at scale that's
48:09 the hard thing that is the fun problem
48:12 to solve I I think you can I think you
48:13 treat it almost as a outbound funnel
48:15 it's like go make a list of the 10,000
48:19 micro influencers or folks that could
48:20 become micro influencers and go do some
48:22 scaled outbound to them figure out a
48:24 process that works like you you could
48:25 solve almost any problem in that way uh
48:27 but short answer is yes what's the most
48:29 polluted Channel today or overrated paid
48:33 search I think everyone goes to paid
48:35 search because they have to but I think
48:37 like in the very very early days you
48:38 have nothing that works sure it's like
48:40 table Stakes but it saturates quickly
48:43 you're paying a tax to Google I think
48:46 paid search is like relatively
48:47 uninspiring of a growth channel that
48:49 will scale long term what growth tactic
48:51 have you done that with the benefit of
48:53 hindsight you're like oh I wish I hadn't
48:56 on
48:58 that oh man um I think like very very
49:01 early on it was event spons like at very
49:03 small companies it's event sponsorships
49:05 like go exhibit at dreamforce or
49:07 something like that like it feels like
49:08 something you have to do because
49:09 everyone else does it it's like just a
49:10 waste of money you're in a sea of other
49:12 companies and no one's paying attention
49:14 to you like your that money would be
49:16 better spent figuring out some gorilla
49:17 tactic to stand out or honestly probably
49:20 be better spent on paid outs we do a lot
49:21 on events I think we're we're at a very
49:23 different scale though um and we do
49:25 things a little bit differently it's not
49:26 just like go exhibit at event at an
49:28 event it's also like get a speaker slot
49:30 do some out of out of home advertising
49:31 around it make sure we do like outbound
49:33 emails immediately after it's it's more
49:35 than just like let's get a booth in a
49:37 corner of the conference hall because it
49:39 was all we could afford and let's hope
49:41 people come to us it's one of those ones
49:43 where it's like if you're in you need to
49:44 go all in and have the billboard outside
49:46 the conference the key cards for the
49:48 hotels as well the speaker slot and the
49:51 booth not just like a small Booth
49:54 exactly and and this and that's why I
49:55 think it changes in depending on size
49:57 like when you're a series a company like
49:59 most events depending upon your go to
50:00 Mark motion most events are not going to
50:02 be high Roi what growth channel did you
50:05 not take advantage of that with the
50:07 benefit of hindsight you're like H
50:09 George that was in plain sight it's a
50:13 good question I I think like direct
50:15 mailing gifting was one that we could
50:16 have actually even attempted even
50:18 earlier uh when we were at samsara in my
50:20 opinion I think that there's probably
50:22 unfair advantages to be gained in like
50:24 display advertising like display
50:26 advertising is so inexpensive right now
50:28 but it's not measured well and most
50:29 people think of it in terms of like
50:31 direct response like if someone no one
50:32 clicks these ads so like the response in
50:34 the ROI is not very good but the reality
50:37 is like figuring out how to serve imp
50:39 the right number of Impressions with the
50:40 right message into an account or into a
50:42 contact actually does have a halo effect
50:44 so there's definitely an advantage there
50:45 but solving and solving for that
50:47 Advantage is probably dependent on how
50:49 you measure it uh so I I would say
50:50 display advertising is an opportunity
50:53 what have you changed your mind on in
50:55 the last 12 months I I would say it's
50:57 very much brand advertising um I think
50:59 this is something that or brand
51:00 investment in general I think this is
51:03 something that gong did really really
51:04 well uh they were willing to invest in
51:07 things that elevated the brand even if
51:09 it was completely unmeasurable even if
51:11 it was something that like we would
51:12 never be able to measure um but you saw
51:16 that in the fact that like you saw that
51:17 in the inbound numbers um and how big of
51:20 like a channel like inbound was for gong
51:21 because they invested in their brand so
51:23 much and I think that like it doesn't
51:25 make sense when you're very early
51:26 startup and you don't have that type of
51:27 resource but once you reach a certain
51:29 size or scale like if you're not
51:32 investing in brand as part of like your
51:33 long-term horizon or long-term bucket of
51:35 bets then you're going to screw yourself
51:37 over in the future final one for you
51:39 George what growth strategy have you
51:42 been most impressed by in the last 12 24
51:46 recent times but which company were you
51:49 like that was smart it's it's usually
51:52 the non-traditional stuff so like cold
51:54 calling I think a lot of people thought
51:55 cold calling is said and like cold
51:57 calling is like a tremendous channel for
51:59 a lot of companies because it is hard to
52:01 do door too I think is something that's
52:02 really interesting and I've been
52:03 thinking about this for a while now like
52:05 the medical devices space is interesting
52:08 because a lot of their sales and a lot
52:09 of their their marketing or sales is
52:11 door too um and I think there's actually
52:14 now that folks are moving more back into
52:15 the office and working from offices
52:16 again I think there's probably an unfair
52:18 advantage to having a true field sales
52:19 team that goes door to door and brings
52:21 like a cake with them or whatever it
52:22 might be uh so that's that's something
52:24 I've been thinking about that any door
52:26 daor salesman that want to bring a cake
52:28 I love cake all in George as I said like
52:32 I love it when you get concision with
52:35 quality it's very hard to do but you've
52:38 been fantastic so thank you so much for
52:40 this I've loved having you and you've
52:42 been amazing no thank you I'm honestly
52:44 great questions I love the conversation
52:46 hopefully I wasn't too concise uh but no
52:48 it's been a fun time
Chrome Extension