171: The Founder Flywheel: Retention First, Revenue Second | Richard White

0:00

fathom is one of the most successful SAS products in recentyears i know it's one of my favorite and I bet you've seen

0:06

their notetaker in at least one meeting every week but howdid a small team with

0:12

zero salespeople build one of the fastest growing highestrated tools in

0:17

SAS today by delaying sales and doubling down on productwelcome to the Learn It All Podcast the show for today's leaders

0:24

who want to get ahead and stay ahead because we believegreat leaders aren't born or made they are always in the

0:30

making i'm your host Damon Lebby two-time best-sellingauthor and CEO of Learn It a live learning platform that

0:37

has helped to upskill over 2 million people over the pastthree decades today I'm joined by Richard White a three-time

0:44

founder who has turned frustration with meeting notes intoan AI engine that replaces note takingaking with clarity

0:51

context and fewer follow-up meetings at the end of the day alot of our products were just forums so there's you know they're they're webapps right now with

0:58

AI like you have a different totally different way you gotto build it it's like we're building backwards from what's technically possiblein our

1:04

conversation today I'm going to ask Richard to share histhoughts on why solving for retention before revenue was

1:09

Fathom's secret to early success how moving fast evenwithout all the answers

1:14

can be your biggest startup advantage the hidden risk offollowing expert advice that's stuck in the past why

1:20

onboarding is the real first impression and how it fuelsword of- mouth growth

1:25

and what it really takes to ship AI features that are morethan just good enough when did you realize that you

1:32

weren't just going to be a builder or a coder but somebodywho's going to keep founding things uh I mean I think I had

1:38

my I founded a bunch of things when I was like I would sayfounded but like I built a lot of stuff when I was like in

1:43

school like actually in high school like made websites backin the late 90s and sold fake Oakleys to like students at

1:50

high school and I think I actually knew when I tried to getmy first real job and I was realized oh I'm kind of a

1:56

terrible employee um that I realized maybe I need to stickto this like start startup game um and so yeah and so I

2:03

think I had two stints where I was an employee right aftercollege and then ever since then I' just been founding companies what was itabout being an

2:09

employee that you just couldn't stand i you know I if Ididn't understand why we were doing something like I like you

2:14

know I probably wouldn't do it right it was you know I wouldbe really good if you gave me all the context of why we

2:20

were building something and then I'd build something prettygood but I think I was just like you know if I didn't get that and mostmanagers didn't want to

2:26

spend the time to do that they're like just go build thisthing shut up and build it right kind of thing our clients are asking forthat's all the why you

2:32

need sort of thing and that wasn't sufficient for me itturned up so now that you know that how have you been a

2:39

better leader by being able to articulate that why better imean that's the number one thing I feel like I

2:45

supposed to do for the company right is share context andmake sure everyone knows why we're doing the things you

2:50

almost like you know from the five W's like what are the allthe way down like why are we building this feature why is

2:56

that feature matter why does that metric matter etc um youknow because I'm

3:01

obsessed with like how do we get the most out of having asmall team and part of that's having really senior folks

3:06

that are you know really autonomous and really you know shipcode fast in case of engineers and they do their best work

3:13

when they like me like really understand what impact theirwork should have so I spent a lot of time on that super

3:20

critical because especially these days people want to feellike they have purpose and they have belonging and that

3:26

what they're doing matters that they're not just they're notjust paycheck players so let's just take it back a

3:32

little bit so you were there for the early YC YC scene whatdid being around

3:38

that environment and being in the mix teach you about or didyou learn from building timing and just maybe even just

3:45

staying in the game yeah it was interesting so I I grew upin North Carolina i tried to do a bunch of like startup stuff while I was incollege

3:51

while I was a teenager and I think at some point I reallyreally struggled to find people that also wanted to work on startup thingsactually most people

3:57

around me were super happy with the like you know the stablepaycheck sort of situation and I got very fortunate that

4:03

through a cold email I ended up working with company in thefirst batch of Y cominator which at the time was based in Boston i then likefor working with them

4:10

flying up to Boston working at the office with like thefounders I worked with the people that went on to found Twitch but like thefounders of Reddit

4:16

were in the same room as me and honestly that in some wayslike saved my life or saved my career in some ways because I

4:22

think I probably would have given up on the startup path butthen I kind of found my tribe of people and these folks

4:29

you know they I didn't have to convince them to take risk ofdoing a startup they were kind of like me that doing a startup was like thedefault assumption

4:34

like I couldn't imagine doing anything other than buildingsomething um and so I think it was really cool and I mean it

4:39

was amazing community of folks that I'm moving out to SanFrancisco with you about 20 years ago and it was kind of

4:45

during the combust wall right um and kind of

4:50

reminds me actually of the last couple years here right likepost 2022 where you had you know all these companies

4:56

kind of go bust and everyone's like oh you know tech boom'sover sort of thing and that's what I feel like all the real

5:01

builders kind of get to work and so that was what was greatabout that that cohort of people I came here with about 20 years ago is theyweren't driven by

5:08

getting rich they were 100% driven by just like this thingshould exist in the world and I'm going to build it and uh

5:14

you know you need a fair amount of hubris to do that butlike it was just this really genuine like I just love to build stuff kind ofmentality amongst

5:21

everyone in that group that was really infectious so fromthat opportunity to to be around

5:26

those types of individuals and you know being part of it hasit helped you in

5:31

return kind of give back to others in uh others who want tobe founders and have conversations with them yeah I mean I

5:38

think a lot of you know I I try to go like back as much aspossible right with the day job right because a lot of my

5:44

the best things have happened like I said cold emailingother founders uh you

5:49

know joining their companies them helping me out uh and solike that network that like founder to founder network is just so so powerful ithink

5:56

it's also I'm pretty competitive guy so it's pretty goodbecause you're like kind of friend of me it's like I'm

6:02

competing with all my friends too in a fun way right likeand so I think it really pushes each other to be better

6:08

and I'll say I learned a lot from watching these guys righti think a lot of these folks had a level of self-confidence that I originallyonly

6:14

could aspire to when it comes to like trusting their owninstincts a lot of the most successful people I know in

6:20

their mid20s had such supreme confidence in their owninstincts and decision-m which was mindboggling to me at the time

6:25

but it produced really good outcomes for them and so I thinkI've learned a lot just from watching how all these pretty

6:31

successful people operated even before they had the successto really validate that what they were doing if that makes

6:36

sense my background I was a baseball player you know andthat's what I always thought I wanted to be but I even kind of lacked someconfidence there a little

6:44

bit uh what advice do you have for somebody who maybe hassome good ideas

6:49

but maybe they lack the confidence to take it to the nextlevel i mean I know one trap I think I fell into early in my

6:55

career is kind of confusing experience for wisdom um uh orfor intelligence

7:02

really you know and it's you know when you're in your mid20sor in your early 30s and you meet folks that have been in the game for 10 moreyears they sound

7:08

really smart you know and you know you have this almost wantto like oh wow I should not only do what they say but I

7:15

should always seek out someone that's been there done thatright but you tend to find that like a lot of that stuff tends to sometimes belike that that

7:22

advice tends to over represent the way things used to beright and so I think the best folks you know there's always

7:28

this thing like don't ask for advice ask for stories andthen integrate those stories into your own kind of experience

7:33

and figure out what they mean right don't just take them asgospel of it's always done this way right because all the best folks you knowthey would get

7:40

they would hear other people's stories but then theypersonalize it to them and they make their own decision and I think what I dida lot of was was the opposite

7:46

where I just like spent a lot of time every decision I madeall I need to go talk to three experts on this and the

7:52

best friends I the most successful friends I had didn't dothat right they did some of that but they were also like you know I'm going tojust move forward

7:58

because it seems like the right thing to do um and I thinkit turns out in the startup game speed is almost the most

8:04

important resource right time is the most important resourceeven more than like there's very few decisions that you

8:09

have to get exactly right right just getting themdirectionally correct and moving quickly is by far like game

8:14

optimal and I totally agree with that you you can't sitaround think about perfection and have your seven-year

8:19

business plan you'll just get passed up the other thingRichard I like what you said is you know you can ask for a

8:25

couple people for advice but at some point you got to youknow these are just concepts or people's opinions i had this

8:32

uh one friend of mine I think he's a legendary leanconstruction guy Paulo Napalito who said even stop reading

8:38

books after a while you know and and and start coming upwith your own point of view and perspective and find an empty

8:45

space and do things different especially if you want to getin and and create something whether it's disruptive or even you know 10x betterthan what's

8:52

currently out there i think that's really true too i meanthe other reason why I like being in kind of these founder cohorts which is whywe actually

8:58

did YC again with Fathom was because I think there's so muchyou can learn from other folks that are just like trying

9:04

new things from first principles by the time there's a bookthat says here's how to do this thing marketing wise it's usually probably outof date right like

9:11

because that's how you know quickly the world moves you'vegot to find you kind of have to ziggs type of thing right

9:17

yeah and so this Fathom is your third startup really rightkiko and then uh if

9:22

I said that right voice user voice uh in my research uhabout you what I really like is how you call uh user voice your

9:30

your finishing school um so what were some of the mostimportant lessons you learned there that you then pivoted and

9:37

used getting into Fathom i mean I do think so I I ran thiscompany voice for about 12 years uh it's kind of a

9:43

platform for product feedback you think like Reddit forproduct feedback and over those 12 years we kind of did

9:49

everything because it was like a tough market a good butmaybe not great product sort of thing like when I was

9:54

there and so you had to execute really well and we had kepttrying everything to try to make this thing grow right and

10:00

so we did PLG before they called it PLG uh and then we werethe to the end of

10:05

its lifespan where I was there end of my tenure it was anenterprise business in

10:10

between executives I ran every team i ran the marketing teamfor a minute i run in the success team for a minute i run the sales team for aminute and so

10:16

it made me a very good like T-shaped entrepreneur right mybackground is product and engineering and design so

10:22

I'm like really deep in that and go back 15 years beforeuser voice i was really deep in that but I didn't know anything

10:28

about sales marketing you name it and that was really Ithink I think that really held us back because I was so

10:34

differential to folks i kept looking for people to come inand white knight those things for me because I was just like I

10:39

had in my head I can't figure these things out now afterdoing that for 12 years I'm not a great head of sales but

10:45

I know enough to like know okay here's a good V1 of how tobuild this and here's what a good first leader looks like

10:50

things like that so I think it makes you really you know ifyou can be kind of crossunctional you know enough about how

10:57

all the little things all those departments work you canmake quick decisions you can figure out who to hire

11:02

like you know it just allows you to move so much faster thanwhen you're you have to delegate all that stuff because you

11:07

just are in your head that like I don't know anything aboutthis like I can't do anything about it it turns out it's so much easier for ifyou're technical to

11:13

like learn some of these other disciplines than the reverseright right so if you're coming from a technical background you already have Ithink one

11:19

of the hardest things which like this deep technicalexperience trust yourself that like sales is rocket science yes

11:25

you're not going to do an A+ job but in the beginning youdon't need A+ job you need like a C plus job right um and same

11:31

thing goes for like a lot of these disciplines likemarketing and and customer success you don't need an A+ job you just need toget yourself out

11:38

there and and just kind of test a market and see what peoplethink now I've got several buddies who are you know

11:44

fractional uh CRO and one of their biggest frustrations tendto be working with

11:51

technical founders who really know nothing about sales umand they always

11:57

think that products can sell themselves um what advice orfeedback do you have

12:02

for technical founders who maybe have that perspective ithink there's a tendency for technical founders to

12:10

undervalue the human relationship right um they tend toundervalue things like customer support they tend to undervalue

12:16

things like sales um and again I think sometimes if you'rekind of a little if

12:21

you're like the maybe the classic kind of slightlyantisocial you know technologist

12:27

archetype you yourself undervalue these things right and soit's asking to do the thing that's like the hardest thing to do is justempathize with people that

12:33

are not like you right um and so I think you know my generalthought on a lot of

12:39

these things is go and see like go try it right and the samereason I think like most

12:45

companies always benefit from their first round of fundingmaybe not their 10th but their first round of funding like because you knowyou've got more

12:51

resources to do something so what I'd say to them is likeyou know understand that the way most people not everyone buys the same way andin fact there's

12:58

probably two or three major archetypes of how people buysome people hate talking to sales some people won't buy

13:03

without talking to sales um and so if you want to build abig business you're going to have to build them you have to

13:09

build a business that can cater to all the different wayspeople want to buy i think it's a great point so let's talk

13:15

about Fathom and I I mentioned to you before we went on airI'm a big Fathom fan uh our team's been using it at Learn

13:21

It for several years now what made you decide that Fathomwas needed i mean I

13:28

think it was right before the pandemic i was just on Zoommeetings and I was doing a lot of you know meetings with

13:34

customers and they're back to back and I'm talking tosomeone and I'm hurly typing out some like you know shorthand

13:40

notes and spending a lot of time stressing out that I'mmissing some key quote looking at those notes two weeks

13:45

later and saying gosh I don't remember which meeting thiswas and more importantly like try to share those notes with my team andrealizing how

13:51

much was lost in transmission right like I got to have anamazing conversation with a prospect or a customer and you

13:58

know they have really positive or negative reaction tosomething we're building two weeks later I put together the synthesis and it'slike three bullet

14:04

points on the page and the team just kind of shrugs theirshoulders and so it seemed obvious to me that like gosh this

14:09

is like such a bad experience from all angles notes are noone likes taking them they're not effective for reading

14:15

them i ran our sales team for a minute i mentioned it uservoice and I remember the number one thing I I struggle with

14:20

as a sales leader is like trying to get good notes out of myteam some people write too little some people wrote way too much actually buteven when people

14:27

wrote a lot I still found myself being like "Yes butwhat did that person actually say when they said they're going to buy beforethe end of the month

14:33

right?" Could I hear the tone in their voice um and sojust in general I kind of came around to this belief that like

14:39

gosh notes are this weird artifact that we don't really needanymore we have technology can do this for us right we

14:44

can we can record the entire meeting we can now transcribeit at pretty high quality and in 2020 we didn't yet have

14:50

LLMs that we do today and AI but I was pretty confident thatthat was coming um

14:56

based upon what I knew from my YC network about what washappening at OpenAI and companies like that and so we

15:01

really bet this whole company on you know everyone takingeveryone hates taking notes everyone would benefit from

15:06

having an AI in the meeting that's like recording and takingnotes for you and even though they doesn't exist to do

15:12

this today we think it'll exist in a few years and so we'regoing to start building the product in that direction

15:17

so when AI does show up we'll have built all the hard partsreliable video recording and transcription

15:22

crossplatform etc etc we'll get to kind of drop in AI like anew engine in an

15:27

old car sort of thing uh and that's fortunately exactly whathappened right we ran out of window that was closed and

15:33

it opened right as we were kind of jumping through it um andso it's been very gratifying to see that and I like

15:39

the story about how you brought on I think it was your oneof your first sales reps and you said "Look there's

15:44

nothing for you to sell here for a while i just want you tocome in and experience and kind of get a feel for

15:50

it." Talk talk to a little bit about how that workedbecause most people put revenue as key number one when you're

15:57

first getting off the ground especially after you raisemoney i tend to like to take the key things that make up a business right likeretention

16:03

acquisition monetization and I like to kind of attack eachone of those in order when starting a new company and so

16:09

when I looked at Fathom I thought well the hardest thinghere is and I think for most companies is retention it's

16:15

actually getting a product people use day in day out or weekin week week out so let's do that first let's really

16:20

focus on that first once we nail that great now we're gonnaactually focus on how do we get more users in right think about distributionchannels and stuff

16:26

like that and once we have all of that now we'll focus onmonetization because it's really hard to know what people will pay for beforeyou know who you can

16:33

acquire for how much and whether they're going to stickaround right um and I also think there's always this pressure

16:39

yeah to like get revenue immediately but I'm also a big fanof like I'd rather show no metrics to investor than a bad

16:45

metric and when you try to tackle all three of these thingsat once you're not likely going to get you know you're C

16:51

minus across the board right so try to take these things onestep at a time and then yeah I think you know the other

16:57

thing I did with Fathom is like I noticed from my experiencewith user voice the value of you know people

17:02

building relationships right and hearing from customersfirsthand and again I think it took 12 years working another

17:08

company for me as a technical founder to really believe inthat and now I believe in it so much that we did a lot of things in the earlydays of Fathom where

17:13

we would pay people to talk to us hey you just had yourfirst call with Fathom we'll pay you 25 bucks to jump on a

17:18

10-minute like Zoom call and tell us what you thought and atsome point I was like I can't continue to do this i don't

17:25

have the skill to do this gosh let's bring in someone whocan and so this woman Amanda who was my kind of top SMB

17:31

rep and manager at UserVoice I was like yeah you come inhere and basically I'm going to ask you to be a CSM because we

17:37

have nothing to sell for at least a year maybe two but Iwant you to get in you really get to know the product get to know our users getto understand the

17:43

problems they have and at some point we'll flip the flip theswitch and we'll start charging money for it and you'll

17:48

already be so ramped up because you'll know exactly how thisthing works and what people like what people pay for and

17:53

you'll help me figure out actually what we should chargemoney for um and so that was been a luxury we had because as

17:59

a second time founder I could raise enough money to hirepeople maybe at well ahead of when we needed them but it

18:04

worked out fantastic i think we actually had a team of threesalespeople quote unquote that weren't selling at all for

18:11

over a year before we started monetizing but when the marketshifted in 2022 we immediately had to start monetizing

18:16

because now the market's giving us this feedback we needrevenue and now I was very fortunate to have oh I've already have three rampedup sales people on my

18:23

team they haven't been selling anything and so we're justlike great it's time to go sell so worked out really well are

18:28

you still doing health checks with your customers now thatyou have uh what I mean by health checks are having

18:34

somebody like Amanda or somebody who's been a customer ofFathom for a while and saying "Where are you where you

18:40

finding value with us?" or does that help with yourroad map as far as features and benefits even yeah I mean

18:45

we still try to build out a lot of these kind of feedbackloops both in the product and through service um we invest

18:52

a lot in customer support i think that's a place wherecompanies underinvest in oddly it's not grant that expensive to

18:58

do great customer support um and it's a huge it's a huge Ithink differentiator in the market um we still have all the

19:05

the things like how's your first call go uh that goes backinto our support team now we don't pay you $25 anymore for

19:11

that feedback um but if anything you say anything negativelylike oh something

19:17

doesn't work and immediately gets escalated to our supportteam we do kind of NPS surveys uh throughout the life

19:23

cycle anything that comes up negative there goes back to oursupport team or our success team and so I yeah I I love

19:30

building these loops that kind of do a lot of like one tomany kind of engagement right because 19 out of 20

19:35

people are probably going to have a great experience one outof 20 are having a bad one great we want to like kind of be drilling test wellsall the

19:41

time to figure out where those folks are get them into somesort of support or success experience so we can kind of

19:46

remediate that and then you know a lot of learnings fromthat get fed back into the product road map you also have to be

19:52

aware that you can't please everybody all the time so notnot saying you should discard somebody's feedback

19:59

because all feedback is data but some people you just can'tmake happy so if you have 29 out of 30 happy may maybe

20:05

you know you could look at the case as far as growth hasgone for Fathom how

20:11

did it look differently from you know your first thousand uhusers to you know

20:17

the hundreds of thousands you probably have now i love thisquestion because I think this is another like key start of insight which islike everyone thinks

20:23

you have one channel that takes you from like the first userto the millionth user right and I actually think our

20:29

experience has been you know the first thousand users wasliterally me going to my network on LinkedIn and just harassing people intousing Pathom right

20:36

like hey will you try this out like I think someone Ioffered money like hey I'll pay you 50 bucks I think we also

20:42

used like interview panels like we went on some sites whereyou could hire people to do user testing and I just like hey will you you Iactually pay you

20:49

like you know you're a sales rep you're a CSM I'll pay you50 bucks 100 bucks to

20:54

try to use this product in front of me and then some of themkept using it right so first thousand very scrappy a

21:00

lot of LinkedIn messaging badgering my friends you knowhiring interview uh

21:06

folks and honestly I think out of the first thousand peoplewe got signed up

21:11

maybe about 800 we only got we got to about a group of about50 people that are using it day in day out and that was

21:16

actually like cool I was like great That's that's like winwe've like won the first stage right and so you know we

21:22

had 50 MAU when we technically launched um but I felt veryconfident because those folks had you know uh were were

21:29

kind of really high repeat users and then we got veryfortunate that Zoom was launching kind of like a marketplace

21:35

right around when we got started that's pretty lucky notmany big companies are launching marketplaces and even less

21:41

than generally let startups that have no users or 50 userslike we did basically participate in the launch um and so that

21:48

got us probably the first you know the next 10 to 50,000users sort of thing right and then from there it really just

21:55

spread kind of word of mouth or virally because you bringFathom into your meeting people have to kind of explain why it's there ohtaking notes for me

22:02

you know do you want the copy of these notes too i can sendyou the notes um and that was the thing that got me

22:07

excited about this business in the beginning uh one was Ithink there's a business model that really fit where it's like we could givethis product

22:13

away for free and if we can give it away for free then itcan spread virally because it's like the most naturally viral thing I can thinkof right in

22:19

front of people that aren't users constantly um and so thatwas our whole thing but viral models are tough because

22:25

the first you got to get to like 10,000 users before thatlike flywheel really kicks in and so that's where kind of

22:31

like getting involved with Zoom marketplace was so helpfuland having a a premium model is great but even even

22:37

with a premium model doesn't mean that people are going toactivate it and continue to use it because sometimes

22:42

they feel like they don't even have skin in the game so youhave to stay on them to show them the value to be able to uh

22:49

start leveraging it i think there's two things like one is Ithink everyone underestimates how good your product

22:54

needs to be to be like viral like it's not good enough forjust like it does the job it needs to like be like really

23:00

lovable and blow people's minds right like and so weintentionally I think you know we spent a year building this

23:05

before we ever launched it uh and the whole thing is likecan't launch something that just kind of works it needs to be really reallygood at day

23:12

one um and then we're fortunate with Fathom it's like youyour aha moment happens pretty quickly right after your

23:18

first meeting with Fathom when you get those notes and yousee how good they are you're like "Oh gosh I guess I never need to takenotes again right?" Um and

23:25

that's the big aha moment and happens you know sometimeswithin 30 minutes right of sign up right whenever your

23:31

next meeting is so I think we're very fortunate in thatregard that that's true but I think back to your point it's

23:36

like it's not enough just to make it free um I think free isa superpower i think especially if it's a really good

23:41

free plan i think a lot of people try to like cheat and belike "Oh we have a free plan but it has you know it's it's

23:49

clear you can't use it for more than a week or a month orsomething like that." And I think most consumers are pretty savvy aboutthat now and like no they're

23:56

honestly they're skeptical right because they've they'veseen this model a lot oh it's free yeah yeah sure but if I keep using it it'sgoing to become paid we're

24:03

not like that we were like "No no." In fact we hadno paid plans for the first two years right no no it's free be free

24:09

use as much as you want um and to me that's really importantbecause if you want word of mouth there's a huge

24:15

difference i saw this at UserVoice between me recommendingsomething that's like truly free and me recommending something to you thatcosts 50 bucks so

24:21

that's 50 bucks you're less likely to recommend to yourfriend because if they don't like it then you feel like you you

24:27

you know cost them 50 bucks if it's free it's like you gavethem a gift and also what I've seen psychologically it's like

24:32

if I give you a product that's free you as a user feel thislevel of like like you want to like in like it's like a

24:38

social contract where you feel like oh I want to dosomething nice for them they gave me this this you know this gift of like thisfree product that I use get so

24:45

much value out of i'm going to go rave about them to otherpeople you know and if you go look at our G2 scores we have

24:51

a which is like Yelp for for software we have a 5.0 ratingand we now have 5,000

24:56

reviews which is insane um in fact we were the highestscoring product across all of G2 last year and it's because of

25:03

this mechanic where we just give away so much value in ourfree product that it leads to people wanting to write reviews

25:08

wanting to tell their friends and that builds this likereally amazing kind of snowball effect but it only works if you

25:14

really invest in making it's like really high value rightgreat customer support great experience and it costs you nothing and so that'sa real winning

25:22

combination well a couple things that stood out to me withFathom and I'm not a paid sponsor here or anything but

25:29

number one was the immediate not uh the email you get withlike the takeaways

25:34

and action items uh when I saw that that was one of the veryfirst things I was like this is this is cool and we

25:40

switched over from um I mentioned Chorus over over to Fathomthe second one was

25:46

that um your team did an excellent job on onboarding becauseyou know even if it's free or it's inexpensive if they're

25:53

not using it or not using it properly I mean what's the useit's it's frustrating so as a learning development

25:59

company we spend a lot of time making sure we onboard ourcustomers because if if they're not you know logging in and

26:06

taking classes then they're not going to renew and andthey're going to have a bad experience now one other thing that

26:12

one of my sales reps brought up before this call uh thisinterview today was it

26:17

makes us look like a trusted adviser with our customersbecause a lot of times Richard we'll get on calls with

26:24

clients he says he does this all the time and they'll ask"Hey what is that that recorder?" Uh and we'll say "It's

26:29

Fathom by the way." And and what you guys did wasbrilliant here's like a referral code or a link where they can

26:35

just go in right there and so we're adding valueautomatically to our customers or prospects by giving them a

26:42

tool that we all need because we all have meetings we're onall day long yeah i think a lot to unpack through but you

26:48

know I think onboarding is also a place companies underinvesting like we spent nine months of meeting I think one or

26:53

two engineers on our team just every day improving theonboarding process to make it really really good i think that's

26:59

overlooked we also do a lot of stuff where like we can't fixeverything but we notice you know you hit some bug or

27:05

you hit you know some experience like maybe you kicked thebot out of the meeting we have all these emails we call proactive supportemails where it's like

27:10

if we notice something goes wrong we actually email you andsay "Hey we saw this went wrong here's how to fix it if you have anyquestions let us know." And

27:15

so I think I'm really happy to hear that like you guys had areally good onboard experience because I think that is so so

27:22

important to these things um and often overlooked how do youthough as a as a

27:27

founder keep the patience to do really focusing

27:33

on onboarding or some of the fixes when you probably get alot of pressure from people saying look we have to get this

27:39

to market or or we have to generate more revenue how are youable to you know manage that

27:45

without moving forward i mean that is one of the harderthings right it's like the the balance of sanding down rough

27:51

edges versus investing in big new features um for me it'sless of a the

27:57

tension between like people saying we need to drive morerevenue and more of like there's so many things we can build with AI now ai hascome so far in the

28:03

last 18 months 18 months ago the state-of-the-art for us wasyou know giving you great uh meeting notes and

28:10

action items and stuff like that and now thestate-of-the-art is we're rolling out things that can you know an AI that

28:15

can watch every meeting your team has had and you know writeinternal documentation answer questions right we

28:21

did the show me every moment that we ever seen in Fathomwhere someone had like a negative reaction to our pricing

28:26

or had a really positive reaction to one of our features oryou know write me a fivepage article explaining the history

28:33

of this feature at Fathom or blah blah and so right now thetension for us is there's so much you can now do with AI

28:38

that you couldn't do six nine months ago sitting on top ofthis data set of your company's meetings versus going in and

28:43

like sanding down the rough edges um the only way I've seenit to do it is basically just to basically dedicate x

28:49

number of engineers or percentage of your engineeringbandwidth to you know we're going to dedicate 20 to 35% of our

28:56

engineering bandwidth at any given time to just small UXfixes right um we've tried it sometimes where we'll do we

29:02

have less engineers when we had like three engineers wewould do a couple big features and once we have a big feature done we spend amonth on great it's like

29:09

little fixes month um now that we're a little bit larger wejust have like okay there's a few teams are dedicated to a

29:15

lot of these small fixes um but I think it's reallyimportant I think that's also something companies underinvest in like Well thefeature works for most

29:21

people and it's like so we're not going to invest in it umand so I think it's it has to requires you from an executive

29:28

level to say no no no we're always going to invest thispercentage of our resources into fixes one of the things I

29:34

like I heard you say is we suffer through especially sinceco meeting inflation how talk a little bit for

29:41

those out there who aren't using Fathom how uh by leveraginga tool like Fathom

29:48

it can cut down on the number of meetings people have andactually make your organization more efficient yeah I

29:55

think so my here's my big contrarian take i actually thinkmeetings are great um I think what what we don't like about

30:01

meetings are two things one is we don't like all the workthat comes with them the prep work the postwork the action

30:07

items the documents creation and what we also don't like isbeing on meetings where we didn't really need to be there

30:12

right it's like oh I'm on this recurring meeting i only talkabout 10% of the

30:18

time i 80% of the content is not relevant to me right and sofor the

30:23

former obviously AI what we do with them gets rid of allthat right you don't no longer have to do all that writing notes

30:28

action items that sort of thing and now what we're startingto do is focus more on that latter piece which is like how do we how do we helpcompanies have less

30:35

people in a meeting because now you basically have thisartifact it's like oh you know Tim in marketing one time

30:40

needed to hear this thing and so now poor Tim's gettingstuck on this recurring meeting every week right just in case something elsecomes up that Tim

30:47

needs to hear well that happens a lot right especially inhybrid companies or remote companies and the next thing you

30:52

know a meeting that should be three people is 10 peoplebecause what if someone needs to hear it but now we got

30:58

this artifact we got the recording the transcript thesummary and so now you can uninvite Tim from that meeting tim

31:03

will be very grateful and you know if something does come upTim needs to hear you like "Hey Tim you know check this

31:09

out minute five right?" Or Tim can go in and use ourlike ask them like our chat GPT like "Hey did they discuss anything

31:15

about project X?" Right great watch you know minute 3to minute 7 um and so it

31:21

really is going to allow us to you know I think companieslike Loom try to make meetings asynchronous by like oh record

31:26

things it's tough because like having a meeting is such anatural thing we do in New York we're coming in a different

31:32

direction which is no no have the meeting and the meetingitself then becomes an asynchronous piece of content you can use so you canbasically share

31:39

knowledge about your work and I think where we're going iseveryone will eventually have an AI agent of their own

31:44

that knows what things you know knows that Tim runs projectX Y and Z and so anyone mentions project XY and Z in the

31:50

meeting he's going to get a notification and get to watchthat clip of the meeting right and follow up so I think

31:55

we're going to get this really cool world in the next yearor two where information just starts coming to you without you having to go outand seek it

32:01

without you having to sit through a bunch of synchronousmeetings and so you get to consume all this stuff kind of at your own leisureone area it's helped us

32:08

a lot and become more efficient is around interviewinginstead of having to have multiple people go through and

32:15

interview either group interviews a lot of times somebody onmy team will

32:20

interview a candidate and then they'll send like a piece ofthat meeting to see

32:26

uh what they think of the candidate and it just saves a lotof time and and gets it to the point faster that's one of my

32:31

favorite use cases right because like I think mostcandidates experience one of these things is like oh I had three

32:37

meetings for this company and they asked me 70% the samequestions in every meeting right like and so now you can

32:42

kind of get a quick summary like double click on a thing andit becomes now we're actually each meeting feels much

32:47

more there's much more continuity right recruiter screenmeets recruiter meets them they do the screen great I pushed

32:53

on this question here's the answer you may want to doubleclick on it right I'll all the time I'll get on with the candidate be like okayI know you talked

32:59

to Nick about this let's go deeper on that right youmentioned this and this and this and I think at first they're

33:04

kind of blown away right they're like wait you know how didyou know that it's like you weren't in that meeting it's like no yeah we'rerecording it so I

33:10

think That's really one of was one of our first use cases ofFathom actually and so it really helps that like you get

33:15

to ask better questions the cans get a better experience andthen everyone also feels more read in right it's not like

33:21

you know hey I thought this thing was weird should we alllook at it and see if we should ask more questions about if we're concernedabout it great you don't

33:26

have to just take my word that their response was wasinteresting or maybe not what we're looking for you can go

33:32

watch it yourself right and so I think it really almostdemocratizes a lot of those hiring decisions a bit more and

33:38

makes them a little less uh fog of war i think what you saidwhich is very important is it makes it a better

33:44

experience for the interviewe they're not going through thesame redundant questions and they're like this company

33:50

has their their act together right i mean they're they'recommunicating well um they're not wasting my time asking

33:57

the same things and and they're advancing the conversation ithink that that that is that's a huge help what's

34:02

in the future for Fathom it's really exciting this yearbecause only in the last six months of AI got to the point

34:07

where I can do some of the things we wanted to do five yearsago right which is help me as a manager as a founder

34:14

basically get insight across a bunch of meetings I'm not inright so you know our head of success now has we're beta

34:22

testing this feature that's basically kind of like an AIsearch right it's an AI listening to every customer call and

34:27

looking for things and the AI understands intent andunderstands emotion so he's listening for like you know tell me every timesomeone gets

34:34

really frustrated with something right it could be just thatopen-ended just someone gets frustrated on a success call great he's going toget a real-time

34:40

alert about that um our VP of sales gets an alert every timethere is like hey anytime there's a pricing discussion

34:46

that didn't go well right uh anytime there was an objectionthat we didn't have a good answer for um the AA

34:51

understands what those things mean and can give you likebecause no one wants to watch a whole meeting so now what you're getting islike almost like a

34:56

highlight reel at the end of the day here's you know as aproduct guy I get a lot of like what were the big feature blockers forcompanies of the size right

35:03

and so uh that one's really exciting we're also looking atlast year we brought this thing called Ask Fathom

35:09

which is basically we saw a lot of people taking transcriptsfrom their Fathom meetings dropping the chat GPT and asking follow-up questionsright so

35:15

we kind of just build that directly into Fathom so you knowI'm watching this recruiting meeting hey did they discuss

35:20

salary oh did they discuss you know work experience and itcan kind of summarize it send you a report of the meeting now

35:25

we're going to be able to do that for your entire company weactually did a prototype of this a week ago where we

35:31

said "What's the what's the history of like we'retrying to make a a decision like design decision on the engineering

35:37

team we said "Hey what's the history of transcriptionengines that we've used at Fathom?" And went back over four years

35:42

of all hands and engineering meetings and I wrote this likefive-page document explaining every you know summarizing

35:48

every discussion we've ever had about this part of ourapplication kind of mind-blowing stuff right we don't have

35:53

to you know everyone hates updating the knowledge base andall the documentation again that comes out of meetings to kind

35:59

of memorialize decisions and discussions we had now we don'thave to do that work ai will just you know there's an AI now

36:05

let's watch every meeting your company's ever had and cananswer questions for you in near real time and so I think

36:10

that's going to be a huge game changer also for you knowshort-term things hey I'm on a meeting what's the status of

36:15

this project and it's like yeah there's AI that's beensitting in all the product meetings and knows the answer to that so there's somuch coming now

36:22

that's only possible with kind of the AI improvements thathave happened in the last 6 months or so i was having a

36:28

conversation with uh Courtney on my team who runs our uh LXteam and one of the

36:34

things I think this is kind of tapping right into whatyou're talking about is taking all of the conversations with a

36:40

client from the the discovery call through the completion ofa of a program and turning it into a case study by just

36:47

throwing all of the transcripts into it i don't know ifthat's available yet but I mean that that's a that's like a perfect uh use casefor it and then the

36:55

other one Richard I'm just kind of giving you my wish listhere so so you know is across companies so if if if we

37:03

could throw all the transcripts in across you know all ourcustomers and hear what their challenges are or their

37:10

pain points it helps us really build our value prop yeah uhso we actually today in the product as of last fall we have

37:16

you can look at a a specific company you can look at if youare using Salesforce or HubSpot you look at a specific deal

37:22

and we actually give you a summary for that deal we alsogive you like a ask fathom for that deal or company where you could ask youquestions right like

37:28

what are the top feature requests for this company or writethe case study for this company as you pointed out and yeah what we're reallyexcited about that's

37:34

in kind of alpha testing now is the latter thing which likeit can sit across your entire or so each department right like you point outmarketing might

37:40

be asking questions that inform messaging sales might beasking questions that inform sales training and so you know this idea has beensomething

37:47

we've been excited about for a while right if you can get100% of your meetings into Fathom you all of a suen get these really reallyamazing kind of

37:54

AI that really understands what's happening in your businessyeah and that's the thing you got to get them all into Fathom because you knowpoor data

38:01

or no data isn't going to get you anywhere but if if you'reconsistent and and you run the meetings through Fathom

38:07

and eventually can make that happen so that's awesome as wewrap up here what

38:12

is something that you want to share with our listenerswhether it's about the direction of Fathom or as a founder that

38:18

we haven't spoke about yet this age of building AI productsis is really interesting right i've been building

38:23

software for 20 years and now we're you know for most ofthose 20 years your

38:29

road map was just defined by what you wanted to build rightoh I want to build it because at the end of the day a lot of our products werejust forms or

38:34

there's you know they're they're web apps right now with AIlike you have a different totally different way you got to build it it's likewe're building it

38:40

backwards from what's technically possible right um and so Ithink that's a really important thing the other thing

38:45

with AI is like I think it's stress testing a lot of howpeople build product because there's no good way to

38:52

test for All right um it's really easy to get the ad to saysomething right and

38:57

so what's the difference between it saying somethingmediocre versus really good right if you're company like us

39:03

developing meeting summaries what's the best meaning summaryis it you know it's hard for it's hard to basically make a

39:10

test to do that you actually have to hire really good folksthat are really diligent read that and say "Oh this is

39:15

accurate but too verbose this is concise butinaccurate." Um and I think we're in this interesting era where likequality

39:21

of output really really matters i think it's also why a lotof the bigger companies struggle with producing some

39:27

AI features because they've lost that like artisal kind offeel they don't of like how to you know sit down like our

39:34

AI team will literally look at hundreds and hundreds of testoutputs to figure out this is the right incantation of the

39:39

AI models to use to generate good action items goodsummaries etc and so if you're building AI products or you're

39:46

someone that's consuming AI products I think this is thenumber one thing to think about is if I'm buying a product how do I test theaccuracy of the output

39:52

how good it is if I'm building it how do I build a team thatis really incentivized to make sure we ship

39:58

something that's really high quality and not just somethingthat kind of Yeah that looks like it right like spit out something um and so Ithink this is

40:04

going to define the good versus great companies in the AIera right because with AI I mean it will answer you it

40:13

just you know it whether or not it's a good answer or eveneven truthful is a completely different thing and what I

40:18

hear you saying is the importance of pushing back and andbeing intentional

40:24

uh and and and taking it from the next level of good togreat which could really separate uh a lot of the

40:30

companies out there who are who are building or producingwith AI you know is going to be a game changer moving forward so I mean we'vehad an awesome

40:36

conversation we we started off talking about your journey uhwhat you learned from other founders and mentors and and

40:43

really your cur curiosity of what it took to to get Fathomwhere it is now

40:49

and and the direction it's going to go in the future wedidn't talk much about uh building and scaling teams but you

40:55

mentioned the importance of having a strong team and notnecessarily a large team all the time and uh where we're

41:03

going with AI so Richard I you know really appreciate youbeing here today i'm like I mentioned again a huge fan of

41:09

Fathom we're going to put a link uh to the Fathom of coursein our in our show

41:14

notes but where else can our uh listeners connect with youat yeah I'm on LinkedIn uh as I mentioned earlier

41:21

like some of the best things of my career have come fromcold outreaches so I try to try to return the favor so if

41:27

you have any questions about Fathom feedback on Fathom orjust you know thoughts or questions about you know

41:32

startup game um I'm here on LinkedIn happy to hear from youand listeners I want to leave you with if you know

41:39

somebody who is a sales leader a founder or even running Idon't know a a

41:44

marketing team and they're not using a tool like Fathom sendthis um episode to

41:50

them there's huge value in it for them because it's a toolthat we should be leveraging uh across the board

41:55

throughout all the departments and until next time everybodystay curious keep

42:01

learning have a great day thank you Richard are leadinglaboratories they're incredible scientists in their own right

42:08

but they've never really sat in a program around leadershipso fascinate

‍