Latest / The Prompt / Ep. 550 | How Intercom fixed bad Salesforce data and revived outbound prospecting

Transcript

Ep. 550 | How Intercom fixed bad Salesforce data and revived outbound prospecting · The Prompt

• 0:01
  This is on base your one stop shop for solving your most
• 0:04
  pressing B2B go to market challenges.
• 0:07
  Each week, host Chris Moody and Paul Gibson talk to sales and
• 0:09
  marketing leaders to get you in depth insights and creative
• 0:12
  strategies to set your revenue teams up for success.
• 0:18
  Everyone, welcome back to the Own Base Podcast.
• 0:20
  I'm your host, Chris Moody, and today I'm very excited to be
• 0:23
  joined by Alexander Domuan to talk about smarter prospecting.
• 0:29
  Alexander, thanks for joining us today and welcome to the show.
• 0:32
  Thanks so much, happy to be here.
• 0:33
  Well, we've already talked about this first product.
• 0:36
  Plug the glasses and tell folks what you wish you had for the
• 0:39
  glasses. We have to get this out of the
• 0:40
  way. I mean seriously.
• 0:42
  So I've this fun new thing happening lately where whenever
• 0:44
  I look at a screen, my phone, my laptop, my monitor, I
• 0:49
  immediately start getting a migraine.
• 0:50
  These kind of help a little bit. It's been a bit of an
• 0:52
  occupational hazard, but I am looking for a sponsorship
• 0:56
  opportunity. These are like cheap Amazon, but
• 0:58
  I think there's an opportunity to get some like really cool
• 1:00
  branded Valenciaga some sort of situation because I these might
• 1:04
  end up being a long term thing. So we'll see if anyone's out
• 1:08
  there we'll chat after. We will attack them all in the
• 1:11
  comments. Maybe that's what we do.
• 1:13
  We just try to turn this into AB to C initiative. 100% all.
• 1:17
  Right. Well, I love it.
• 1:18
  Well, we like to start first by understanding your journey.
• 1:21
  What led you to what you're doing today?
• 1:23
  Yeah, totally. This is my favorite story to
• 1:25
  tell. I am a failed salesperson, so my
• 1:28
  first job out of college was in like university fundraising.
• 1:30
  So it was similar to what I did in like as an intern in college
• 1:33
  and then after a year that I I got my first tech sales job.
• 1:36
  I was an AE for a company up in Seattle.
• 1:38
  I lasted 10 months. I was terrible at it.
• 1:41
  It was the hardest thing I've ever done and I got fired for
• 1:44
  not even coming close to my quota.
• 1:45
  Honestly, they were really nice to keep me on for 10 months.
• 1:48
  Then I went and and became an SDR. So I moved down, and I was
• 1:52
  at a company for like a teeny bit over a year, also in
• 1:56
  Seattle. And then, yeah, I was terrible
• 1:58
  at that, too. Like, I had this deep fear of
• 2:00
  inconveniencing people, which is not easy when you're
• 2:03
  prospecting. So I also got fired from that
• 2:06
  one and then got to move into my mom's unfinished basement in
• 2:09
  Portland, OR my hometown. And then I ended up getting a
• 2:12
  kind of like a very ground floor temporary gig at a software
• 2:16
  company there called Smarsh where they had kind of an
• 2:19
  abandoned part on instance. Like they wanted someone who's
• 2:22
  familiar with Salesforce just come in and do like general OPS
• 2:25
  work. Ended up staying there for
• 2:26
  almost 4 years and really realized, oh, I like marketing
• 2:29
  OPS. I like sales OPS.
• 2:31
  I like, like before then, like revenue operations hadn't really
• 2:34
  been a term that was known, but like this kind of work, I
• 2:37
  genuinely enjoy it. It also seemed like it was the
• 2:40
  work that everyone around me in actual marketing and sales like
• 2:43
  strategy positions did not want to do.
• 2:46
  So it felt safe. And that's definitely been the
• 2:48
  case. So yeah, after about four years
• 2:50
  and I realized like, OK, even though I definitely can't do
• 2:52
  sales, I can do this. I moved on to San Francisco, got
• 2:55
  my first job where I was the only and 1st marketing OPS hired
• 2:59
  and I was at a company called Circle CI for about 2 1/2 years
• 3:02
  where I learned a lot, made a lot of mistakes, really cut my
• 3:06
  teeth, became a people manager, learned how to like actually set
• 3:09
  a strategy and, and develop a point of view that mattered for
• 3:12
  the business. And then I moved here to
• 3:14
  intercom. It'll be four years in
• 3:16
  September. So like 3, almost four years now
• 3:20
  where initially I LED marketing OPS here for about a 2 1/2
• 3:25
  years, I think. And then for the last little
• 3:27
  over a year, I've been a director of Rev OPS.
• 3:29
  But really I call myself a director of go to market OPS
• 3:32
  because that's really my function.
• 3:33
  It's all the things that I used to own with marketing OPS.
• 3:36
  But then I also own a sales OPS function that focuses on new
• 3:39
  business and prospecting and outbounding with a real emphasis
• 3:43
  on all the technology that the growing landscape of technology
• 3:48
  and that supports those functions.
• 3:50
  So yeah, I absolutely love what I do.
• 3:52
  I frequently feel like I've no idea what I'm doing, but I think
• 3:56
  that's my favorite. Like honestly, that's my
• 3:58
  favorite story to tell because like you can truly fall down the
• 4:01
  stairs. And I think failing really
• 4:03
  quickly and learning like this is absolutely what I'm good at.
• 4:07
  Like this is where I'm very bad. It really forced me to figure
• 4:11
  out how to specialize and and triangulate based on the skills
• 4:14
  and the things that I did have. And it's turned into this career
• 4:18
  that it sometimes it feels like I'm falling up the stairs.
• 4:20
  I'm like, I don't know how this is happening, but I think like
• 4:22
  that's, I think a lot of OPS people are feeling this way
• 4:25
  because technology and like just the, the unsexy yet really
• 4:29
  important things of just making sure data is flowing from A to B
• 4:34
  and it's what one team is putting in is understood by the
• 4:37
  other team is still so foundational.
• 4:40
  And so many businesses are really recognizing that.
• 4:43
  Like it's good when it, there's a good strategy behind it, but
• 4:46
  it's awful when you don't have that running well at all.
• 4:49
  So. I love that journey and you have
• 4:52
  your book title figured out too. From the basement to go to
• 4:55
  market OPS. I think that would be a cool
• 4:57
  story. We have since actually renovated
• 4:59
  that basement into a full finished Airbnb apartment.
• 5:02
  If you are ever in Portland, you can stay in a beautiful
• 5:05
  neighborhood. Irvington.
• 5:06
  Yeah, so. The link in the comments.
• 5:09
  And there you go. Just kidding, you have a long
• 5:11
  term tenant. I don't know but.
• 5:13
  Well, that's awesome. I mean, look, that's indicative
• 5:16
  of your success too, right? Like the basement is now it has
• 5:20
  it's glow up this awesome place. Well, I love that.
• 5:23
  And you talked about the ability to fail.
• 5:26
  I think that's such an important thing that we all have and
• 5:29
  embracing strengths. I know personally, I used to
• 5:32
  focus so much on the things I didn't do 10 out of 10 and think
• 5:36
  in my head, I have to get better.
• 5:37
  Yeah. And well, let's talk about
• 5:40
  something that a lot of people fell at.
• 5:41
  And that was a terrible podcast segue, but it's, it's related
• 5:45
  because the list building process, what makes transforming
• 5:51
  that into a strategic priority successful for sales?
• 5:56
  Because a lot of folks just start with, well, we want to
• 5:58
  close, you know, Apple and Amazon and Adobe and Oracle and
• 6:04
  they just list the biggest accounts they can think of and
• 6:06
  that becomes the list. Yeah, great.
• 6:09
  Yeah, go for it. Now this is such an important
• 6:11
  thing to talk about because of the things that like, again,
• 6:15
  kind of like what I was talking about earlier, like the things
• 6:17
  that are like least sexy are what matter the most or a good
• 6:22
  list. Because what's fundamental here
• 6:24
  like this is this is a key process where marketing is
• 6:27
  involved in setting strategy and making decisions and then
• 6:31
  handing this off to sales to execute.
• 6:34
  So that trust between the two teams becomes absolutely
• 6:38
  paramount. And so the reason why I've done
• 6:41
  this multiple times in my career, I think I've tried to do
• 6:44
  some version of like, let's get good outbound lists ready for
• 6:47
  sales. So they always have accounts to
• 6:49
  call. I think maybe 5 * 3 times here
• 6:52
  at intercom. And most times they've been an
• 6:55
  absolute failure. And it's fundamentally came down
• 6:58
  to the fact that sales did not trust what we were sending over.
• 7:02
  They didn't feel as involved in the process.
• 7:05
  And the worst part about it is each time that we failed, that
• 7:09
  trust got lower and lower and lower to the point where like
• 7:14
  that good faith we're like we wanted to try an interactive
• 7:16
  campaign, some sort of a new function or initiative.
• 7:20
  Sales was really hesitant and for good reason because they did
• 7:24
  not want to give us their time on something unless they felt
• 7:28
  that it would very clearly go towards them hitting their
• 7:30
  number. So yeah, I mean, without getting
• 7:33
  too much into like maybe you need to do a a segment or a
• 7:37
  transition, but that's why I'm here because I'm going to talk
• 7:40
  about like what I did this time around to to try to really build
• 7:44
  back that trust through the process of building lists and
• 7:48
  how that worked out for us. Yeah, I will.
• 7:50
  Before we segue, we've got to dig the dagger a little deeper.
• 7:54
  The the. Pain points there, right?
• 7:57
  What were some of those lowest moments where you had that
• 8:02
  crisis and and thinking, how are we going to do this?
• 8:04
  That drove the need to do the radical transformation.
• 8:08
  So I think what has always impacted us that we've just
• 8:13
  never been able to fix at scale is like this might be
• 8:16
  unsurprising or unfamiliar for some people, but every single
• 8:19
  company I've ever worked for has a dirty Salesforce instance with
• 8:22
  a lot of dirty data, a lot of duplicate, a lot of unreliable
• 8:26
  enrichment. And so as a result, like no
• 8:29
  matter what, if you identify the absolute best accounts, like say
• 8:33
  you've got the best intent data, you've got the best kind of like
• 8:37
  firmographic enrichment, like you've identified who's
• 8:40
  absolutely the best fit. The Achilles heel was really the
• 8:43
  fact that we could never 100% trust that what we were sending
• 8:46
  sales was truly an account that they could call.
• 8:48
  It was solving that unsexy yet extremely important, like Roe
• 8:53
  standpoint of can I trust that this account, there isn't a
• 8:58
  duplicate of it that has an open opportunity or there isn't a
• 9:00
  duplicate of it with a slightly different website that an SDR is
• 9:04
  already working? Or there isn't a duplicate of it
• 9:06
  with a very, very, very slightly different clear, but domain that
• 9:10
  also has a like that just spun up a free trial.
• 9:13
  Like I don't think sales trusted us to be able to encapsulate all
• 9:16
  the context, not about whether the account was good to like,
• 9:20
  you know, they they could trust like thermographic information.
• 9:23
  I think what they couldn't trust is that we could actually
• 9:25
  operationalize these lists at scale, which for US meant
• 9:28
  ensuring that everything that we sent them was something that
• 9:31
  wouldn't be a waste of their time.
• 9:32
  After they're, you know, they're, you know, a few days
• 9:34
  into prospecting and then realize, oh, this is an account
• 9:37
  I can't go after. So like, I think that's where so
• 9:41
  many good intentions went to die.
• 9:42
  Like every single time we tried to do this, we couldn't
• 9:45
  operationalize this. I think this is something that a
• 9:48
  lot of, I think a lot of tools, a lot of intent tools and other
• 9:51
  tools that really help with account based marketing and
• 9:54
  identifying and targeting the right accounts.
• 9:55
  Like at the end of the day, everything has to be matched
• 9:58
  back to our own existing data and no tool can clean that up.
• 10:02
  So this was always really difficult because cleaning this
• 10:05
  up at scale or even getting to a place where we can understand,
• 10:09
  like can we truly trust that like this account is a workable
• 10:15
  account? That took a really long time and
• 10:18
  it honestly, it was something where I had to get on systems
• 10:20
  road map. I had to make sure that we had
• 10:22
  the sales OPS resources. I had to block out two weeks on
• 10:24
  my calendar to be in spreadsheet hell.
• 10:27
  It just wasn't tenable. And so I think that's
• 10:29
  fundamentally where we always fell flat, you know, no matter
• 10:33
  like marketing could have done. And often it was something where
• 10:35
  I felt really bad because like marketing did everything they
• 10:38
  should do, like they would generate really good lists.
• 10:40
  We had some intent providers that I think actually did a
• 10:42
  great job of identifying the right accounts.
• 10:45
  But when it came back to like, OK, can sales actually take
• 10:48
  meaningful action against these accounts, the answer was always
• 10:52
  like, yeah, you sort of, we hope so.
• 10:54
  And that's that's just not going to cut it.
• 10:56
  Yeah, my brain was on. We hope so.
• 10:58
  And then you said it. That's exactly what I was
• 11:00
  thinking. You know hope is not a strategy,
• 11:01
  right? No, it's not.
• 11:03
  How did you leverage unique enrichment tools Clay Gen.
• 11:07
  AI to enhance the accuracy and relevance to start to improve
• 11:10
  that data and ultimately build that trust?
• 11:13
  Totally. So yeah, I'll set the stage like
• 11:15
  we were developing a regular reliable outbound pipeline
• 11:20
  engine has been a priority for us for the last couple of years.
• 11:23
  I think it's it's a priority that any successful SAS business
• 11:26
  should have. Like no matter how good your
• 11:28
  inbound or your PLG is should always make sure that like
• 11:30
  they're your top target accounts, like you have some
• 11:32
  sort of an outbound program against them.
• 11:35
  So we've done multiple iterations of this at intercom
• 11:37
  and like intercom's changed a lot in the last four years.
• 11:39
  And so it like intercom's passed.
• 11:41
  We tried this and then most recently last fall, we did our
• 11:44
  first kind of like iteration of this where, you know, we got
• 11:47
  that buy in and sales said we're going to Yep, we want to start
• 11:51
  outbounding like will you help get us the list?
• 11:53
  We'll make sure we're targeting based on XY and Z.
• 11:56
  And so we did that and September 1st we signed out the accounts
• 12:00
  and then like a couple weeks later, like I got a slack from
• 12:03
  someone high up being that this was Sunday morning or Saturday
• 12:05
  morning, I got a slack that was like, I'm hearing some really
• 12:09
  unfortunate, like really troubling things from the field
• 12:11
  about like the accounts. Like I know we spent a lot of
• 12:14
  time here like making sure these were good, what happened?
• 12:17
  And so immediate, immediate panic because again, like that
• 12:20
  my worst fears which were like, are these actually good accounts
• 12:23
  that felt like those were coming true?
• 12:25
  So like, I think a lot of good things that come out of OPS,
• 12:28
  like the initial impetus was like an uh oh moment.
• 12:33
  Like, oh, I really messed something up or this absolutely
• 12:36
  failed. I need to get creative.
• 12:39
  I need to figure out a way that I can ultimately do this right.
• 12:43
  So there are two parts to this. Like there's the actual tactical
• 12:46
  part, which is where clay comes in.
• 12:48
  I think the beauty in clay is that it truly allows a single
• 12:52
  person who is alone in their apartment over a weekend,
• 12:56
  stressed out to ideate, to try things out, to create
• 13:01
  prototypes, to try to understand what could be possible at a
• 13:05
  really small scale and then makes it really easy to scale
• 13:08
  that out. And so that's kind of the
• 13:09
  journey operationally for what we did with this specific
• 13:14
  program. More important, however, was the
• 13:17
  fact that this for this list, even though sales was involved,
• 13:21
  we did not go nearly far enough to make sure that we were
• 13:25
  listening to them to understand what does a good account look
• 13:28
  like rather than built like we've got excellent data
• 13:32
  scientists and and folks internally at intercom who've
• 13:34
  built some like really cool models and like can truly do
• 13:38
  some amazing work at predicting the likelihood like propensity
• 13:41
  is something we can do really well here at intercom.
• 13:44
  But what matters here, because of the history of all of that
• 13:47
  lost trust, all of us saying trust us, this is going to to be
• 13:51
  a good list. It's going to oh, oops, sorry.
• 13:54
  Like sales obviously was like, yeah, no, we don't trust you.
• 13:58
  So I had to make sure anything we were delivering them was
• 14:01
  optimized towards what they thought was a good account.
• 14:05
  So like, even if they said something like we only want to
• 14:07
  call companies that make peanut butter and like we didn't have
• 14:11
  that propensity score. OK, you know what we needed?
• 14:13
  We needed to start that. We need to make sure that sales
• 14:16
  could start actually outbounding so that we could generate that
• 14:20
  feedback. We could start seeing what's
• 14:21
  working, what's not working. We were at an impasse where, you
• 14:24
  know, sales didn't want to call out on accounts that they didn't
• 14:27
  trust. So if they said we only want to
• 14:29
  call accounts that have peanut butter on their website, great.
• 14:32
  What can I do to try to source as many accounts that are
• 14:35
  talking about peanut butter? Now, obviously that's a
• 14:37
  ridiculous example, but I went to sales directly.
• 14:39
  I had multiple calls, worked with them on multiple fronts to
• 14:42
  say like if you had a magic wand, if you were able to like,
• 14:46
  if we could give you look alikes of like your ideal accounts.
• 14:50
  What does that look like? What are the things that you're
• 14:52
  looking for when you get a lead, when you get a new opportunity
• 14:55
  from an SDR? What are you doing to like for
• 14:58
  your own kind of Spidey sense sanity check?
• 15:00
  Every salesperson kind of did something different.
• 15:02
  And so we aggregated those together and got folks to agree
• 15:06
  on two different sets of criteria, which I think is
• 15:08
  really important. The 1st is OK, who should like,
• 15:12
  if anyone meets any of these criteria, they're just not,
• 15:15
  they're not a good fit. So you could be amazing, but
• 15:19
  sorry, you're located in a country that we just don't ever
• 15:22
  sell to successfully. So we're going to exclude you.
• 15:25
  We had to make sure we got that right and that was established.
• 15:27
  Or we could say like, OK, the accounts that truly aren't
• 15:30
  workable, we're going to make sure we don't assign you any of
• 15:32
  those. The second part was OK, if the
• 15:34
  accounts that need all that criteria, so like baseline
• 15:38
  they're workable, how do we prioritize the ones and assign
• 15:42
  out the ones that most closely resemble what you sales have
• 15:47
  said are the best fits? And so that's where Clay came in
• 15:50
  and actually allowed us to realize a lot of those like,
• 15:53
  well, if I had a magic wand, I would want XI would want Yi
• 15:56
  would want Z. And a lot of what sales told us
• 15:59
  like this was an iterative process.
• 16:01
  We assigned out the first list, then the next month we assigned
• 16:03
  out the second list. After getting that feedback, a
• 16:06
  lot of those things that sales actually told us anecdotally, we
• 16:09
  have turned into programmatic enrichment.
• 16:11
  Like we are always enriching accounts on an always on basis
• 16:15
  for these like wacky unique things that you could never get
• 16:17
  from a legacy data provider. But it's because sales told us
• 16:21
  these are the things that we actually care about.
• 16:22
  These are the things we're looking for when we're scanning
• 16:24
  a website. And with a tool like Clay, we're
• 16:26
  actually able to go through and programmatically scrape a
• 16:30
  website, determine if, for example, something that's really
• 16:33
  great for us. If you have a login button, if
• 16:35
  you have a sign up now button, if you have some sort of a My
• 16:38
  account button, something that indicates there's another
• 16:40
  customer portal on the other side of the marketing website,
• 16:43
  you're gonna be a good fit for us.
• 16:45
  That means that there's an Ave. where people interacting with
• 16:47
  you, they're creating support tickets.
• 16:49
  And so that, for example, how are you going to scale that out?
• 16:52
  How are you going to enrich 50,000 potential accounts to
• 16:55
  determine which ones have a login button?
• 16:56
  I mean, there are ways, but not ways that I could do a loan in
• 16:59
  my apartment on a Saturday, but with Clay, I absolutely could.
• 17:04
  Awesome. You answered my next question
• 17:06
  too, talking about no, no, no, it was perfect and you nailed it
• 17:11
  about building trust. I mean, the first thing I was
• 17:13
  going to say, I was really hoping you would have said
• 17:15
  basement instead of apartment just for the call back.
• 17:18
  You know, just my OCD was like, oh, go basement, go basement.
• 17:21
  Sorry I have a but. We didn't get there.
• 17:23
  You moved out of the basement to an apartment.
• 17:25
  So good on you. We'll let you have that when.
• 17:28
  OK. I don't live in the basement
• 17:30
  anymore. The part that I love though, the
• 17:33
  trust and feedback, one of my favorite customer interviews
• 17:36
  I've ever had here at Demand Base was Jody Weibo at Hexagon.
• 17:40
  And there was this quote. As soon as she said it, I wrote
• 17:43
  it down, put it on a slide, started sending it out to
• 17:45
  everyone. And it was, I don't ever want
• 17:47
  sales to come to me and say, why did you pick this account?
• 17:50
  And having those conversations is so important to hear that
• 17:54
  first hand. No matter what you're doing with
• 17:57
  the tech, you have to hear it and act on it.
• 17:59
  And that was a great answer of how you did that.
• 18:02
  What were the biggest operational or technological
• 18:06
  challenges that you faced in doing that?
• 18:08
  So you hear the feedback, you're working to improve it.
• 18:11
  How did you address all of that feedback?
• 18:14
  Yeah, I mean, it was tough. I think the biggest thing is
• 18:16
  this is before we ended up doing like full scale programmatic
• 18:20
  enrichment using clay as like the orchestrator and in some
• 18:23
  cases the source of a lot of the data that we did for accounts.
• 18:26
  A lot of this work that I had kind of prototyped that and then
• 18:29
  helped us identify what did we want to enrich on a regular
• 18:31
  basis. But I think again, it was
• 18:33
  solving those problems that weren't like like it was
• 18:35
  actually, it was fun and it was kind of easy in comparison to do
• 18:39
  things like write the query to figure out which counts are like
• 18:43
  have these cool things or have X or have Y or have ZI think the
• 18:46
  hardest part was once again, making sure that anything that
• 18:49
  we assigned was actually workable.
• 18:51
  So I think historically this has been really difficult, at least
• 18:55
  for me as like as like a point and click operator who can
• 18:58
  barely do sequel without ChatGPT.
• 19:00
  Like it was always really hard to solve that main issue
• 19:04
  blocking the trust, which was like, can we ensure that like
• 19:07
  these accounts are callable? There's they don't have a
• 19:10
  previous existing relationship with intercom that would prevent
• 19:14
  me from being able to reach out in terms of their already
• 19:16
  self-serve. There's already an open
• 19:18
  opportunity. They recently turned.
• 19:20
  This might surprise you. We have a ton of duplicates.
• 19:22
  We have a ton of duplicates in our sales force account or in
• 19:25
  our sales force instance. Everyone does.
• 19:27
  Everyone does. I feel like a company of our
• 19:29
  size. If we had a perfectly clean
• 19:30
  Salesforce instance, it would mean that we were not great with
• 19:33
  allocating budget for headcount and we were spending too much
• 19:36
  money cleaning up Salesforce data rather than on more
• 19:38
  strategic initiatives. It's just par for the course and
• 19:41
  solving all of our Salesforce data problems, specifically
• 19:44
  around duplicates, if that is necessary to assign out these
• 19:48
  accounts and to get a good list of sales, we're never going to
• 19:51
  get there. We're never going to get there.
• 19:52
  We're making changes, we're making improvements, but there
• 19:54
  will always be that level of unreliability and there will
• 19:57
  never be such thing as perfect data and Salesforce.
• 20:00
  What we had to get to is managing the expectations with
• 20:03
  sales to say, yeah, we've got dupes, they're in there, they're
• 20:07
  annoying, we're working on them. We've got bad data and
• 20:09
  Salesforce, we totally get it. But we can confirm that every
• 20:12
  account that we're assigning you.
• 20:14
  There isn't a mystery duplicate. Not only like this account is
• 20:17
  workable, there isn't a duplicate out there that isn't.
• 20:20
  And so that was something that I think was absolutely hardest to
• 20:23
  do with clay. It's extremely difficult to do
• 20:26
  with any other provider. I haven't figured out a way to
• 20:29
  do this. It's essentially you can pull a
• 20:31
  report in Salesforce of good accounts, but how do you know
• 20:34
  that there isn't another one out there?
• 20:36
  And so that's something that I'm actually most proud of.
• 20:38
  It's again, something that's not as like glamorous as some of the
• 20:41
  other things, but it's a series of Socal queries where
• 20:44
  essentially we're saying we have Clearbid, Clearbid's kind of our
• 20:47
  our source of truth for like what is a company.
• 20:49
  And one thing I like about Clearbid is that it'll enrich an
• 20:52
  account down to just it's simple domain.
• 20:55
  So for intercom, for example, you could have intercom.com or
• 20:58
  intercom.aiorintercom.co.uk or some really long URL and have
• 21:03
  all that in the website field. Clearbid enrichment will run in
• 21:06
  return just the simple domain. For better or for worse.
• 21:08
  It's not perfect. That's kind of what we use to
• 21:10
  identify a unique company. So what I did was just create
• 21:13
  like 6 columns in this clay table that like pinged
• 21:17
  Salesforce to look for any account with the same clear bit
• 21:20
  domain as the account in this list that I think might be a
• 21:23
  good fit for sales. Do any of those, do any accounts
• 21:26
  with the same clear bit domain have an open opportunity?
• 21:29
  Are they currently paying us self-serve revenue?
• 21:31
  Are they currently paying us contracted ARR?
• 21:34
  Have they recently turned in the last 90 days?
• 21:36
  Do they already have a sales strategy assigned, which means
• 21:39
  that they're an account that we're already focused on for
• 21:41
  some sort of reason? Just basic?
• 21:43
  Oh, and then finally, has there been activity in the last 60
• 21:46
  days, the most basic of things? That's something that we
• 21:49
  actually got very right. Like, and I'd say like I around
• 21:53
  95% of the accounts that we have signed out met all that
• 21:56
  criteria. There's always going to be the
• 21:57
  mysteries. There's always going to be the
• 21:59
  weird ones in there, but that was huge.
• 22:01
  I'm most proud of that because I think that was quintessential.
• 22:04
  That was absolutely necessary for sales to have that initial
• 22:08
  willingness to start reaching out to these accounts.
• 22:11
  They were still skeptical as they very well should be.
• 22:14
  So I think they went to a lot of these accounts and did their
• 22:16
  initial like, OK, I got this account.
• 22:18
  Let me let me just check, let me try to find in Salesforce.
• 22:21
  Is there another one? Is this truly one in that I
• 22:23
  think they started to realize, OK I can can start to reach out
• 22:26
  to these. That's awesome, I love it.
• 22:28
  It's such a great example. It's very specific, which I
• 22:31
  love. What were your favorite examples
• 22:33
  of the benefits that came from it?
• 22:35
  It could be metrics, it could be stories, it could even be quotes
• 22:38
  from sellers. What stood out to you?
• 22:41
  Yeah, I think in genuine, like I don't think I've ever had a
• 22:44
  salesperson ever tell me this is a really good list.
• 22:46
  Thank you for sending it to me. I can't wait to cold outbound to
• 22:49
  it because obvious, I mean from my like fired salesperson self,
• 22:53
  I'm like, no one wants to cold outbound to anyone, no matter
• 22:56
  how good the list is, right? But I started to get those
• 22:59
  comments and like like I mentioned, this was iterative.
• 23:02
  Like sometimes like we have to address the fact that like our
• 23:06
  Tam is limited and you know, the accounts that look a lot like
• 23:09
  the accounts that we've always sold to, there just aren't that
• 23:12
  many of them, especially kind of post 2022.
• 23:15
  A lot of acquisitions have happened, a lot of contractions
• 23:17
  happened in the SAS space. So we in order to outbound like
• 23:21
  we're going to start pushing the limits of that and start
• 23:23
  outbounding to some accounts that might not look exactly like
• 23:26
  the customer profiles that we have from some of our existing
• 23:28
  customers. So the best part has just been
• 23:32
  that anecdotal feedback that like I mostly the best part for
• 23:35
  me, for my ego and just for my sense of relief was that like we
• 23:39
  did it. We were actually able to solve
• 23:40
  this problem. That I think at the back of my
• 23:42
  mind, I always thought like, we're never going to be able to
• 23:44
  do this. Like this is just always going
• 23:45
  to be a thing. And then that trust will never
• 23:47
  be able to be built back. What I think is much more
• 23:50
  important is like on a broader scale, this, I think this was
• 23:52
  really important for just that general sense of trust between
• 23:56
  sales and OPS, like my function and also between sales and
• 24:00
  marketing. Because in order for like you
• 24:03
  know this better than anyone working in a company like demand
• 24:06
  base, like in order for any sort of account based marketing or
• 24:09
  really just good marketing to be successful, sales and marketing
• 24:13
  absolutely need to be in lock step.
• 24:15
  They need to be understanding the strategy, they're part in
• 24:17
  it. It's a shared goal.
• 24:20
  When that trust isn't there, it is next to impossible to get any
• 24:23
  sort of true collaborative initiative like this off the
• 24:27
  ground, especially something like a like a wide scale
• 24:29
  outbound program like this. So because we were able to get
• 24:33
  over that initial hump of like bad data, we were able to
• 24:36
  understand what does it mean. When you say bad data, do you
• 24:38
  mean like genuinely like I can't call this account or like the
• 24:41
  information on this account is wrong Or there are like the this
• 24:45
  is not an account that like fits what we could, what we should be
• 24:48
  calling like those are kind of the three different things that
• 24:51
  bad data can mean. We didn't think this was going
• 24:54
  to solve the problem. We just needed to get this out
• 24:57
  of the way so we could figure out, is there something else
• 24:59
  that's like blocking outbound from being successful?
• 25:02
  Because right now it's just bad data.
• 25:05
  That's all we're hearing. And so as we were able to start
• 25:07
  solving that bad data problem, or at least solving parts of it,
• 25:11
  we could start to under get that feedback on other things like,
• 25:14
  oh, like this messaging is not landing or like these verticals,
• 25:17
  like we really need more support from product marketing or like
• 25:20
  more collateral. Or maybe we should, we should
• 25:23
  roll back on this or even some circumstances where sales was
• 25:25
  more than willing to be like, we thought this would be a good
• 25:27
  type of account to go after. We aren't seeing success.
• 25:30
  So let's no longer focus on these.
• 25:32
  Let's let's filter these out and instead focus on XY and Z.
• 25:35
  So it was by no means perfect. This is still not perfect.
• 25:39
  Like we just recently assigned out all the accounts for you 2
• 25:42
  and we've got vibrant and voluminous feedback.
• 25:46
  But what we have is that that trust both ways, where we feel
• 25:50
  like we can safely, you know, call out what's working, what's
• 25:53
  not working. And that's not just one sided,
• 25:55
  it's it's not even bi directional.
• 25:57
  It's like pan multi directional. Like marketing is involved,
• 26:00
  sales is involved, OPS is involved.
• 26:02
  We're getting our data science teams involved to see like how
• 26:04
  can we start to tweak scores? And it means that like sales is
• 26:08
  able to give us that really good feedback on what works, what do
• 26:11
  they want to do? And we're also able to give them
• 26:13
  feedback to BIC. OK, we gave you the accounts
• 26:15
  that you said were were good and we're not seeing activity.
• 26:18
  What happened? So what else is blocking are
• 26:21
  there? You know, like sometimes it's
• 26:23
  just like I had too many inbound deals that I had to focus on.
• 26:26
  OK, that's a good problem. That's a good problem to have,
• 26:28
  but maybe we'll take a look at that.
• 26:30
  But then other times like, you know, accountability needs to go
• 26:33
  both ways. If we're investing in this kind
• 26:34
  of a program, we want to make sure that everyone's on board.
• 26:37
  And so again, none of that would have been possible had we not
• 26:41
  just come back to basics and really indexed on that first or
• 26:46
  well, I guess that second list after the first one that really
• 26:49
  flopped to see like a good account is an account that in
• 26:52
  your eyes as a salesperson, you think is good because just we're
• 26:56
  trying to solve what will get you to start outbounding.
• 26:59
  Once we solve that, that'll open the door, where, as we've
• 27:02
  discovered, there are plenty of other problems.
• 27:05
  There are plenty of other reasons why outbound, you know,
• 27:07
  might not be as effective. And so we've got plenty of job
• 27:10
  security, plenty of other problems to solve.
• 27:13
  But with this bad data mountain and this lack of trust in the
• 27:16
  way, we couldn't even get close to those.
• 27:18
  Yeah, that is such a good example.
• 27:20
  And you've hit on failures, you've hit on flops.
• 27:23
  We've both been fired from jobs. Most people listening have no
• 27:26
  shame in that. It's happening straight.
• 27:29
  Yeah. What's a business risk that you
• 27:32
  took that didn't work and what were your learnings from it?
• 27:36
  Yeah, a business risk that I took here at Intercom.
• 27:39
  You know, I don't know how much of a cop out this answer is, but
• 27:42
  I think I've often been too afraid to take really risky
• 27:44
  bets. And that in and of itself has
• 27:46
  been something where I've really prevented myself from being able
• 27:49
  to do something like what I did here where I did kind of take a
• 27:52
  risk. I was like this process that we
• 27:55
  like all collaborated on to create a list of accounts that
• 27:58
  ended up failing. I kind of just like took this on
• 28:00
  my own and I was like, let's let's try this, let's move
• 28:03
  forward with this. I would say in the past, it's
• 28:06
  I've been too afraid to challenge when I thought
• 28:08
  something wasn't going to work. So I think in some of these past
• 28:11
  iterations of, of like this mass outbound assignment, there have
• 28:15
  been times where I thought like, these are good accounts or I
• 28:19
  don't think these are the right data sources or like I think
• 28:22
  we're sending something over to sales that isn't going to work
• 28:26
  out. But I felt like, OK, there's
• 28:28
  already some cost. This is what everyone else feels
• 28:30
  like. It's the right approach.
• 28:32
  So I said nothing and I supported those.
• 28:34
  I did those listed boards. I signed out those accounts and
• 28:37
  those contacts. So I don't know if that really
• 28:39
  answers your question. It's not so much a risk, but
• 28:41
  it's more like my unwillingness to take risks I think absolutely
• 28:45
  has cost the business. That was a great answer and
• 28:48
  Alexander, you have made it to the Lightning round.
• 28:51
  We have a couple quick hitters, so these are really fast.
• 28:54
  Keep us on schedule. The first What is the single
• 28:58
  biggest challenge in B2B sale? Oh my God, this is coming from
• 29:03
  me as an operator. I mean, like personally,
• 29:06
  salesman is really, really hard and I don't know how to do it.
• 29:08
  So that's the biggest challenge, but we've already talked about
• 29:10
  that. For me, I think it's like I, I'm
• 29:13
  still hammering the point home. Like there's a lot of AI noise.
• 29:16
  There's a lot of like potential paradigm shifts and a lot of
• 29:19
  like potential FOMO that folks have.
• 29:21
  I still think the biggest challenges are the same
• 29:25
  challenges that we've always had, which come down to how do
• 29:28
  we get the right information in front of the right person at the
• 29:31
  right time. AI can help us get at the right
• 29:34
  information in front of the right person at the right time.
• 29:38
  We're starting to see some real successes with it, but still the
• 29:41
  biggest challenge is making sure we understand what is the right
• 29:43
  information. It all comes down to the, you
• 29:45
  know, the garbage in, garbage out philosophy.
• 29:48
  Like, I think we're still really struggling to serve our sales
• 29:51
  people in like 1 of the number one ways that we need to serve
• 29:55
  them as OPS people, which is making sure they have everything
• 29:58
  they need at their fingertips so that they can be as successful
• 30:01
  as possible. I, I'm like the technology and
• 30:04
  the ability out there as is insane now, like things that you
• 30:09
  can do, things that you can do instantly, things that a single
• 30:12
  person can set up a single operator.
• 30:15
  But I think at the end of the day, if a salesperson doesn't
• 30:17
  find that information useful or the source data, just, you know,
• 30:21
  you haven't done the unsexy yet necessary work to make sure that
• 30:24
  that is clean and in a good state, you're still going to,
• 30:27
  once again, risk losing that trust.
• 30:29
  Risk sending a bunch of really cool signals that sales doesn't
• 30:32
  trust. Risk saying, like, look at all
• 30:34
  this different intent that's basically just noise.
• 30:36
  Because like, what you really need is someone to sit down and
• 30:39
  ask you what really matters. Can we isolate this to the top
• 30:42
  three or top five? So I think this is just once
• 30:45
  again, a perennial problem. I'm looking at this from a very
• 30:48
  myopic OPS lens, but I think that the potential is greater,
• 30:54
  but the blockers are still the same.
• 30:56
  Love it. What are your favorite tools or
• 30:58
  technology? You've hit on a few already.
• 31:00
  Totally, totally. I mean, like I am very big on
• 31:03
  clay. I think like as I've mentioned
• 31:05
  here and I will say like a little bit about clay, like it
• 31:07
  takes a while to get used to. So if you're out there and
• 31:09
  you're like, you've gotten access to it, you have a free
• 31:11
  version of your company just acquired it and you're like,
• 31:13
  what do I do here? Like I was in the same boat.
• 31:15
  I was skeptical and I just had to spend time playing around.
• 31:19
  But other than that, I think my favorite thing about clay is
• 31:21
  that like I, I got into OPS because I find it creative.
• 31:23
  I find it fun. It's problem Solvy.
• 31:26
  And I think for a while I felt like I was kind of getting away
• 31:29
  from that. Like Clay allows a single
• 31:30
  operator to go in and test things out, play around, have
• 31:33
  fun. Like I said, prototype some
• 31:34
  stuff, get wacky, but then actually some of that might turn
• 31:38
  into, oh, this is valuable information that I'm going to
• 31:40
  hand off to our systems team to make sure we're enriching at
• 31:42
  scale. Clay is absolutely huge.
• 31:45
  Other than that, I would say just like any sort of AI, web
• 31:50
  scraping is extremely important. Like any sort of just Gen.
• 31:53
  AII know that is probably a cop out answer because to be honest,
• 31:56
  my philosophy around tools is there are usually way too many.
• 31:59
  They're we're usually paying too much for tools that aren't
• 32:02
  providing nearly enough value. I am really big on understanding
• 32:05
  like with a few simple like functionalities and like a
• 32:09
  really good clean snowflake database that has everything
• 32:12
  that we absolutely need. How much can we do with very
• 32:15
  little? So I think like for us, the
• 32:17
  ability to scrape websites with a prompt and the ability to just
• 32:20
  overall generate account information with a prompt.
• 32:24
  So a lot of that isn't like you can't buy this off the shelf,
• 32:27
  but I think a lot of the tools out there, you can't buy their
• 32:29
  value off the shelf. Again, I'm an operator, so I'm
• 32:32
  looking at everything with like, what do we need to have in place
• 32:35
  to ensure something like this is successful?
• 32:37
  So I take that approach. That's why like I love Clay
• 32:39
  Gent, which is Clay's AI web scraper.
• 32:41
  We're able to like, hey, does this website say the word peanut
• 32:44
  butter? Great.
• 32:44
  Then they're a good account. Like never been able to do that
• 32:47
  at scale before. But also internally like
• 32:49
  thinking through things like, OK, we've got an awesome data
• 32:51
  infrastructure and data engineering team that has like
• 32:54
  all this product usage and like basically full 360 view of
• 32:57
  anything we want for a customer or a trial and prospect.
• 33:00
  How can we just simply consolidate that and immediately
• 33:03
  create like succinct call prep for every single one of our
• 33:06
  relationship managers or like recent news?
• 33:09
  So like, what is the tool that we would use?
• 33:10
  I don't know one of the GPTS, probably something internally,
• 33:13
  but that's what I'm most excited about.
• 33:15
  Like this new functionality allows us to do so much more
• 33:17
  with what we have and not need. You know, a different tool for
• 33:21
  everything. Awesome.
• 33:23
  We're going to close right on time, so this is 30 seconds or
• 33:25
  less. We call it Goodreads, but it
• 33:27
  could be book, blog, newsletter, podcast.
• 33:29
  What stands out to you? Yeah, absolutely.
• 33:31
  I would say my favorite is probably Kyle Playar's blog,
• 33:34
  just because it's very, it's very down to earth.
• 33:36
  It gives me information that's useful and I think it does the
• 33:38
  best to just help me understand that, you know, like I'm working
• 33:41
  for for intercom, which is I will, I'll pat us on the back
• 33:44
  like we're a successful AI company and we are figuring this
• 33:47
  out. Like the pricing model is new,
• 33:49
  like the go to market is new and we're all we're we're kind of
• 33:53
  trying to benchmark with each other to understand what's the
• 33:55
  right way to do this. So I believe it's called growth
• 33:58
  unhinged. I don't, I might need to Fact
• 34:01
  Check that, but I found that one to be like a very useful, like
• 34:04
  succinct piece of or or newsletter that gets me like,
• 34:08
  oh, that's interesting or like I've got some some ideas from
• 34:10
  that without being overwhelming, without trying to sell me
• 34:12
  anything, without being too abstract.
• 34:14
  Awesome, we have one left. Most important, one of all, Eggs
• 34:17
  Xander. How can everyone get in touch
• 34:19
  with you after this episode? Yeah, absolutely.
• 34:21
  I'm not really on social media. I don't have a presence on X,
• 34:25
  but please find me on the LinkedIn.
• 34:28
  My LinkedIn handle is.linkedin.com/in/a De Mulan,
• 34:32
  my last name ADEMOULIN. Awesome.
• 34:37
  Well, it's been a pleasure. Thank you so much for your time
• 34:39
  today. Absolutely appreciate it, Chris.
• 34:43
  Thanks for joining us for On Base.
• 34:45
  If you enjoyed this episode, please be sure to share, rate,
• 34:48
  review and subscribe on Spotify, Apple, YouTube, or wherever you
• 34:52
  get your podcasts. This show is brought to you by
• 34:54
  Demand Base, the only account based experience platform built
• 34:58
  for sales and marketing alignment.