Latest / Sales Pipeline Radio / AI-Native Demand Generation Best Practices

Transcript

AI-Native Demand Generation Best Practices · Sales Pipeline Radio

• 0:17
  Matt Heinz: All right, well happy Thursday everybody.
• 0:19
  Welcome to another episode of Sales Pipeline Radio.
• 0:21
  I'm your host, Matt Heinz.
• 0:22
  Excited to have you all here in the middle of your workday, in
• 0:25
  the middle of your work week.
• 0:27
  We are somehow, Alex, into late September now as we record this, which is nuts.
• 0:32
  But here we are so excited to have you here.
• 0:34
  If you're joining us live on Sales Pipeline Radio today,
• 0:37
  we're excited to have you here.
• 0:39
  If you wanna be part of the show, make a comment, ask a question, make a rebuttal.
• 0:44
  Any of those are fair game, we will see those.
• 0:47
  We can respond to those.
• 0:48
  We can bring your comment or question up on air, onto the screen.
• 0:51
  So we'll make you part of the show.
• 0:52
  And thank you for doing so.
• 0:53
  If you're watching or listening on demand.
• 0:55
  Thank you so much for being part of the Sales Pipeline Radio community.
• 0:59
  Every episode is available, past, present, and future at Sales Pipeline Radio dot
• 1:03
  com, and very excited to have with us today, the founder and CEO of SalesboxAI.
• 1:09
  Alex Roy, Alex, thanks very much for joining us today.
• 1:12
  Alex Roy: Hello.
• 1:13
  Nice to be here.
• 1:13
  Excited to be here, Matt.
• 1:15
  Matt Heinz: All right.
• 1:15
  So, when we talked earlier and you brought up this idea of AI native
• 1:21
  demand generation, I don't think I'd heard anyone describe that before.
• 1:25
  We've seen and worked with a lot of companies that are heavily incorporating
• 1:28
  AI into various go to market motions.
• 1:31
  But I'm curious, maybe let's just start with just an introduction yourself and
• 1:33
  like, what is AI Native Demand Gen.
• 1:37
  Alex Roy: Sure.
• 1:37
  Yeah.
• 1:38
  it's about, you know, AI has been around for several years and, when we talk about
• 1:45
  AI native Demand gen, it's about having an AI agent that is ingesting signals,
• 1:52
  realtime signals and making decisions.
• 1:55
  And, making sure that it is orchestrating other agents to make
• 1:59
  sure that, the right person gets the right message at the right time.
• 2:03
  And, yeah, so it's, it's integrated.
• 2:06
  You know, you don't have different silos of, systems in place.
• 2:09
  So the data is available as well as the signals go to an AI agent
• 2:14
  that takes those decisions.
• 2:17
  Matt Heinz: So how, I mean for those that are thinking about incorporating
• 2:20
  this, like how is this different than just sort of traditional demand gen?
• 2:23
  Is this new or is this an evolution?
• 2:25
  How should people think about it?
• 2:27
  Alex Roy: Sure.
• 2:28
  Yeah.
• 2:28
  So, I would say in the, in the past, you know, the, the fundamentals
• 2:32
  still remain the same, right?
• 2:34
  But what's different is that, probably, you know, and end of
• 2:38
  the day we have to stay human.
• 2:40
  We have to stay, you know, make sure we each account and each person and
• 2:44
  each contact has the right experience.
• 2:46
  But unfortunately, we couldn't do that at scale.
• 2:49
  We had to make sure that, you know, we had to kind of divide it into personas
• 2:53
  and, you know, probably just look at a few, you know, maybe a handful of them and
• 2:58
  have one set of message for each of those.
• 3:01
  But with an AI agent, what you are able to do now is that give each
• 3:05
  account and each contact the right attention that, that they truly deserve.
• 3:10
  So it can be orchestrated at scale and helps you, make sure
• 3:15
  that you know, you are earning the trust of that buying group.
• 3:19
  Matt Heinz: You know, is the complexity of go-to market motions and complexity
• 3:22
  of buying journeys continue to grow.
• 3:24
  You know, that we're also seeing, of course, board members and
• 3:27
  investors saying, do more with less,
• 3:29
  be more efficient.
• 3:30
  And yet the motions
• 3:31
  Matt Heinz: we're trying to execute often are really
• 3:33
  struggling with speed to market and
• 3:35
  efficiency.
• 3:35
  Right?
• 3:35
  Can you talk about
• 3:37
  Matt Heinz: how AI can really directly address that?
• 3:39
  Not in a future state, but right now?
• 3:42
  Alex Roy: Sure.
• 3:42
  Yeah.
• 3:43
  So, bandwidth is usually the biggest constraint that,
• 3:47
  that we have as marketers.
• 3:49
  And, you know, an AI agent helps, you know, gives you that scale wherein
• 3:54
  you are able to be more efficient.
• 3:56
  'cause now you have the agent take on some of the laborious tasks.
• 4:01
  Be it, you know, from signals, you know, you have thousands of signals.
• 4:04
  But, you know, just distilling down that into, you know, the right ones you
• 4:10
  know the, the noise out of the signals, that's where the agent helps you.
• 4:14
  So it makes your overall more efficient, in terms of that.
• 4:17
  Matt Heinz: And what does that mean in terms of like realtime intent signals?
• 4:21
  I think, you know, I think we've had signals around for a while.
• 4:25
  How does AI, how is AI different?
• 4:27
  How is it different in terms of speed, in terms of context, in
• 4:30
  terms of, you know, everything.
• 4:33
  Alex Roy: Yeah.
• 4:34
  So, when it comes to intent signals, the key is time, because it's
• 4:37
  ephemeral, you know, it doesn't, you have to act on it quickly.
• 4:41
  And, a good
• 4:41
  example of an intent signal is somebody showing up on your website,
• 4:45
  Alex Roy: and there's thousands of visitors each month that a company has.
• 4:50
  And, these signals needs to be, inferenced.
• 4:53
  And, you know, you have to identify which account it is, which person it is, and
• 4:57
  then have personalized conversations.
• 4:59
  So an AI agent helps you facilitate that.
• 5:03
  Matt Heinz: So walk me through an example of that.
• 5:05
  Right.
• 5:05
  So like, let's say I'm an, let's say I'm an inside sales rep. I'm A
• 5:07
  BDR, You know, is this following up on leads I already have?
• 5:11
  Is it identifying new leads I should engage with?
• 5:13
  Is it a little bit of both?
• 5:15
  Alex Roy: Yeah, it's a great question, right?
• 5:17
  So typically what you have in an organization is, you know, there's
• 5:20
  a sales focus set of accounts that, you know, sales is looking at.
• 5:23
  There is a wider set of accounts that marketing is looking at, and then there
• 5:27
  are these white space accounts, right?
• 5:29
  So, what the agent is doing is, you know, it's listening for signals and
• 5:33
  first party signals is very important.
• 5:35
  And as you know, and off the website, it listens for that and, it tries to
• 5:40
  find, you know, does that match your ICP?
• 5:42
  Is that part of a sales focus account, or is that a marketing focus account and
• 5:47
  then it hands it over to another agent that is going to have that communication.
• 5:51
  So for example, if it were a sales focus account, you know, it's
• 5:55
  going to be having a play which is on behalf of the sales rep.
• 5:59
  And, the goal is to, you know, to get to a meeting.
• 6:03
  And whereas if it's a marketing focus, it's a wider account, you know,
• 6:06
  it, it goes into a nurture stream.
• 6:08
  So it's not just a single agent, but starts from a signal, but then you know
• 6:13
  the next best action is handed off to another agent, depending on, you know,
• 6:18
  how relevant it is within that ICP.
• 6:21
  Matt Heinz: Talking today on Sales Pipeline Radio with Alex Roy, he's
• 6:24
  the CEO and founder of SalesboxAI.
• 6:27
  It's one thing to talk about what the technology can do.
• 6:29
  It's another thing, Alex and you know, directly to, to get people
• 6:32
  to actually do this at scale.
• 6:35
  From your work, what do you see as some of the biggest barriers to
• 6:38
  getting sales and marketing teams to not just experiment with AI?
• 6:42
  This isn't thinking about this not just as a tool, but as infrastructure, as
• 6:46
  part of a consistent, complex playbook.
• 6:49
  What are some of the barriers keeping people from doing that?
• 6:52
  Alex Roy: I mean, you know, there's, there's the number one,
• 6:54
  I would say, resistance to change.
• 6:56
  There is a bit of an unknown.
• 6:58
  You know, so it's about taking that first step.
• 7:00
  You know, it's about, we always recommend to our customers to list down the tasks,
• 7:06
  you know, that's part of, you know, your Go to Market that you do today,
• 7:10
  and then, you know, the first step is to make sure, identify the task from
• 7:15
  that list, which could be automated, that could be, you know, where you
• 7:19
  could have assistance of an AI agent.
• 7:22
  And, what the agent really is doing is helping you to get to more detail than
• 7:27
  you could possibly have done earlier.
• 7:29
  So, probably, you know, it's not like it's taking off just, it's doing more work
• 7:35
  than what you were able to do earlier.
• 7:38
  Matt Heinz: Yeah, and I think, is there a phased work program you recommend people?
• 7:42
  I think if you're thinking about doing AI native demand gen, I mean
• 7:46
  oftentimes if we're doing a test with a client or sort of recommending it's
• 7:49
  not just launch everything right away, do it in a particular market, do it
• 7:52
  a particular industry, pick a segment of the sales team to start it with
• 7:56
  kind of Is there a phased approach?
• 7:57
  You recommend people think about that?
• 7:59
  Alex Roy: Absolutely.
• 7:59
  Absolutely.
• 8:00
  Yes.
• 8:01
  Yep.
• 8:01
  Yep.
• 8:02
  So we recommend, you know, it starts with signals.
• 8:05
  So, you know, we recommend a subset, a specific product page on your
• 8:09
  website where you add a tag so that the agent is listening for the
• 8:13
  intent on that specific set of pages.
• 8:15
  It's probably, you know, perhaps product centric.
• 8:18
  And, you know, so you have, you know, you look to have the agent identify who
• 8:24
  is on the website, which company is there, and then orchestrate plays based on that.
• 8:28
  So.
• 8:29
  Yes.
• 8:29
  A phased approach is the always the recommended way to proceed.
• 8:35
  Matt Heinz: I've seen some reports already, of, you know, like Forrester
• 8:38
  for instance is saying that they might be seeing some budget cuts in
• 8:41
  2026, which I think might be just sort of exhausting to people that have
• 8:45
  been through what they feel like has been two, three years of doing this.
• 8:48
  I feel like we're getting to this point where this is not just optional, right?
• 8:53
  That adopting these programs is kind of mandatory if you want to survive.
• 8:58
  And I think, you know, can you talk about not just the program implications you
• 9:03
  mentioned earlier just the culture change is required internally to make this move.
• 9:08
  What are you seeing from CMOs and CROs that are operationalizing
• 9:12
  this quickly, but also managing the human element behind it?
• 9:17
  Alex Roy: Yeah.
• 9:17
  So, you know, budgets is a challenge as like, you're, like, you rightly
• 9:20
  pointed out, and it, it's forcing teams to be more efficient.
• 9:24
  It's about unifying the MarTech stack that you have.
• 9:27
  It's about, you know, making sure that, you know, there is, you know, you are.
• 9:33
  You are, you know, there's, there's less inefficient systems in, you
• 9:37
  know, within that MarTech stack.
• 9:38
  So yeah, so we work with them in a phased approach, like I mentioned
• 9:41
  earlier, and it starts with, you know, a specific use case.
• 9:45
  And usually that's about, you know, the inbound leads that you know that's
• 9:49
  coming in onto your website and, you know, how can an agent help you?
• 9:53
  There is where we start.
• 9:56
  Matt Heinz: So I think about this, I mean, we've been talking sort of
• 9:58
  almost with the assumption that this is a sort of a direct sales engagement,
• 10:01
  but it strikes me that this can impact partner and channel sales as well.
• 10:05
  Have you seen those use cases?
• 10:06
  Alex Roy: Yeah.
• 10:07
  So, one of the challenges which, which companies have is that the partner has,
• 10:12
  you know, the, the product knowledge that the partner has is limited.
• 10:15
  They have multiple products they're looking at also.
• 10:18
  So, you know, an agent that is knowledgeable makes the partner more
• 10:23
  knowledgeable indirectly as well.
• 10:24
  So the use cases where, you know, we give an agent to the partner that,
• 10:28
  that's grounded to reality by, by the product collateral, which is available.
• 10:34
  So the partner readily is able to access that information
• 10:39
  and, be more, knowledgeable when presenting that to customers.
• 10:44
  Matt Heinz: Can you help with a little expectation setting
• 10:47
  for programs like this?
• 10:49
  You know, whether you're taking a phased approach or implementing
• 10:51
  a technology like this.
• 10:52
  Like what, how much time does that take?
• 10:54
  How many people does it take to implement it?
• 10:56
  How much time does it take to implement?
• 10:58
  And then what should people's expectations be about how and
• 11:01
  when it starts to deliver results?
• 11:03
  Alex Roy: Yeah, I mean, you know, in terms of rollouts, typically we see like two
• 11:06
  weeks to four weeks, somewhere in between that in terms of timelines and, part of
• 11:11
  it also has to do with, you know, there's some back and forth with, with the
• 11:15
  marketing team and the IT team as well.
• 11:17
  But yeah, but generally we see about two weeks to two to four weeks.
• 11:21
  Matt Heinz: And then what... I love asking the question about doing a
• 11:24
  sort of a pre postmortem, right?
• 11:26
  Like if we were, say, if someone implemented sort of these programs
• 11:29
  today, if we're talking about three months from now and they have not
• 11:33
  been as successful as people want.
• 11:35
  If you do the postmortem, then what are the things you hear most often?
• 11:39
  Like, what are the things that most often keep a program like
• 11:42
  this from being successful?
• 11:44
  That people that are leaning in and saying like, I need to adopt this, should be
• 11:46
  aware of and prepare for now, so that they increase their chances of success.
• 11:50
  Alex Roy: Sure.
• 11:50
  Yeah.
• 11:51
  So for example, let's say, you know, the KPI is where we wanted to generate more
• 11:54
  intent leads and content leads, right?
• 11:56
  So intent leads is visitors that you convert into leads.
• 12:00
  Web visitors.
• 12:02
  So you have to commit, you know, sending the right audience to those pages as well.
• 12:07
  So let's say you, it's like setting up a mousetrap, but then
• 12:10
  you know, you're not driving, you know, the right folks there, right?
• 12:15
  So, yeah, so it's about, you know, making sure that you also have a marketing
• 12:19
  program to drive the right audience to those pages, and then the agent is
• 12:23
  able to, engage with them, identify them, and then engage with them.
• 12:27
  So that's, you know, that's number one.
• 12:29
  The other is another metric you look at is, so we have an integrated asset hub.
• 12:35
  You know, you have an asset hub where the agent is knowledgeable
• 12:40
  about the assets, so it's able to, so another KPI we typically look
• 12:45
  at is, you know, how many marketing qualified leads were you organically
• 12:49
  able to generate, and, you know, how many conversations were created.
• 12:53
  So those broadly I would say, you know, are the three key metrics, that
• 12:58
  would come out of this as an outcome.
• 13:04
  Matt Heinz: I am curious to hear from your perspective, this is a little
• 13:06
  take a different approach, just the speed of innovation happening in an AI
• 13:12
  forward company right now versus sort of traditional growth and development cycles.
• 13:16
  I mean, from the outside it seems like things are moving incredibly fast, and I
• 13:21
  can only imagine that growing a company in this space is controlled chaos.
• 13:26
  What's it like on the inside, sort of riding this enormous tidal wave right now?
• 13:32
  Alex Roy: Yeah, it's a great question.
• 13:33
  I mean, over the last 10 years, you know, we've you know, gone through
• 13:36
  multiple eras of evolution, I would say started back in the days with, you
• 13:43
  know, you have classification models, you know, it's neural network based,
• 13:47
  and now you have the LLM based model.
• 13:49
  So it's about evolving like you mentioned, and, being engineered to evolve.
• 13:53
  Right?
• 13:53
  So, which is very important.
• 13:54
  So one of the examples is, we look at specific use cases and which is the right
• 14:01
  model for that, is that, should that be using a fine tuned model that we have
• 14:05
  versus, you know, for this specific task, you know, there's another model to use.
• 14:09
  So it's about plugging in the right models for the right use case, I would say.
• 14:14
  And yeah.
• 14:15
  Matt Heinz: And I think, you know, I'm thinking about integration
• 14:18
  of these programs as well.
• 14:19
  Like the volume of tools that marketing operations folks at RevTech, you know,
• 14:24
  people are managing is significant.
• 14:26
  I'm hoping that you are gonna help us build and you envision a day when we may
• 14:31
  have separate tools, but AI actually helps them operate and integrate and communicate
• 14:34
  with each other a little more seamlessly.
• 14:36
  And I'm hoping that that actually solves what many people have as a
• 14:39
  data management problem as well.
• 14:41
  Which inhibits AI from oftentimes doing what it needs to do.
• 14:45
  How do you envision the future of the tools, the data,
• 14:49
  the integrations happening?
• 14:51
  Alex Roy: Yeah, it's a great question.
• 14:53
  You know, like you rightly mentioned for AI, you know, you need to make sure
• 14:56
  the data is accurate, that, you know, it works on the right data, right?
• 15:00
  And in terms of tools, also there are protocols like MCP, which vendors are
• 15:05
  using to enable, the existing agent with additional skills you could say.
• 15:12
  So give it additional context.
• 15:13
  So, it's about, you know, you would see each of the, product in the
• 15:18
  MarTech stack has this connector back to this agent and, you know,
• 15:22
  enables the agent to, to be more, more productive across different systems.
• 15:27
  And, at the same time, IT needs to be able to set certain ground rules,
• 15:30
  access controls, et cetera, also.
• 15:32
  So it's almost an additional, person within the organization that's,
• 15:37
  utilizing these, permissions and making good things happen.
• 15:43
  Matt Heinz: I think maybe more than any technology we've seen before, AI
• 15:47
  has democratized software development.
• 15:50
  Right?
• 15:50
  I mean, the vibe coding is real.
• 15:52
  You can talk to a machine and it will build something for you.
• 15:55
  I think that is changing how many people think about managing
• 15:59
  their own technical expertise.
• 16:01
  As business leaders, what advice would you give someone sort of new, you
• 16:07
  know, sort of recent college grad, who's trying to figure out what skills
• 16:11
  they need to have, who in the past, unless they were a developer, didn't
• 16:13
  worry about needing to build code.
• 16:16
  What advice do you give to that up and coming business leader now?
• 16:23
  I was building to my theme there, and I think I lost Alex, did we lose Alex
• 16:28
  right before my last question to him?
• 16:31
  Oh no, we did lose Alex.
• 16:32
  Well, anyway, thank you Alex, for joining us today on Sales Pipeline Radio.
• 16:37
  Really appreciate everyone joining and listening to us as well.
• 16:40
  Have a great week.
• 16:40
  We'll see you next week on another episode of Sales Pipeline Radio.
• 16:43
  Until that'll be well, take care.