Build a buyer intent engine that actually works

(0:00) James: People say, oh well, you know, we've got buyer intent in place. But what they typically mean is they've got some form of lead scoring or they've basically just turned on buyer intent features in something like HubSpot. There's still a lot of skepticism around buyer intent. What we're talking about is building a buyer intent engine. That's a completely different thing to simply having it turned on. Typically we see improvements up to 2 times what you get from a non-intent based demand gen program.

(0:26) Craig: So welcome to this episode of Truth, Lies and B2B Growth. Today we're gonna talk about buyer intent. So we're gonna talk about what a real buyer intent engine looks like. We're gonna go through the common problems of buyer intent. Then we're gonna give you five components of an effective intent engine.

(0:39) James: Yeah, and we'll be honest with you as well, it does require engineering, it does require time, it does require iteration, and what intent is not is SQLs — they are beyond a typical cold lead but they're not yet always ready for sales, and we'll talk about that a little bit more later. So let's get into the common problems with buyer intent. Buyer intent's been around for a long time as we all know, there's a lot of tools that offer the ability to identify companies that are showing intent through a variety of different signals, but for many, that's probably as far as it goes. Intent signals are quite limited and therefore can create false positives. For example, you might just have one signal based on research that points to an organization potentially being interested in a certain category. But that's just relying on one particular source, so it's not always reliable and it could mean several different things rather than actually a buyer being in an active position.

(1:34) Craig: The main problem is that people confuse intent with SQLs, and if we look at what intent really is — we'll talk about the three different types of buyer intent a bit later — what they are is clues that someone may be within an active sales process. What they are not, though, is absolute guarantees that somebody is within that buying process. Really they're akin to something a little bit beyond the traditional MQL. They're nowhere near yet at the level that an SQL might be determined. So the core problem is people assume that, and therefore they take action on the back of intent being received, such as sending it straight to sales and asking them to pick up the phone.

(2:10) James: And what happens then is sales pick up the phone as they're asked to do, the leads go nowhere because they're not far enough down the line, and then obviously trust is then impacted in the data.

(2:22) Craig: Intent doesn't work, scrap it. We'll go back to the old model, thanks.

(2:28) James: It's not that intent doesn't work though, it's that if you treat it like an SQL, you're gonna be sending it to sales too prematurely, and you're gonna erode trust. Another mistake that's linked to that is when you're driving signals that are based on potentially one person, which is typically through the engagement signals. When, as we know, in complex sales most decisions — if not all decisions — are made through a larger buying committee. Gartner or Forrester did some research to say that 82% of B2B sales — never mind in complex sales, I'm assuming the number is even higher — of deals happen through a buying group, i.e. multiple people within a particular business. So your intent engine — a good intent engine — will factor in multi-threading.

(3:14) Craig: There are legitimate reasons you might want to focus on particular individuals within an intent tool or an intent engine, and it may be that, despite the fact that 82% of B2B buying journeys are taken by buying groups, you may be in the 18% that is not. You may also have a lower threshold for when you want SDR outreach to happen. I don't think it's a black and white thing that you should only ever act on intent when it's multiple people within a business, but I think what we have to be clear about is that only when you are able to gather signals of multiple people engaging with a particular target account should you be really clear that there is probably an opportunity there. I think it's probably a good point to talk about what those intent signals are and explain a little bit about the differences between the different types of intent signal. The first is engagement intent, the second is research intent, and the third — what we call triggers or signal intent — and we'll get into each of those and take them one by one. So the first is engagement intent, do you wanna talk that one through, James?

(4:22) James: Yeah, and that's the one that most people will be familiar with, whether they call it intent or not — it's engagement with things that they own, so first party data, people that are engaging with your website, with your emails, with your various marketing activities and platforms that you own, and that's typically the easiest one for people to get their hands on. It's also, by the way, one of the more suggestive of intent of the three, because ultimately the familiarity barrier has already been crossed. A lot of people may have previously had lead scoring systems set up which was pretty much based on the same thinking, but examples of engagement are website visits, interactions with content — but really it goes beyond that to platforms like LinkedIn, which is obviously a key channel now from an outreach and demand generation perspective. So there we're looking at things like profile views, comments on posts, likes, company page views, etc. So if you can try and get out all that information, you start to get a really good picture of the companies and contacts that are engaging with you. So let's move on to research then, the second type of intent. You want to take that one away?

(5:30) Craig: Yeah, absolutely, and I explained that a little bit earlier, didn't I? Ultimately that is a way of understanding if an organization is interested in researching topics which are relevant to you as a business. So an example would be for us — we're obviously a marketing agency focused on B2B service-based organizations with complex sales. If we were to set up research, and of course we have done research intent, then we would put topics in there such as B2B marketing services, B2B marketing agencies, B2B complex sales services — whatever it may be — because we know those are the kind of things that people are searching for that make them relevant for us to reach out to them, because we can help people that are interested in those topics. What that doesn't do is ensure that that organization knows anything about who we are as a business, so this is where we've really got to take research with a bit of a pinch of salt and treat it for what it is, which is someone researching a particular topic. And then the third type of intent are what we call triggers, which are typically signals that might suggest that an organization is in a position where they may be appropriate for you to reach out to, based on your proposition being a timely solution to a particular trigger — so it could be a category entry point. For us a good example is often we see a lot of change happening when a new CMO comes into post, so that's quite a common one to track — new hires joining an organization. But again, I think to temper a little bit of the excitement around them, we've got to be clear about what they're really saying, and I've spoken to various business leaders recently — they're getting really excited about this intent stuff, and ultimately just because someone has hired a new CMO doesn't mean that they're absolutely gonna invest further in marketing or rip out the old guard and replace it in our case. So there's that barrier to overcome. It doesn't mean it's definitely gonna happen, and B — we've still got to educate them on who we are against probably a list of other agencies they've worked with in the past. So there's still work to be done, and I think we've got to be really clear that an intent signal is the starting point for you to begin engaging with someone at the right time, or potentially at the right time — not the end point whereby you're gonna immediately book deals on the back of it all.

(8:04) James: Yeah, they're like breadcrumbs, aren't they — they're tiny signals, clues to suggest that someone might be about to enter a market or already be active in a market. So I think then, if we think about how you can build an intent engine which actually does work, then we should look at some of the challenges that we see in people that have supposedly set these things up already. Now the first — you know, as you pointed out at the start, Craig — is people say, oh well, you know, we've got buyer intent in place. But what they typically mean is they've got some form of lead scoring, or they've basically just turned on buyer intent features in something like HubSpot. Now of course that is technically true, that they have buyer intent in place, but what we're talking about is building a buyer intent engine. And that's a completely different thing to simply having it turned on. A buyer intent engine considers multiple sources of intent and brings them together so that you can triangulate those sources and have a much stronger signal that somebody is likely — or more likely than most — to be engaged in some form of buying process for the kinds of things that you're interested in. So building in multiple sources of intent is the first one.

(9:01) Craig: That might mean HubSpot plus LinkedIn signals plus others — whatever it is, the more intent sources you've got, the more likely you're gonna be able to identify strong signals.

(9:12) James: Yeah, and it's important that you carefully think through the best types of signals to focus on, and then that will inform the tools and investments you make to make sure that you've got the right tech stack in place.

(9:26) Craig: Yeah, and then what I would recommend there is prioritizing where you spend most of your effort in terms of marketing activity. If your strategy is very heavily focused on generating engagement on LinkedIn, then it's absolutely crucial that you track intent on LinkedIn. If you don't use LinkedIn at all — which obviously would be questionable — then there's not much point in investing in tracking intent on LinkedIn, because you're unlikely to be generating much of it.

(10:00) James: Yeah, so start with the hypothesis, and then start to think about the best sources of those signals that are gonna have the most impact for you. The second important pillar is ICP filtering. And I can imagine people saying, yeah, well, OK, we've got that. In HubSpot's case, it's a case of saying, well, these are the job roles, and these are the companies, and these are the size of companies and the industries that we're interested in.

(10:16) Craig: Now that is a good starting point, and obviously having that as refined as possible is important, so that you're not only looking at the people that are showing intent — you're looking at the relevant people that are showing intent. But in our view there's a layer of further validation that needs to happen there to ensure that is the case, because anyone that's worked with data will know that it's not always gonna give us exactly the right kind of ICP match just based on data points alone. For example, we work with cybersecurity companies, and if you put "security" into any kind of data platform, you're gonna get cybersecurity companies — you're also gonna get people that man doors at nightclubs. So pure firmographic data alone isn't enough to do ICP matching.

(11:00) James: We would recommend bringing in more intelligent tools, such as some of the LLMs or AI, to do a more thorough job of ICP matching, both on the organization level and also based on the persona level.

(11:14) Craig: Yeah, and it's far better to keep the ICP really, really tightly focused, so you're not pulling in loads of accounts that really aren't relevant to you — that's another kind of mistake we see. It's far better to have a smaller group of better-fit accounts than lots of accounts that just don't fit your ICP whatsoever.

(11:34) James: Connected to that is component number 3, and that's around enrichment. So a lot of the tools out there give you company-based data around whether an organization is showing visitor intent, research intent, and of course that isn't enough then to take action on it. The amount of times we've had conversations with people that say, yeah, we've got intent set up but I don't know what to do with it — and that's half the battle. So ensuring that there is some form of automated enrichment happening — that means when an intent signal is received, something is going away and finding the relevant data against the right ICP and the right persona, following that back into your CRM — is a really important step, because without that contact data, no action can happen. Tools like Apollo are really useful there, but it's a really important, critical step that you're enriching those intent signals at an organization level with the appropriate contact-level data, so that you can take action on it.

(12:24) Craig: Component four is activation. As we just talked about, enrichment is a critical step in being able to activate any outbound activity at all, but you need to be able to connect the intent signals that you receive with some type of outbound activity. For us that might be using HubSpot's prospecting agent, it might be using LinkedIn automation tools such as Heyreach or others. It may be connecting with a third-party email platform like Lemlist or Instantly, to ensure that you've got that — you're connecting the dots between when intent is received and some outbound activity happening. And there are multiple routes to making that happen.

(13:05) James: So I've had conversations with clients recently that have fairly sizeable SDR teams, and the first question is, OK, well, where does the SDR team fit into this?

(13:19) Craig: And there are various flavors of how you can make it work. You can make it work in a fully automated model, so that when intent is received, outbound automatically happens, or you can think about it in a more hybrid way. An example of how that could be done — an intent signal could be received, it could be checked against the list of target accounts. If it meets the target account criteria, it could be followed up by the SDR. If it doesn't meet that criteria, and they're still a target, but there might be a third-tier target, then some of that multi-channel automated outreach could kick in. But whatever it is, you need to ensure that there is some method — be it through automation, through AI agents, or through a really robust, established process with your SDR team — that people are following up on that intent which is received.

(14:07) James: Yeah, now we're in a world where, to be effective with outbound, really requires a real tightly coordinated, integrated approach through various different channels, all timed in the right way and all using the prioritization that intent gives you to reach out at the right time and on the right platform. And that's so key — and obviously intent also gives you the ability to personalize a lot more, and it gives you the ability to scale through automation.

(14:37) Craig: Which, at a time where it can take an incredible amount of heavy lifting to do all of this manually, really what it enables your sales team to do is, hopefully, minimize the amount of basic administration, data collection, account research, and really hopefully get on with the main task at hand, which is speaking to people in discovery calls and moving along sales opportunities, rather than focusing all their efforts on early-stage prospecting.

(15:09) James: So that probably leads us on to component number 5 — tracking and reporting. I think this goes for any kind of marketing effort, doesn't it? Unless you can prove the return on it, then—

(15:19) Craig: —it's unlikely to stay in your strategy for too long, and certainly the funding provided for it is likely to be stripped pretty quickly if you can't prove return on investment.

(15:28) James: Now, the good news with intent, unlike some other things like brand, for example, is that it's fairly easy to connect the dots between the intent signals which have been received, why they've been received, what outreach happened as a result, and whether those people are associated to deals in your pipeline. So it's a fairly simple one to be able to prove ROI, based upon — from a technical level — the KPIs that you're looking to achieve being considered and built into that reporting upfront, and the expectations being set with the appropriate stakeholders at the right time.

(16:02) Craig: I think it's important to add that when we've seen like-for-like comparisons between a demand gen program that is running without intent versus one that is, the one that is we're seeing sort of almost twice the effectiveness, and also a significant reduction in the amount of administration heavy lifting that the sales and marketing team needs to do once everything's set up and running. And it isn't always on engine — it's really about timing, and that's really everything today when it comes to demand gen. So what are the tools that you need for the job? I think, firstly and critically, you obviously need some form of CRM, because that's really where all of your data is gonna sit, but it's far from the only tool that you're gonna need in the engine stack. So what else would you add in there, James?

(16:52) James: Well, as we've talked about, there's a really important element in this — you need to be able to activate marketing activity or outbound prospecting, whatever you wanna call it, when a signal is received. So you need a platform that enables you to receive those intent signals, either through integration or embedded within — such as in HubSpot's case — and automatically trigger actions on the back of that, whether it's a manual reminder for a salesperson to follow it up, or enrollment in a prospecting agent or something similar. We know that single-channel outreach — say through email, whether it's through a prospecting agent or otherwise — is not normally enough to really make this as effective as it could be. So, like we talked about earlier, LinkedIn automation tools like Heyreach or others, other email platforms like Instantly, can be really valuable add-ons to an intent engine. And the obvious other one is enrichment — so you need a platform like Apollo, like ZoomInfo, like Cognism — as long as they integrate well with automation tools, they'll be suitable enough for you to go with. Full Enrich is another good one. Whatever you use, as long as it's able to neatly integrate into the process and the automation. But the next one, I think, is the really critical one, and what a lot of people are missing — they might have access to all these different tools, but they just haven't got the orchestration and engineering in place to bring all of them together in one place, which is where you get the real gold. So there's that layer of middleware and orchestration that is needed as well, so obviously there are various tools now that enable you to do that.

(18:41) Craig: Well, it's why Clay exists, isn't it — and for those that can afford Clay, because Clay is an expensive way to achieve this, it really enables that orchestration. You can plug all these things in and manage it all in one place with Clay. But for those — which is the majority — that aren't able to invest that level of funds into something like Clay, you need an orchestration layer that helps you connect all of those different pieces, and importantly, audit what's happening within them.

(19:11) James: So it may be, you know, using something like Google Sheets even, if you want to do it that way, or something that's ingesting data between the different points of integration, so that you can see — say, for example, when a trigger has occurred — what data has been pushed from one system to another, and audit, if there is an issue, why that is. If you completely leave the system to work as it does, then effectively you've got an intent signal on one end and data spat out on the other — it's very, very difficult then to determine what's happened in between. And you need an orchestration layer to make that happen, because there will be issues, there will be things you will need to investigate, and you simply won't be able to do that without that orchestration layer in place.

(20:00) Craig: Yeah, and it's not a case of just buying as many tools as you can and throwing a lot of money at getting all the different tools. If you can't orchestrate them, you're never gonna have a really sophisticated, effective buyer intent engine. Clay can be a barrier to many because of the cost, but there are ways to stitch together the various different tools using forms of AI, such as Claude, n8n, using platforms such as Airtable — all of this can be pulled together in a very effective way that almost mirrors what can be achieved with Clay.

(20:30) James: The cost-benefit analysis when it comes to Clay versus building it yourself is the level of engineering involved in making it happen. It is complex — I don't think we're trying to skirt around that — and it does require a level of engineering capability which doesn't exist in most marketing teams.

(20:48) Craig: In terms of where to start, I think the answer is not to try and boil the ocean. I think start with the foundation — so it may be that you start with HubSpot, and you start by simply beginning to track the signals that you get through the intent tool in HubSpot, if you use it, or whatever platform that you use. It's good to get that in place and start to maybe manually check it regularly, to see what kind of data is being pulled through. Is it right? Are the research topics that you've established bringing through way too many contacts for a particular topic which isn't maybe as intent-driven? Get that set up first, get used to it, refine it — because that intent tool is gonna be the vast majority of the source of what happens next. If that is wrong, and you've got the ICP filtering wrong, you've got the research topics wrong, then ultimately that's gonna funnel through to outbound outreach which is completely irrelevant. So getting that set up first, I think, is a key foundation.

(21:44) James: Then connecting the automation to that in an MVP style way. So once you've got that set up, connecting that to some form of outbound activity via enrichment is the key next step, so that you've got that fully end-to-end loop of intent received, company enriched, outbound happening in a single channel. Once you've got that, then you need to start to think about bringing different intent signals in, to further enrich the kind of intent data that you have available at source.

(22:17) Craig: So for me, the key takeaways are: there's still a lot of skepticism around buyer intent, and understandably so, because most programs — most buyer intent setups — are really only doing half a job. People feel just setting up an intent engine in a tool like HubSpot is enough. If you're gonna commit to a buyer intent process, or trying to build an engine, you need to try and cover all the key components that we've talked about in this video.

(22:43) James: The second key takeaway is: don't focus on one signal. You may want to layer signals in as you build beyond MVP stage, but certainly the final destination should not rely on a single signal of intent data. Third takeaway is: the buying group is the unit, not the lead — so you need to make sure that you're not just focusing your intent on one particular contact, and then basing your outreach on a very thin signal. Make sure that you're trying to factor in, as much as possible, the whole buying group. And lastly, a key takeaway would be that you really need to make sure that you're measuring this — it's really, really key that you line up a reporting mechanism that shows your activity led by intent versus that that's not. And no doubt, as we said earlier, you'll see a difference — typically we see improvements up to 2 times what you get from a non-intent-based demand gen program.

(23:38) Craig: So thanks for listening or watching. If you enjoyed the episode, then please give us a like, add in your comments, and we'd really appreciate it if you could subscribe to our channel. We're pushing out plenty more content, so please check it out whenever you get the chance on your preferred platform, and we'll see you again next time.