About this episode
Eddie Reynolds, a revenue operations expert and co-founder of Union Square Consulting, breaks down how to transform your Ideal Customer Profile from a static document into a continuously improving operational system. Reynolds brings deep expertise from his time at Salesforce and years of consulting with B2B revenue leaders navigating the gap between pipeline activity and actual deal conversion.
The core problem: most ICPs are too broad, too shallow, and completely disconnected from how teams actually qualify prospects and close deals. Companies often find themselves with full pipelines, healthy-looking activity metrics, and strong lead flow—yet deals slip, cycles extend, and customers churn. The root cause is almost always an ICP definition that sits gathering dust in a shared drive, lacking the operational teeth to drive daily go-to-market decisions.
The reframe here is that defining your ICP isn't an annual planning exercise—it's identifying which customers are genuinely easiest and most profitable to win, retain, and expand. This requires moving beyond firmographics (industry, revenue, headcount, geography) into the specific, often counterintuitive data points that actually separate your best customers from the rest. Reynolds walks through how to spot survivorship bias in this analysis, why the biggest or most numerous customers aren't always your best, and how to validate each hypothesis before you start shifting tens of millions in marketing and sales spend.
The episode covers how to build account scoring models reps will actually trust, how to operationalize ICP across sales, marketing, and CS, and most importantly, how to architect a continuous improvement engine that feeds every new lead, lost deal, and churned customer back into your definition—keeping your ICP sharp and aligned to reality as your product, market, and team evolve.
What we cover in this episode
-
1:02
The Pipeline Paradox Activity metrics look healthy but conversion stalls—deals slip, cycles extend, customers churn out the back.
-
1:50
Why Written ICPs Fail Static documents are too broad, too shallow, stuck in a doc nobody uses, and often miss product-level variation.
-
3:29
Defining Your Best Customers Look for lowest CAC, shortest payback, highest LTV-to-CAC, strong product usage, high NPS, and replicability.
-
6:50
Biggest Isn't Always Best Large enterprise logos can require heavy customization, destroy margins, and churn faster than mid-market wins.
-
9:09
Beyond Firmographics Move past industry and revenue range to isolate the one or two differentiating data points that matter most.
-
14:53
The Survivorship Bias Test If your ICP criteria describe 80% of your lost deals too, it's not actually differentiating anything.
-
19:26
Manual vs. AI in ICP Definition Start with human-defined best customers, then feed into AI for analysis, but always spot-check and validate.
-
30:17
Fit and Timing Together Fit criteria (revenue range, industry) are permanent; timing triggers (new hire, funding) are transient signals that matter now.
The lines worth sharing
If it describes 80% of your lost deals too, it's not differentiating. We want that data point that really goes to differentiate.
Eddie Reynolds · 14:53
Every single time a lead comes in, a deal closes or is lost, a customer churns or expands, push that signal into an AI engine that continuously reevaluates the ICP.
Eddie Reynolds · 29:52
A human being has to make a decision of whether we're going to change how we invest tens of millions of dollars in sales and marketing. That's not something you want to take lightly.
Eddie Reynolds · 37:12
We just changed the voice of the content and started speaking to mature companies. Overnight, we started getting the right types of leads. It was really shocking.
Eddie Reynolds · 27:51
Common questions from this episode
Why does my ICP document not actually change how my sales and marketing teams work?
Static ICP documents fail because they're too broad, never operationalized into qualification criteria, and disconnected from daily decisions. The fix: translate ICP into fields in your systems, MQL/SQL definitions, pipeline inspection rules, and account scoring logic that actually routes leads and flags mismatches.
How do I know which customers are actually my best, not just my biggest?
Analyze your best customers by payback period, LTV-to-CAC ratio, product usage, NPS, CS ticket volume, expansion rate, and replicability. Compare these to your lost deals and churned accounts. The customers easiest to win, cheapest to serve, and most profitable win out—regardless of deal size.
Should I use AI to define my ICP or start manually?
Start by manually defining which customers are best using financial and CS data. Then feed that list into AI with your chosen data points (revenue, industry, technology stack, etc.) and test if new criteria differentiate wins from losses. Always spot-check AI output before operationalizing it.
What's the difference between fit and timing in an ICP?
Fit criteria are permanent (revenue range, industry, company size) and stable for years. Timing triggers are transient signals (new hire, funding round, product launch) that create urgency now but expire in months. Both belong in your ICP definition and account scoring.
How often should I revisit and update my ICP definition?
With an AI-driven continuous improvement engine, you can review signals daily. Flag anomalies when close rates drop or churn spikes on certain account types. Formally revisit ICP quarterly at minimum, more often if your product, market, or competitive position shifts.
How do I get my sales team to trust an account scoring model?
Build it transparently using historical wins and losses. Keep the model simple—three to five key criteria max, not 50. Explain the scoring logic to reps. Create feedback loops where reps who spot bad scores can report back, and you update the model. Trust builds when reps see scores improve over time.
SEO meta description
Eddie Reynolds on turning your ICP from a static doc into a living system that sharpens every quarter, with continuous AI-driven improvement loops and human judgment.
Target keywords
ICP definition best practices
ideal customer profile example
account scoring model how to build
Eddie Reynolds go-to-market
Union Square Consulting
firmographics vs. behavioral data ICP
ICP continuous improvement engine
deal qualification criteria sales
fit vs. timing signals revenue
AI in sales operations
pipeline quality GTM ops
customer churn analysis ICP
Read the complete conversation
Auto-generated from the episode audio and lightly cleaned. Expand to read or search inside.
Read the full transcript · 66 KB
EDDIE0:00a lot of what we talk about in this podcast is the fundamentals. And I think this is where people fall into a trap of like, let's just go whip up an AI
EDDIE0:07What's great? Like, let's do that.
EDDIE0:08I'm all for it.
EDDIE0:09the reason we're talking about ICP today is this is the foundation of it. And what's the foundation of ICP is figuring
EDDIE0:14which of our customers are most profitable, and how do we define those customers with reliable data.
EDDIE0:19we need to really think hard about that. And then we can go build the AI agent with a human in the loop to do that work.
EDDIE0:26at the end of the day, like a human being has to make a decision of, are we going to change the way that we potentially invest tens of millions of dollars in sales and marketing to target or not target a group of companies that look like this?
EDDIE0:37That's not something that you want to take lightly.
EDDIE0:51consultants for our clients here at Union Square. Consulting and candid conversations with revenue leaders in the market that have been there. Now let's get into it.
RACHAEL1:02On paper, everything looks healthy. Reps are hitting activity. Leads are coming in. The pipeline looks full and yet deals are slipping, cycles are running long, and customers keep churning out the back. So much of that comes back to an ice that's too broad, too shallow, too outdated and stuck in a dock that nobody uses. We're going to get into how to make your ICP definition into a living system that sharpens itself every quarter, and actually points your teams at the accounts worth going after.
RACHAEL1:30And with me today is Eddie. Hey, Eddie has gone.
EDDIE1:33It's going well. What a great opener. Oh, solid.
RACHAEL1:37Cool. So, Eddie, most mature companies, especially in our audience listening to this, already have an ice written down somewhere. Or all of them probably have an ice period down somewhere. But why isn't that typically enough?
EDDIE1:50Well, a lot of what you just shared is that it's, you know, it's too shallow, it's too broad, and it's stuck in a dock that nobody looks at. And also you needed. You may need a different ICP for each product, potentially each different geography. So we're selling a product into SMB. We know clearly who who buys that. We've expanded into mid-market and enterprise.
EDDIE2:13We now launch new products. Now we're going into a mia. Those customers are going to look different and need to be defined differently. And so that's problem number one. Problem number two is that we don't go deep enough. We say okay it's these industries. It's this revenue or headcount range. It's this geography. And oftentimes we're not getting that much more specific.
EDDIE2:34And when we look at the deals that we've won and lost at the customers that have become really valuable for us and those that have churned, if we look at the customers that we're able to expand in the ones that we can't, then we are going to find that they all fit that ICP criteria. So what differentiates them from each other?
EDDIE2:53How do we know who we should be focused on in sales and marketing, so that we can get the most lifetime value out of the customer, have the most profitable customers, and have the most scalable business. And so we need to go deeper there and pick additional data points that are going to better describe and differentiate our best customers from the rest, so that we can better target them.
EDDIE3:13And then we also need to make this operational. We need to make this part of what drives how we qualify inbound leads, how we identify outbound prospects, how we qualify sales opportunities, and how we look at our existing customers for renewal, risk and potential expansion.
RACHAEL3:29so when you look at a customer base, how do you decide which customers are genuinely the best?
EDDIE3:35So I think what we want to look for is the set of customers that are going to be the easiest to attract when, retain, grow, serve, etc. and the cheapest and the most profitable and also the most replicable, right? And so some of the things that we want to look at are things like payback, LTV to cactus, we want to look at product usage, Net Promoter Score, CSat, customer service cases, etc. and we want to look at the customers that look like other customers or other prospects.
EDDIE4:05And so a good example of this is we've landed some whale enterprise customer and they pay us a ton of money, but their demands of us require us, if we're a software company, to customize our software specific to them, we are bending over backwards on the implementation and onboarding side. We are killing ourselves to like, win the customer and to serve the customer.
EDDIE4:27They're never happy. They're not really using the product well. We don't have a good customer story to tell. We're not making any money, and there's not a lot of other folks out there like them with the same problems, the same needs, the same desired solution. We just have this big fancy logo that we're able to put on our website, and we're scared to death that somebody might actually call that customer and ask them about their experience working with us.
EDDIE4:49And, you know, it's not going to be good. And we've experienced this as a service business ourselves. I can think of one particular customer, customer we worked with, not in the tech industry, and they're one of the largest companies in that industry. And it just felt like everything was so difficult working with them. And I woke up one day and said, like, why are we trying to make this customer happy when we are very miserable ourselves?
EDDIE5:17Like on an emotional level, trying to serve this customer. And from a financial perspective, we weren't making very much margin. So it's just like, not worth it. And I remember that particular customer was a bit abusive. And this is like an extreme example. They had a whole team of people that do similar work to what we do, and that team of people either quit or got fired, and that didn't help.
EDDIE5:36But I remember talking to one of the stakeholders over there, and it was very clear that they just had the attitude of like where the 800 pound gorilla and you should bend over backwards and want to please us. And I just said, look like I lose sleep at night trying to take care of you guys. And I could make way more money just having two smaller customers instead of you guys.
EDDIE5:53And I'm sorry if this is really unprofessional, but, like, you're basically terrorizing us. And, you know, I think that especially in the service business, every single services company has an example like that. When you're selling a product, if it's very easy to implement, maybe you don't have as much of that, but you have people logging customer service cases, etc. that are just taking so much time, energy and money away from your team versus really narrowing in on your ICP and saying like, wow, like these types of customers are really profitable and they're really happy they have success.
EDDIE6:23We have great customer stories. There's a bunch of other people like them. They're telling those people about us and singing our praises, and we can go sell and market to those people and tell these stories and solve the same problems with the same solutions that we are solving for this core group of customers, where we're having a lot of success and making money.
RACHAEL6:40It's the biggest customer isn't always the best. And I think in the same vein of that, the customers that you land the most of can also not necessarily be the best, right?
EDDIE6:50Yeah, absolutely. If they're not profitable. Right. I think this is where CCaC, payback and LTV to Cat come into play. Right. So if we have like a classic example here would be an SAB customer. We land so many SMB customers. I mean, this is an example that I lived and breathed in the first year of this business. We started this company or I started this company and I was selling, I think maybe I had 1 or 2 customers.
EDDIE7:13I sold for like $5,000 and immediately recognized, like not making any money with that. And they would churn immediately like they would spend five grand. There were a startup for like two people. That was it. That was all they could spend. I'd come back and say, like, we could take the ball further for you. Can you spend another five grand to do the next project?
EDDIE7:30And they're like, no, no, no, we're tapped out. This is all the money that we ever plan to ever spend on our go to market. Okay, so quickly I realized, let's move to eight Grant. We do eight grand, and now the average customer is spending another eight grand on the back end because they want to go into project number two.
EDDIE7:45But I was personally doing the sales. And at the time I had left Salesforce and I was running a sales process that required all the skills that I had acquired at Salesforce, plus all the technical skills of a Salesforce architect. And I was looking at this and I'm like, how do I replace myself? And I'm like, so I need to go find somebody at Salesforce and give them like a 200 K plus OT.
EDDIE8:06Then I need to go find another person like Jerry, who runs our delivery team now and pay that person hundreds of thousands of dollars a year and have those two people sell this deal. And this deal is going to net us out $16,000 of revenue. Like from a payback perspective, that doesn't make any financial sense. And I remember talking to competitors that were in the same business and asked them, like, how are you selling?
EDDIE8:30Like you work at some big company like Blue Wolf? Like, how do you sell? And he's like, oh, I have the sales in there and do all the technical stuff. And I'm like, I can't afford to pay two people like, high six figures to sell a $15,000 deal. It doesn't make any financial sense. So I think that that's a really classic example.
EDDIE8:45And then of course there's churn that plays into that. Right? So what we want to do is look at which customers are most profitable, happiest, most replicable. And I think that the way that we identify who those customers are comes down to looking at the financials, the product usage, the CS data, etc.. But how we define those customers is where we need to get more specific on key criteria.
EDDIE9:09And it's different for every company. Right. So we talk about going beyond the standard firm and by extension technocratic data. But what does that mean. Right. Well, it's hard to summarize because it's different for every company. In our case, for example, we try to target companies that are selling products for around 50 to $500,000. Right. That's a unique data point, right?
EDDIE9:31It matters to us how much our customers are selling their product for. And that's not necessarily something that I can just go in and click a button in a tool to filter for, and that's an important thing. If I can't filter for it, it's not super actionable at certain points in the journey. But what we need to do is figure out what are those key criteria that are going to really define our customer and differentiate the best customers from the rest of them, so that we can better target the next batch of potential best customers.
EDDIE10:02And it's unique to every single company and even every product that's being sold by that company. So at the risk of kind of droning on here, you know, I worked at Salesforce as they grew from 5 billion to $10 billion in revenue. And as you can imagine, that's a very complicated, you know, company and structure. And I didn't work across the entire business like, for example, in India are pricing was I don't know if it's still that way today, like 40% off because, you know, just totally different culture in India.
EDDIE10:30They're looking at this and they're like, well, we don't pay our people nearly as much as you do in the US. And so we're not going to spend as much on a tool that automates workflow unless you substantially discount it. So just think about what that does to defining your ICP, what your sales process is, etc.. I mean.
EDDIE10:48You have lots of companies in India, for example, that are like heavy in manufacturing and have really thin margins. And so the way they're going to look at buying a tool for sales, marketing and CS is going to be wildly different from the way an American company would look at that. And you're going to need to define your ICP differently as we go from product to product, from sales cloud to service cloud to marketing cloud.
EDDIE11:12Yes, like the target customer is similar. Buyer personas are obviously different. And that's not what we're here talking about today. But for example, I remember when I was trying to learn how to sell Pardot, part of Marketing Cloud. We were looking at companies that had succeeded on Sales Cloud. And then you look at their website and their websites terrible.
EDDIE11:32And you think, what is the likelihood that this company is going to spend $30,000 on a marketing tool when it looks like they spent $500 on their website? That's a very different definition of an ICP for that particular tool, as opposed to Sales Cloud, where, hey, they have a crappy website, but they've got like ten or 20 or 30 salespeople.
EDDIE11:54They'll buy a sales to all day long.
RACHAEL11:56So for somebody listening to this, wanting to go deeper than from graphics and, and find, you know, the traits across all of their best customers, how would you suggest they approach that?
EDDIE12:08Yeah. So first of all, this is such a hard question to answer because I sat down and mapped out a list of over 150 different data points that you could consider. Now, I am by no means recommending that somebody go and try to identify 50 or 100 different data points to define their ICP. Here's a big problem. If you were to do this number one, just forget like how difficult it's going to be to build the AI engine.
EDDIE12:30And and I think we kind of skipped ahead here on this podcast, I want to talk about how to build this continuously improving AI engine here on this podcast. And in order to do that, it needs to be somewhat simple so that a you can build it and be people will trust it. If you've got like 100 different data points, your sales team is never going to trust it and they're not going to leverage the ICP to, you know, target the right companies for outbound or follow up with the leads, because they're just going to say, like, the leads are crap, the prospects are crap because this is too complicated and I don't understand it,
EDDIE12:59and it probably won't be good because it's too complicated and you probably won't get it right. But with every single company, I think defining that ICP or even every product is different. So we do have a list that you can find in the show notes of over 150 different data points that you might consider that might help you differentiate.
EDDIE13:17And so let's take like the basic stuff, the revenue headcount, the tech that they're using, geo industry, etc.. And then let's layer on a couple data points. Start with a hypothesis based on our sales conversations. What do we think might differentiate our customer. And then what you can do is you can take those best customers that you've identified by running that analysis, and you could have done that analysis in an AI tool by feeding it in the financial data, the customer usage, the customer service data, etc..
EDDIE13:48CSat, NPS feed that into AI say help me identify my best customers based on this criteria. You go and review it, spot, check it manually, make sure it's right now you've got your best customers, and now you say, let me layer on these data points. Let me look at, let's just say, hypothetically speaking, that we have the opportunity to find out which companies have which companies sell their product between X and Y dollars.
EDDIE14:16Hypothetically, if we could get access to that data, we layer that data in and then say, does this differentiate our best customers from the rest are wins from our losses, our renewals from our churn, our expansion, from our stagnation. And run that analysis 1 or 2 data points at a time. And so you've come up with something that really does differentiate.
EDDIE14:37And as I was thinking through this article and reading other articles and doing research, one of the things that jumped out at me was that it said, well, if you say that x, y, z is criteria for your ICP, but it also describes 80% of all the customers that you lost, then that's not a differentiating piece of criteria.
EDDIE14:53So we want to look for that data point that's really going to differentiate. And it's going to take some trial and error by identifying specific data points, looking at the actual list of customers and prospects and opportunities, and turn customers and probably feeding this into an AI engine to do the analysis and help identify what data points are going to best describe your ICP.
RACHAEL15:15So where do you start? If you have a very limited number of data points that you can feed into the AI? To get this.
EDDIE15:21I think having a limited number of data points is totally fine. I don't think that we need even necessarily ten or 20 or 30 data points when we get into account scoring or lead scoring. We might, especially when it comes to sort of engagement with our marketing data. But we don't want to overcomplicate this. We just want to think long and hard about what specific data point is really going to differentiate.
EDDIE15:42And if you're new to the organization and you have no background context, go talk to the sales team, talk to the marketing team, talk to the CS team and find out like, what do they see with their own eyes? I can tell you anecdotally from working with our customers what I see every day. And then what I would need to do is to go test that by actually looking at our data and validating that that matches or better correlates with our better customers than the rest of our customers.
RACHAEL16:07And how often do you find defining criterion for ICP? By looking at things that all the people that churn have in common. I know that there's this
RACHAEL16:19story from some World war I can't remember where they had planes coming back, and they were analyzing the bullet holes and the planes that kept coming back, and they were seeing that there was a pattern of where there were no bullet holes on the planes that came back.
RACHAEL16:33And so they ended up reinforcing those areas of the plane because it was, you know, the planes that got shot down were the ones that actually got hit in those areas. So those were the actual areas plane that were most susceptible to destroying plane. So that always gets me thinking, like, how can you find patterns in quote unquote, the planes that don't come back?
RACHAEL16:54So the people that churn are the people that you lose?
EDDIE16:56I think that's a really, really great example. Right.
EDDIE16:58And on the one hand, I want to run with this and I want to say absolutely, like, let's look at the customers that we churned first, and let's try to filter them out so that when we reach out to new potential customers, we're targeting the people that are most likely to not churn.
EDDIE17:13That being said, at the end of the day, we are in business to make money. And so if I can land a customer that sure like buys today, insurance tomorrow and make a profit off of that, and when I say profit, I don't just mean the gross margin from serving that customer, but I mean the entire cost of selling and marketing to them.
EDDIE17:31Then in some ways that's a great thing. And like, let's rinse and repeat. Right now we are talking primarily to B2B tech companies, and recurring revenue is a critical thing. And not just recurring revenue but expansion. So we really want to look not just at the churns, but the lifetime value. However, with that said, the question that I would always ask is would you want more customers like them?
EDDIE17:55So if you land a customer and they churn after three years, is that a good thing or a bad thing? Sure, 4 or 5 years is better, but does it make financial sense to spend more money and more resources to try to land more customers that are going to churn in three years? Maybe it does, maybe it doesn't, and that's a financial decision for the business to make.
EDDIE18:14If it's just as easy to land customers that stick around for ten years, then you probably want to do that. But I think, like in our own business, what we've seen is some of our larger customers that do more strategic work with us have churned a little bit faster, and some of our smaller customers that leverage us more for just systems admin work retain us for longer.
EDDIE18:34And that creates a really serious question like, which do we want more of bigger customers that spend more money quicker and then churn faster, and they're still sticking around for a few years, or customers that were just updating their systems for five, 6 or 7 years on end. And I like both sets of customers. I understand why a customer, at least for us, were a professional services business, not a tech company would churn.
EDDIE18:58After a few years, they mature and grow and they hire a full time resource and they don't need our strategic help anymore. And that's fine. But when I look at that customer and say, wow, that was profitable, it cost us less money to acquire that customer and serve that customer than they paid us. And by a good, healthy margin, I want more of those customers.
RACHAEL19:17So how much of this analysis should you be doing manually by hand before you start implementing it into AI or using AI to help?
EDDIE19:26You know, it's an interesting thing. I had a debate with Jerry, who runs our delivery team, about this, and I personally like doing things by hand at first. I think there's some value in it. Jerry pushed back and said, no, like we can just as easily dump this all into AI right away. The important thing is to spot check it.
EDDIE19:42And I totally understand where Jerry is coming from with this, right? I think what's most important is that you are validating it. You have a human in the loop and you're saying, okay, the AI tells me that our best customers have this criteria, and the rest of the customers have this. The customers that churned have this criteria, and the customers that we retained and expanded to have that criteria.
EDDIE20:02Are you spot checking it? And Jerry points out that, like, you could have a sheet with a thousand rows of data and you can still easily spot check that. I think where you set the foundation is figuring out how do we define the best customers, literally taking out a sheet and manually assigning these names as these are our quote unquote best customers and comparing them to the rest of your customers, comparing them to the customers that churn, the customers that you lost in deals, the leads that came in and didn't convert, etc..
EDDIE20:35And if we have that definition clearly defined, then we can feed that into an AI engine and quickly do analysis on multiple different factors and then have it spit back and then say, okay, it says that this piece of criteria, it says that at this revenue range or ICP is or the companies are much better, much better fit for ICP.
EDDIE20:52Let's check that does that check. And that prevents us from being victims of the AI, hallucinating or taking us down the wrong path.
RACHAEL20:59What's at stake? If you scale this up and you start using AI to help with your definitions, and you start shifting budget and attention and resources towards these new definitions before you've actually validated what you've landed on.
EDDIE21:14I think everything is at stake. Right? So if you think about go to market, the fundamental like building block of go to market is ICP. And so it doesn't make any sense to just like build an AI engine and just run with it and then divert for the companies that we are talking to with 50 or $100 million an hour or far more, if you're going to divert tens of millions of dollars of sales and marketing spend, like I can't imagine any I can't imagine in the organization doing that without validating it.
EDDIE21:47But if you validate it and you validate it on an ongoing basis, this, I think, is where the opportunity lies and what we're going to get into later in the podcast and later in the article that that is associated with it, is building this continuous improvement engine. What we don't want to do is sit down and define ICP today.
EDDIE22:03And then three years later, we wake up and we realize that, you know, we've launched new products, our product has evolved, the market shifted. I mean, just think like as of 2026, right now we're talking about this. How much has our world been turned upside down in the last six years? We had the pandemic. We have AI. That's I feel like in 2020, AI was this thing that we talked about that like might come and like cool, like Amazon and Netflix had some recommendations for us, but it didn't feel like it impacted like our work.
EDDIE22:31And now it's all we're talking about what is going to happen in the next three, 4 or 5 years. And if we sit down today and define our ICP and then just let it stagnate for five years, I don't think we have any choice but to build this engine that is going to continuously improve. But there has to be human in the loop to ensure that we are diverting tens or potentially hundreds of millions of dollars of spend in sales and marketing into the right areas.
RACHAEL22:58When it comes to enriching this data and making sure you have
RACHAEL23:02comprehensive and correct, what kind of data do you trust databases for and what did you not?
EDDIE23:10Oh, this is a great question. So I think where you're going with this is our recommendation that we start with the deterministic data such as, you know, revenue, headcount funding, things like that that can come out of a database like zoom info where it's just hard data. Right. And then there's more, you know, AI driven, nuanced data of like, do they have a crappy website?
EDDIE23:33You know, can we go to their website and figure out, like, what kind of customers they are targeting, what's the culture of their company? All kinds of things like that. Right. And so and this is where it gets really murky because even these tool, deterministic databases that have hard data also have layered in AI, because every single software company on Earth has layered AI into their product.
EDDIE23:51So whether they're AI native or not is a debate. And so we're going to have both in every system that we use. But we want to start with that hard data before we go into sort of the fuzzy data. Right. We're building that foundation of the hard data so that we can define our ICP as narrowly as possible.
EDDIE24:10And then within that go and say, okay, here's the last piece that we need to look for. We are trying to sell to companies that value their employees. Like how do we how do we gauge that? That's a very fuzzy thing to gauge.
RACHAEL24:22Do you have an example of an ICP criteria? That's like one of these fuzzy things that ends up being the most differentiating.
EDDIE24:29Like I said, I think it's really different for every product in every company. The thing that I noticed when I sold part OT for Salesforce was, do they have a shitty website? Like what it really came down to is, does this company look like they invest in marketing? And some of that's in a way hard data couldn't necessarily.
EDDIE24:47I guess you could get it out of a database because it's techno graphic data. But we would go to the website and you would see all the tools they have installed that we're tracking you on their website. We see a bunch of tools pop up. It's like, okay, well, they clearly invest in marketing and it's a really nice looking website.
EDDIE25:03That's one criteria I'm trying to think about, like for us. And this one's really hard. I don't know exactly how you would even gauge this
EDDIE25:12even with AI, so maybe it's a bad example, but with us we work best with crows that quote unquote get it? The crows that value a data driven, process driven, go to market over the revenue leaders that are more old school, and they just want to hire sales reps and spend more money on ads and have more activity, etc. that's a really hard thing to gauge, but it influences our sales process.
EDDIE25:35We know after the first or second call, if we're talking to somebody who's mentality is more of the I just want to like set up Salesforce and HubSpot and outreach, and we need to like log phone calls and, you know, close deals and that's it versus somebody that says they really buy into what we are talking about here.
EDDIE25:56We know that we're going to have, you know, a radically different experience, both in the sales process and like with them as a potential customer. But how do you how do you gauge that? Like that's that's a harder thing. Right. And so I want to give really tangible examples here on this podcast. But it's just I think through I've worked in numerous companies in numerous industries, and every single time I've been selling something, the best customer always has this like really unique nuance that is something unique to that particular product and customer.
RACHAEL26:26I think that example is pretty good, because it also
RACHAEL26:29makes me think of another important part of ICP, which is, you know, marketing. So not not just like how you sell to a person or how you're enriching your account or tier list data, but how you're marketing to them as well. Because for us, knowing that thing about our ICP changed a lot for how we market in the content that we put out, we put out a lot of content trying to like, care to that audience, the crows who quote unquote, get it out and or trying to convert more crows into crows that get it so that potentially maybe down the road they could be
RACHAEL27:02customers.
EDDIE27:02Yeah, absolutely. And that's the whole point here, is this is the cornerstone upon which all of go to market is built. And for us, what we saw tangibly was that leads were not converting. And what we could have done is we could have done a massive operational exercise to think about, you know, I mean, just everything that you can possibly think of that we could have done in marketing and we probably would have done that if we had to.
EDDIE27:28But what we actually did was we just changed the voice of the content, and we started trying to speak directly to the ICP that we wanted to address these more mature companies, the crows in those companies. And like overnight, all of a sudden we started getting the right types of leads. And it was really shocking. Like, I wouldn't have believed it would have happened that easily, but it just so happened that it did.
EDDIE27:51And if it hadn't worked out that easily, we would have gone further down the path and tweaked a bunch of other things until we got there. Because, you know, this is the whole thing that we're talking about. It's like, why are you spending all this money in marketing to generate a bunch of leads from the wrong companies? Yeah.
EDDIE28:05And unfortunately, so many organizations do that. And then you've got your sales reps chasing these leads, and they're trying to close them, and they're having so much trouble. I mean, I feel like everything I'm saying is very obvious, but we lived in breeds that for the first year or two that we built a real marketing engine. And it's like, man, we've spent all this time and all this money, we generated all these leads, but none of them are converting.
EDDIE28:26And the thing that fixed it the most was just thinking really hard about who's our ICP and what messaging is going to speak to them.
RACHAEL28:33absolutely. And even to this day, we're still tweaking it like it's not something that we figured out and decided on and then never revisited it again. Like we're constantly working on it and finding new angles to approach
EDDIE28:45I don't even know if I've told you this, but I literally talked to Jerry the other day and we're doing another pass at this.
RACHAEL28:50Oh, no. This news.
EDDIE28:52Yeah.
RACHAEL28:53I feel like it's an.
EDDIE28:54Everyday updating it.
RACHAEL28:55Yeah. I feel like for me it's like an everyday thing that I'm thinking about. Like, is there something that I'm missing? Is there something more that we could be doing? But that's marketing.
EDDIE29:05Well, and I think that this is the exciting thing, right? And this is why I want to talk about building this continuously improving AI driven engine
EDDIE29:12But as we do this exercise manually, again, it's really obvious, especially for a larger company with larger set of data, that you can semi-automated all of this and you still need the human in the loop. But every single time a lead comes in, every single time an opportunity gets qualified and then is one or lost.
EDDIE29:30Every single time a customer churns or is expanded, you can push that signal into an AI engine that will then continuously reevaluate the ICP and say, look, we thought the ICP was this a year ago, but it turns out that these companies with this criteria are churning out faster or losing these deals, etc.. And now what we're seeing with the companies we're having the most success with is they look like this.
RACHAEL29:52Quick pause. Everything we talk about on this show. Diagnosing go to market ops. Prioritizing projects for revenue. Impact processes. Metrics. Insights. Building a predictable go to market engine. We've built frameworks for all of it. They're free and undated on our website. Union square consulting frameworks. The link will also be in the show notes, so make sure you check that out.
RACHAEL30:15All right. Back to the episode.
RACHAEL30:17And so I want to switch gears a little bit and talk about fit versus timing.
RACHAEL30:22This is kind of related to, you know, the more recent ICP changes or more recent things we've been thinking about in our own ICP, that usually it's crows who are either new to the role or new to the company, and they have inherited messy go to market operations and they need execution.
RACHAEL30:41Help figuring it all out and getting it all stood up. Those tend to be, again, another part of our ICP that's really valuable for us. So. So Eddie, how do you think about fit versus timing and how they both come together and your definition for ICP?
EDDIE30:57I think it really comes down to permanence. Right. And so if we say hypothetically that our ICP or companies with 50 to $500 million in revenue, then the companies that fit that criteria are going to stay companies that fit that criteria for quite some period of time. Right. A company that's at $200 million in revenue has been interact for many, many years and will continue to be in our ICP for many, many years.
EDDIE31:21The timing event is something that is much more timely, right? They hire new Crow. Well, that's relevant for maybe a few months, maybe a few more, maybe less. And that's it. And so what I think you want to do is when you think about your sales and marketing efforts, you're saying, who are the companies that are a fit?
EDDIE31:41So in this example, the companies that are between X and Y dollars of revenue plus all the other things, right. Like I'm literally contradicting myself because I say you got to go deeper with this. I'm just trying to provide a quick example for the podcast. So how we got our our ICP defined by fit. And then within that, what are the timing triggers such as a new hire, a new fundraise, a new product launch?
EDDIE32:00Whatever it is, it's relevant for the thing that you're selling that is going to indicate that they are potentially a much better opportunity for you right now. And then, you know, that's maybe separate from intent signals as well. And so then you've got folks coming onto your website engaging, etc.. And so you put all of these things together and that's where you can, you know, prioritize who you're targeting, especially when it comes to manual human outreach via your sales team.
RACHAEL32:29And of all the intent or timing signals out there, which do you trust the most?
EDDIE32:34I think it just depends so much on the product. But like as a general rule, I would I would trust the the first party marketing data the most. Right? So if somebody is on our website, they're engaged with our content. That is a lot more relevant signal than with like what we're going to see with third party intent signals.
EDDIE32:53So or at least that's the hypothesis. Let me caveat this. The whole point of this entire thing is to test these things, right. And so, you know, we literally wrote that in the article. But then, you know, maybe your data shows something different. Maybe there's like intent data out there and you can map that to your wins and losses and see that, wow, this is a really, really strong signal.
EDDIE33:14I think for most companies, it's not going to be as strong as the engagement with your first party marketing data. But it depends on your marketing data and it depends on the intent signal. You know, there could be a situation where you put it on an ad for a white paper and it means absolutely nothing, and that's the only thing you've done in marketing.
EDDIE33:29But these intent signals out here are far more valuable. I think it depends on so many factors. And the whole point of this entire thing is to build an engine to test that against your actual data.
RACHAEL33:40I was just going to ask what kind of signals tell you that account is ready right now. But I think, yeah, you just answered it. It depends.
EDDIE33:47Yeah. And it's like I want to give very concrete examples because we always try to do that in this podcast. And this particular topic is so hard because it's different for every company and every product.
RACHAEL33:56So moving on to like building the actual account scoring model. How do you take all of this that we talked about already, the timing and ten signals and turn it into a single thing that are Rep can act on and build on?
EDDIE34:11Yeah. So I think there's two separate questions in there. One is like how do you build the model and then how does the Rep act on it and prove on it. Right. So I think the first thing I'll say is, as somebody who's been a sales rep for my whole life, including now, sales reps should always be taking things the last mile, at least at least in the world that we play in, where we're selling a product for a reasonable amount of money such that there's actually like a pretty developed process that the sales rep is executing.
EDDIE34:37Right. I sold cell phones in college for, like, you know, 50 bucks a month or whatever, and I think I sold like 14 per week. And that's a different game, right? But we're talking about selling a 50 or $100,000 product. I personally believe that the rep should always be taking everything the last mile. Before that, we want to build an account scoring model that is essentially very similar to what we've just been describing, and it's also very similar to a lead scoring model in the sense that you have to start with an initial hypothesis and keep testing it.
EDDIE35:08Right. And so we're going to take a look at all of our historic data. And we're going to try to set some kind of a point value for various criteria. Now one way to do this is to say that, you know, if certain factors aren't there, then it's just an automatic zero. This account is not something that we want to target.
EDDIE35:24The score is zero, right? Other things we're going to then add up and say okay. Like if they have these factors they're going to get this many points. If they have this level of intent on our website, maybe third party intent gets a score, etc.. I think the most that we can do to centralize this, but also keep it somewhat simple.
EDDIE35:42We can then explain to the sales team. This is why the account score has this score. And then here's a nice debate that I had with Jerry about this is how often do we want to update this score? I think the answer we landed on is however often we're going to update our outreach. So let's say that like we have sales sequences where we reach out to folks over a span of 30 or 45 days.
EDDIE36:06Well, if we can build this continuous improvement engine, and I think the account score is a somewhat separate but obviously very closely related topic, that then also segways into ABM, not to mention inbound and outbound.
EDDIE36:18If we can build this continuous improvement machine, then we can continuously update the account scores and every 30 or 45 days that we launch new sequences, because you want to finish the sequence, right.
EDDIE36:30Unless there's some signal that just says like, absolutely, stop here. Like if we found out the company just went bankrupt, that might be a good reason to stop calling them. But we want to finish the sequence. And then now we can reprioritize the accounts based on all of that stuff. And what I think is really interesting about combining all this data, especially with the use of AI, which has a serious advantage because you don't have to combine the data like we did in the past, where we integrate databases and map the data, can literally feed it into the AI and have the AI analyze it for us in real time, which I think is really,
EDDIE37:00really powerful. We can also see how our sales outreach might translate into marketing engagement, such that it then increases the account score, and it's telling us, hey, this person hasn't responded to you.
EDDIE37:14but they're going to the website and they're reading all the marketing content. And now we have increased that score because even though nobody responded to sales outreach, their level of engagement has elevated.
EDDIE37:25And that's something that I think is really exciting about putting all this together and thinking holistically about accounts. But that's kind of a separate topic. Did I answer your question, or do I do I need to go further here? Oh sorry, I actually I remember the second piece.
RACHAEL37:39Yeah.
EDDIE37:39How do the reps actually prioritize. So I think that ultimately, you know, it depends on the reps and it depends on their emotion. It depends on what they're doing. Are we talking about stars. We talked about A's or we're talking about SMB or are we talking about enterprise. But I think tearing is really important. You might want to say, okay, I'm going to tear these different accounts counts with this score.
EDDIE37:58Go into tier one accounts with the score, tier two, tier three, etc.. And then I'm probably going to go and I'm going to research these accounts manually. Now we can use AI to help us out a lot with that. But at the end of the day. I'm never going to fully trust a machine to tell me which companies I should be targeting.
EDDIE38:13And so I, as a rep, would want to go in and make sure that my tier ones are indeed my tier ones, and then I'm going to allocate my time accordingly. Now, if I'm an STR that's more operating off of a sequence or a cadence, then that's going to be different. But I think that's part of the answer to your question is, I think depending on the reps role, they might want to do a little bit of research and take it the last mile, validate that score.
EDDIE38:36And what's really important is that we have this constant feedback loop where rev ops and the sales team are talking constantly, and sales team is updating rev ops on where they're seeing these scores land out and giving them feedback so they can optimize the scores such that then those scores get better. And then there's this trust built between rev ops and sales and also marketing.
RACHAEL38:58So when you say a rep should be, you know, double checking this information, is it just they should be checking that, you know, the company is actually $50 million in revenue. Or do you mean they should be checking that this is actually a score? That is a good signal for an ICP account or both?
EDDIE39:17The latter, the latter. Right. So we say that, you know, here's this account. It's got a score of 800 or whatever. This is your top tier one. And then the rep goes to the website and starts researching it and finds out, wow. Like, this is this is a terrible account. Well, that feedback needs to go back to rev ops.
EDDIE39:34And then the question is was this a one off. Like there was something there. The company just went bankrupt and otherwise it was an ideal company or the model was just like completely off. And I think that's the problem with scoring models, is that they need to be constantly updated because you're testing a theory. But hopefully if we've done a really good job of defining our ICP, we're not like light years off because we've looked at historic data.
EDDIE39:55And the feedback to the rep is, yeah, this is the right industry, the right headcount, the right Geo, the right revenue range. And as well as these four other factors that are really critical to our success. So you tell me what's wrong with that account. But ultimately, I think it's really important that you have this direct feedback relationship between sales and rev ops and marketing and rev ops to optimize that account score.
RACHAEL40:20And how much time should a rep be spending on analyzing this and ensuring it's the right account score? Because I know the whole point of this is to give reps their selling time back and make it so that they don't have to really deeply research every single account that they look
EDDIE40:36I think it depends on the sales motion. So the first question is like, is there rep going to do upfront research on their territory? So when I was at Salesforce, we didn't have the advent of a lot of these advanced tools in AI. You'd go through each account and go website by website and validate your like tier ones.
EDDIE40:50Your tier two is your tier three spending more time on those tier ones? Now we can literally have Claude say, go to the website and like, look for these things and come back to me. And there's no reason rev ops can't, you know, get the ball rolling on that first. But there's I still think you need a human loop to validate it.
EDDIE41:06So the question is if I'm a rep and I'm just being assigned a territory for a year or six months or three months, am I going to spend like at least part of a day, sit down for a few hours and review my accounts? Absolutely. Am I going to use AI to help me do that? Of course. But I want to validate that I'm going to invest all of my time in the right accounts.
EDDIE41:25If I'm a 22 year old SDR that just joined the organization, I have no idea what I'm doing, and I've been given like 500 accounts to like, work a sequence on over the next quarter or 30 days or whatever. Yeah, I'm probably not going to do that. I'm probably just going to like, trust it and run with it.
EDDIE41:39However, as I get into those accounts, the question is, and this kind of deviates into a totally separate conversation about sales processes, like what level of research should the rep be doing? We have always advocated that there should be some level of personalization and relevance, and that Rep should go to the website and research the account. I don't know why we have sales reps, if not that.
EDDIE41:59Like we have AI tools, we can just let launch artisan have it right the AI and just hit send. So if the rep isn't going to do any research, what's the point of having a sales rep? If the rep is going to do some research, then they need to decide if they find an account that looks like it's not worth pursuing.
EDDIE42:15Then you know they should cut bait and provide that feedback. Back to revamp so we can optimize the model. And then what I would say is, is that if these are one offs, there's unique situations. Fine. You're always going to have that. But if reps are consistently running into this thing and they're saying, look like I see x, y, z on the website or in their LinkedIn profile or what have you, and that tells me that this is a terrible account to pursue.
EDDIE42:38We should update the model.
RACHAEL42:39Beyond a target account list? What else in the revenue process shouldn't ICP be driving operationally?
EDDIE42:46Okay, so it's a drive target account list and territories which are obviously pretty interrelated. That includes everything for customer success, right? Who are we focused on trying to retain and also who can we expand? We may find out that we have customers, that we don't want to lose them because they're paying us good money, but they're very, very difficult to expand.
EDDIE43:07We may find other companies are going to churn no matter what we do. And I'm not saying that we necessarily give up, but you want to focus on your customers accordingly, right? So when I worked at Salesforce, I had both new companies and existing customers, and one of the first things that I would do is just tear my accounts primarily based on usage, which would go back into that account score.
EDDIE43:28And I'd look at it and I'd say like, half of these customers are a waste of my time if I call them, I can see already they're not healthy accounts. I'm not going to get very far. And so I'm going to push that to the CSM and hope that they will turn that account and make it healthy. I might even make some calls on, you know, on their behalf and try to set up meetings so I can get that going, but there's nothing that I can sell to them right now.
EDDIE43:49And now I'm going to focus on these healthy customers that I can actually expand. So, you know, it goes across all of go to market. It goes into how we qualify inbound leads, the outbound prospects we want to go after and especially how do we qualify opportunities. And so if we have updated our ICP, we need to update our sales qualification criteria to make sure that that lines up.
EDDIE44:12And obviously like we don't want to take a company that's far outside of our ICP and put it into our pipeline and forecast it to close. If what we know is that we have a very low chance of closing it because it's so far outside of our ICP.
RACHAEL44:24And so you mentioned, like all the different departments, organization that's connected to revenue. Where does an ICP definition need to live for sales, marketing, CS product, everyone to all be able to see it and work off of it and work towards the same targets?
EDDIE44:41I mean, I think in this case, to answer your question, it can quite literally live in a document somewhere, right? Like it's in a shared drive. No big deal. But when we talk about not leaving it in a document, what we mean is, is that we have also documented our definition for marketing qualified leads, sales accepted leads, sales qualified leads, sales, qualified opportunities.
EDDIE45:01And that then translates into fields in our systems. It translates into how we actually qualify those deals and what deals get routed and pass to sales, and what deals get moved into pipeline and qualified, and how those deals are inspected in the pipeline. And so if, for example, let me try to think of a good concrete example like, let's say that our ICP has a sales team of 20 or more people, and we are able to get access to that data and enrich the data.
EDDIE45:30And then a deal moves into the pipeline, and it says that it only has two salespeople. Do we have any flag, whether it's automatic or manual human, where someone's going to catch that and say, wow, our ICP is 20 salespeople and above, and this is only two salespeople. Why is this sitting in pipeline? That's how it translates and becomes operational.
RACHAEL45:52So when we think of treating the ICP as a living operational system that's constantly evolving, how often do we say it needs to actually be resisted?
EDDIE46:02So I think with the ability to build an agent to do this, it can be quote unquote revisited every single day, every single time that we have a win or a loss or a churn or a renewal or expansion or even a lead that converts or doesn't convert, we can feed that signal into the AI engine. And then the question is, how often do we want to have a human review it if that's feeding into the engine, and the engine is just looking at this and saying, this validates everything that we've always been saying, then no big deal.
EDDIE46:33But we can also build an agent that's going to flag things to us and notify the VP of Rev Ops, the Crow, the CMO, etc. and say, look like companies that look like this are not converting through the funnel at this stage. Do you want to revisit this? And if we suddenly start seeing that like as an example that are closed rate is is down right?
EDDIE46:57Like I'm kind of assuming that go to markets working. Right. So let's just assume that our clothes rate is 30% or 25% or even 20%. And now all of a sudden the AI agent flagged something and it says, look, every single time you get a company that looks like this and you've had 100 companies in your pipeline in the last quarter that look like this, the close rate is only 5%.
EDDIE47:20That's a really good time to be asking. Do we want to revisit how we qualify deals? Do we want to revisit that ICP definition? And I think the whole point of the AI agent is just to process the data and flag issues and then ask the humans, do you want to look at this and take action on it?
EDDIE47:39Do you want to change the way that we go to market, or is everything okay? And we have now have the ability to automate the enrichment of our data and automate the analysis of our data such that we can do this on a virtually continuous basis.
RACHAEL47:53it's crazy to think like where this all might be next year at this time or God, five years from now, I wonder how obsolete or how much of this that we're talking about right now will be obsolete, and that's fine.
EDDIE48:07Well, I think that the tools are always evolving and changing, but the fundamentals somewhat stay the same. Now, if you think about over the past 20 years, we have access to a lot more data than we used to. I mean, the whole idea of like measuring marketing engagement and having like an MKL, all of that came about because we suddenly had the ability to get access to this data.
EDDIE48:30Intent data, you know, is a somewhat new thing in a longer term horizon. But at the end of the day, I think like a lot of what we talk about in this podcast is the fundamentals. And I think this is where people fall into a trap of like, let's just go whip up an AI agent. What's great? Like, let's do that.
EDDIE48:47I'm all for it. But we need to like, the reason we're talking about ICP today is this is the foundation of it. And what's the foundation of ICP is figuring out which of our customers are most profitable, and how do we define those customers with reliable data. And we need to really think hard about that. And then we can go build the AI agent with a human in the loop to do that work.
EDDIE49:08But at the end of the day, like a human being has to make a decision of, are we going to change the way that we potentially invest tens of millions of dollars in sales and marketing to target or not target a group of companies that look like this? That's not something that you want to take lightly.
RACHAEL49:23I think that answered. One of my last questions I had was, where's the AI roll stop in your start?
RACHAEL49:31Unless you had anything else to add to that?
EDDIE49:33No. I think at the end of the day, like the AI is giving you signals, it's the orchestration layer. It's not the layer that provides the judgment. So the humans in the organization, the VP of the Crow, the CMO, the CEO have to make the final call to say we are or are not going to change our definition of ICP and who we target and marketing and who we target in sales and who we target in CS to, you know, drive expansion, for example, we're spending tens of millions of dollars there.
EDDIE50:03This is where we're going to focus. The AI can provide the signal. The humans need to make the decision at the end of the day.
RACHAEL50:10And so the last question I have, we might have talked about it a little bit before, but just to summarize at the end here, where do you think the richer learning is when thinking of ICP? Is it in the accounts you win or in the ones that you lose in churn?
EDDIE50:26Man, this feels like a philosophical question.
EDDIE50:31I don't know, it's like I feel like in some ways you always learn more by losing. So I'm going to go with that answer. But at the same time, like somewhat philosophically, you don't want to spend all your time losing, like, let's, let's get some wins here. And I think, like one of the things that is really important here that I think a lot about is customer stories,
EDDIE50:58because at the end of the day, what you want is to land a great customer.
EDDIE51:02That then becomes a customer story that you can then share with other companies that are just like them because you want more of those customers. And maybe I think about this a lot because I have like so much scar tissue from doing this the wrong way. I can't tell you how many times, like in the early years of this company where we landed a customer and we made all this money, and we're really proud of the work we did, and they were happy with us.
EDDIE51:23And then it's like, okay, like, do we want to do a customer story? And then we say, not really, because we don't want other customers like them. This isn't who we're focused on. This was a one off for us. And now it feels like there's like a amount of compound interest that we are losing on that work. Okay.
EDDIE51:46We landed the company, they're happy we did work, we made money. But there's no compound interest that we can roll into, like ten more like them. And that's the whole point of having ICP, right? We want to own a niche. And the way you own a niche is by defining a niche that no one else has defined.
RACHAEL52:03Yeah, and I'm really glad you brought that up because I think it comes into every department, no organization. You know, obviously customer stories affects marketing, but also it affects products. You know, being able to know what's working for your ICP and where you could be making your product better if you're not focusing and working on the right IPS or the right people, then you're missing out on those opportunities to get that feedback and hone in your product or service or whatever it is you're doing.
EDDIE52:34Yeah, absolutely. I mean, this is the whole I go to market flywheel. It's all all tied together. You know, we've we've touched lightly on finance, sales, marketing and CS. Obviously product ICP matters immensely to every single team in the company, even like even accounting like thankfully we don't deal with this a lot. But like as a small business owner, I can't tell you how many people I talk to that are other small business owners like complain about how their customers never pay them on time or take 90 days to pay them, and it's like we don't have that problem at all.
EDDIE53:10Yeah, I got to report the other day because there's always some people that are like a couple of days late, and I get this email every single day to Charlotte invoices just to make sure, like nothing's out of hand. And I got an email the other day and it just is like overdue invoices, $0. And I was just like, this is amazing.
EDDIE53:26It's really nice having good customers. Yeah. Just pay you on time.
RACHAEL53:30Yeah. That makes a huge difference for sure. Especially like I used to be like a small business owner just freelancing, right? Yeah. That's you get like 1 or 2 customers like that, and that sucks.
EDDIE53:41And there are correlations with like, who fits that bill and who doesn't.
RACHAEL53:47Yeah, yeah. Well, I think that's all that I had for my questions on the topic. Eddie, is there anything else you wanted to. Any more words of wisdom along to leave us with?
EDDIE53:57I don't know that I have any words of wisdom. These are just tactical tips, like build a machine that can help you generate a little bit more revenue. I think. I think that we've, I think that we've covered it. I think just to recap, sort of the basic idea is think really hard about who your best customers are and how what data points are going to define those customers.
EDDIE54:22You know, keep analyzing and enriching the data until you land on that. Build a machine that can do this, build a machine that can go and take the next lead, the next opportunity, etc.. Enrich the data with the right data points. Feed it into the machine to analyze it. Build an account scoring model which you're going to have to continuously optimize.
EDDIE54:41Turn that into an operation such that you know that's going to influence how you qualify leads, how you create target lists and territories, who you target for expansion, etc., as well as how you qualify opportunities and move them through the funnel, and then build a machine that's going to continuously improve by evaluating every single new input and flagging things for the humans to review, and then decide how often you want to review and revisit ICP and tweak things.
EDDIE55:11I am a firm believer that the more that you do things like this, the more that those tens of millions of dollars you're spending on sales and marketing are going to turn into more revenue. And it's really easy to think like, we don't have time for that. We just need to focus on like landing more customers. But that translates into telling your salespeople, like make more calls, set up more meetings, try to close more opportunities.
EDDIE55:33And if you're doing that with the wrong set of accounts, by definition you're going to have diminished returns. So like, why not take a moment to take a look at who your real customer is in 2026? As of the time of this recording and in 2027 next year. If you're listening to it, then and make sure that you're firing your arrows at the right targets.
RACHAEL55:52And we will most likely have the newsletter on this topic published by the time this podcast goes out. So the link to that will be in the show notes as well.
EDDIE56:02Cool.
RACHAEL56:02Awesome.
EDDIE56:03Thanks for putting this all together, Rachel.
RACHAEL56:05Yeah. Thank you so much, Eddie.
EDDIE56:07Thanks for listening to the episode. If this resonated. Please give us a five star rating and a follow. It helps us reach more people, and you get our latest and greatest content without having to search for it. And if you're looking for hands on help and go to market strategy and or rev ops, please reach out to us.
EDDIE56:21We help our clients with everything from annual planning to improving processes and go to market, implementing systems to support those processes and go to market AI. We're always happy to offer a free consultation to help you identify the best opportunities to improve your go to market engine, with or without our help. You can find us at Union Square consulting.com and the info will be in our show notes.
