[00:00:00] Dan McDermott: All right, so welcome back to the Startup Growth Podcast. My name's Dan McDermot, I'm the CMO here at Vouris I'm joined as always by Kyle Vamvouris, our CEO , Kyle, how are you? 

[00:00:17] Kyle Vamvouris: I'm doing excellent, Dan. How are you? 

[00:00:19] Dan McDermott: I'm doing fantastic. Today, we'll be talking with Nimah Gardideh, who's the the co-founder, president and CTO over at Pearmill. Nimah,

[00:00:26] how's it going? 

[00:00:26] Nima Gardideh: Good, good. Thanks for having me. 

[00:00:28] Dan McDermott: Yeah, great to see you. We spoke a little bit earlier this week, so this will be a great chance to catch up again and go into stuff on the marketing side, whereas on Monday we spoke about the sales side of things. So I'm excited 

[00:00:37] to talk to you today. 

[00:00:38] Nima Gardideh: Yeah, let's do it.

[00:00:39] I'm excited to get into it. 

[00:00:41] Dan McDermott: Super, so

[00:00:43] yeah, Kyle, go ahead. 

[00:00:43] Kyle Vamvouris: Yeah, I was gonna say, this is gonna be super fun. I'm excited. I'd like to learn a little bit about your organization, why should we be talking to you? We know, but our audience doesn't. So why don't you share a little bit about what you guys do over there, what your process looks like, and then we'll use that as our launchpad here to a good conversation.

[00:00:59] Nima Gardideh: Yeah, [00:01:00] thanks. The way I think about Pearmill is that we are a growth studio. Focused on paid acquisition. And what really that means is that we bring on four different disciplines together to do the work that it takes to scale up early stage companies from right around, a hundred, 200 K of ad spend to millions of ad spend per month.

[00:01:21] And the four disciplines are ad operations, so media buying the more classical sort of marketing, creative production, so we have our own studio, we have everything you need there to produce the content. And then the second last, the two are conversion rate optimization. So we, CRO folks who both design and implement acquisition flows.

[00:01:44] And then the last piece is engineering in a large, in a large sense. So there's data engineering, there's data analysis, there's data science. And that discipline is, has been the part that was very different about us at first, and now I think there's more growth teams doing it, or they're combining all these [00:02:00] disciplines together.

[00:02:00] But that's how our company's set up. And, in terms of why it's interesting to talk to someone like me is we just see a lot of spend. We spend hundred plus million dollars a year for startups. We help them scale quite rapidly and it's always useful to see what trends are like and what is working across the board.

[00:02:19] And especially because of our model of experimentation we have like very strong proofs that this specific structure on Facebook is working right now, for instance. Yeah, that's maybe. I'll pass it back to you, but I think that's maybe the context that maybe 

[00:02:31] Kyle Vamvouris: That's 

[00:02:32] fantastic. One, one thing you mentioned there is you take 'em from a hundred to 200 k, it's ad spend to a million, and I think you said a month, right?

[00:02:39] Nima Gardideh: Yeah, these are all 

[00:02:40] monthly spends. Yeah. 

[00:02:41] Kyle Vamvouris: So I'm, 

[00:02:41] I'm curious if any, but what is the biggest difference between a campaign when somebody's spending a hundred k a month versus a million a month? Is it simply, we just shove more money into the exact same campaign? I'm sure complexity is added at a certain point, so I'd love to hear a little bit of your 

[00:02:57] thoughts around that.

[00:02:58] Nima Gardideh: Yeah, that's a good question. I think [00:03:00] there's two general differences. One is that the process in which you have to go through to continue increasing the spend is different at a, at an organizational level. Like how are you thinking about prioritizing your experiments is all of a sudden different.

[00:03:13] And the second thing is, and the reason we cut it off at that, Scale is because you get access to this other part of specifically Facebook and Google that have machine learning auctions where you can use data as a tool for targeting and optimization. And before that level, there's just not enough conversion or volume of conversions going in.

[00:03:35] So it's actually quite hard to use that lever, the data lever and so, we wait until people have figured that part out. There's also other connotations that comes with that. Usually you are much closer to product market fit if you're spending that much money and you're ready to scale and the market is going to actually be receptive to your product.

[00:03:53] But the most important thing is the process part, I would say. Where you have to be a lot more careful. You spend more time debating. [00:04:00] On what to do as opposed to doing it. And the reason for that is because the business is now officially reliant on paid acquisition to some extent. So if we just mess it up, all of a sudden everything falls apart on the business side.

[00:04:12] So it's almost as if you're handling someone's money. It's almost capital allocation at that point. So that's how we think about it. 

[00:04:19] Kyle Vamvouris: Yeah, I imagine it's a big responsibility. I actually, I have a bunch of questions now swirling in my head. Okay. Fascinating to me because I know basically nothing about paid acquisition.

[00:04:26] I know a little bit about some organic channels, but mostly I'm just this dumb brute force sales guy, and that's what I've been doing most, most of my career. So I. I have a lot of questions around this, so I imagine, so somebody comes to you, they're spending a hundred K a month, like we have something working.

[00:04:40] Once you get to a hundred K and ad spend a month, like obviously we have something working. I would imagine then are the people coming to you like, Look, we've tried to scale up past this and we haven't been successful. Like we actually see decrease in results. Or is it that, hey, my hundred K is now yielding less and less, and I need to figure out why.

[00:04:58] I'm just curious what the [00:05:00] motivator to have a conversation with your team is. 

[00:05:01] Nima Gardideh: Yeah, 

[00:05:02] I think that's a good question. There's a few different approach reasons people come to us. One is what you mentioned, right? The people have found some ceiling in their spend and they cannot figure it out, right?

[00:05:13] And this is where it's actually useful to work with agencies because we just see more , we have more information. 

[00:05:19] Kyle Vamvouris: Right.

[00:05:19] Nima Gardideh: And so we can look at your account and say, Oh yeah, you know what? We actually think this is your ceiling. You may see mostly diminishing returns by increasing it in this channel. So maybe our recommendation is to

[00:05:30] try another channel or something like that. The other reason people come to us is that, hey, we figured some things out. Look, we have these pieces of creative that are working this specific set of audiences are working. We just don't know how to take this and double it or triple it every couple of months, and that is much more common.

[00:05:48] The second one, I think if you're stuck at that level of a hundred K, there is probably product market fit issues or fundamental production issues on the creative side, we do come in [00:06:00] and look at those and sometimes it's so obvious that we would do better, that we still take those on. We'll look at it and say, Hey, just you're not doing most of the things that you should be doing at that scale already.

[00:06:10] If we do those, we're probably gonna get a 30% incremental value out of the money. So let us take over. In a handful of cases, we were much better than 30%, right? We had this client come in, they were spending a lot more than a hundred K. We were spending $7 million a month. And 

[00:06:24] Kyle Vamvouris: Wow. Wow.

[00:06:25] Nima Gardideh: And they came to us and they said, Hey, we found, we think we found the ceiling.

[00:06:29] And we thought maybe there is. You have found the ceiling. So they are an interesting company in healthcare. So the market is actually massive, but the account was built by a bunch of folks that we're quite smart and finding a scalable way to do what their thought process was, which is, hey, they have all these like different keyboards they need to go after on Google, and they had structured in such a way that was easy to manage, but not necessarily the most optimal way to get Google to.

[00:06:57] allocate the capital. And so when we came in, we took them from [00:07:00] seven to 1.5 with, for, with the same top line numbers. So we like one sixth their CAC in a 60 day period or something like that. 

[00:07:09] Kyle Vamvouris: Wow. Geez. 

[00:07:10] Nima Gardideh: This made us look good, but actually we made much less money than we thought on that deal cuz we do a percentage of ad spend.

[00:07:15] So it was a tough, it was a tough month, I think. I remember telling the team, like you guys are very good at your job. Too good. Almost. And that's a, that's an interesting problem to face. And we've actually since changed our pricing model quite a bit with a lot of folks to be more incentivized towards the same thing, which is, 

[00:07:31] Kyle Vamvouris: That makes sense.

[00:07:31] Nima Gardideh: A combination of scale and CAC, but there are like best practices that are currently working in the market that we always apply, but those things actually change every six months. I think there, the, in the sales world, maybe a parallel I have and my head of sales recently told me is that usually people used to think that you need to have seven touches with someone.

[00:07:51] Now people are talking about 12. That is like a heuristic you have in your mind of okay, I need to have this level of depth in my sequences. For, let's [00:08:00] say Facebook marketing. There are similar things like, Oh, you need to have 50 conversions per ad set, per week for the ad set to be in its almost optimal state.

[00:08:10] For instance, on the machine learning side, now that number is 20, so that's at recent change because of the iOS 14.5 stuff, Facebook's tried really hard to work with less data, and so they, they've been changing that, right? But that stuff changes every six months. So we have to rediscover these rules, both by literally calling up Facebook and saying, Hey, what's going on?

[00:08:29] And but on the other side, Literally running experiments and getting data. 

[00:08:33] Kyle Vamvouris: Yeah, I imagine the experimentation's huge and I really wanna dive down this best practice thing, but I have one more question but I need to tee it up a little bit. I'm curious from the people who come to you once they've gotten to a hundred K at ad spend a month.

[00:08:45] Cuz a lot of the companies we work with, they vary, but a hundred K is probably less common than people are spending less than that on ad spend per month. So I'm trying to think, if somebody comes to you and they're doing about a hundred k a month in ad spend, were they typically working with another [00:09:00] agency before?

[00:09:01] Or were they managing the ad spend themselves in house and they just self-taught how to do it, got themselves there and decided to go to a 

[00:09:07] real firm. 

[00:09:08] Nima Gardideh: Yeah, 

[00:09:08] that's a good question There. I, we see a mix. My personal preference is that you've been trying it by yourself. 

[00:09:15] Kyle Vamvouris: Okay. 

[00:09:15] Nima Gardideh: And the reason I actually generally recommend this to marketers and founders of early stage companies is because what'll happen when you try to do it yourself, you start building the

[00:09:26] understanding on how to have these conversations about capital allocation in on these different channels. So then when you talk to agencies, you can actually judge how good they are. So if you've gone immediately to an agency and started doing it unless you already know how these things work, I would say you're not in a great position to continue interviewing new agencies.

[00:09:48] Kyle Vamvouris: Yeah, that makes sense.

[00:09:49] Nima Gardideh: It's same thing probably with sales, right? You want the founder to be selling. And then hiring the sales people and tell 'em, here's what I've learned so far. And totally being able to actually hand off in a super useful way. 

[00:09:59] Kyle Vamvouris: Yeah, 

[00:09:59] You run [00:10:00] into this with founders who try to hire salespeople way too early and you're like, you should really be doing these calls.

[00:10:05] Are you booked out, until, for two weeks? No. Okay, let's get you back. Busy . Yeah. Okay. So that's helpful. And this is like the real question I wanna ask then. Cuz I was curious to see if what you preferred more agencies, if agencies make sense earlier or self learning. I tend to lean more to self learning for most things in general.

[00:10:21] Cause I think it's important to have some kind of competency for the exact reason that you mentioned. But I'm curious then, how do you suggest people learn how to run ads themselves? 

[00:10:32] Nima Gardideh: Yeah, that's a really good question. I think there are companies out there trying to teach you, right? So if you go to, Demand Curve has some decent courses there.

[00:10:40] Reforge has some decent courses, but the reality is that the best thing you can do is to actually spend money on it and learn what the different knobs are that you can move up and down and mess around with it. It's not rocket science, it's not like it's investing. It's not as [00:11:00] complicated as actually investing in, let's say, stock market.

[00:11:02] Because there's a big difference between stock markets and these auctions. The auctions want the same thing that you do, which is they want you to get users at the cheapest costs possible. It's actually how they're done. They use them. The phrasing they use is discounted pacing algorithms where they try to get you the cheapest clicks as much as po possible for as long as possible, so they're actually on your side.

[00:11:27] So you should be to some extent getting some results immediately and then iterating from there. And the second thing I want everyone to get comfortable with is losing money to learn because you literally are paying to learn, and it's just the only way to do it. And so when you get into yourself, you inherently end up losing some money and you feel the pain, which hopefully gets you to a point where you are more structured about your experimentation.

[00:11:53] Which is what you really have to be very good at. It's the only thing that I tell people is, can you tell me exactly how you [00:12:00] experiment, how you get results in a way that you can trust these learnings over time and compound the learnings. And most people come to us having spent maybe hundreds of thousands, or in that example, 7 million, not having a process in place that could tell me, here's exactly what we've learned, here's what the truths are about our market, about our user base, about our creative

[00:12:21] that works. 

[00:12:22] Kyle Vamvouris: Yeah. I wanna say something, and I've been hoing this, so Dan, I'm gonna kick it off to you in a minute, but something you just said just made me think a little bit about the mindset of an entrepreneur where you mentioned a client that came to you spending $7 million a month in ads, and you were able to get them down to 1.5 million with the exact same results.

[00:12:42] I think a normal person would go, I can't believe I've been wasting $5.5 million every month for the past, let's say six months when for a real, like for a real entrepreneur, that's now just a story. [00:13:00] Hey, I was wasting 5.5 million a month for six months until we worked for this company and figured out how to not, And I just think I, I wanted to call that out.

[00:13:10] Cause I think it's just I dunno, I like to highlight the different, in different thought processes that like entrepreneurs and founders like tend to have. And that just stood out to me as 

[00:13:18] something that's. 

[00:13:18] Nima Gardideh: Yeah I would say this is not something, I don't recommend you wasting that much, but but there, it certainly was in the time and the it made sense at the time for them, the unit economics actually did work at that rate.

[00:13:28] We just made them much more profitable as a business. The capital markets were such that, they didn't care that much, right? Yeah. Yeah. So they really wanted the top line growth and they were getting rewarded by it, by the capital markets where they were getting more money raised every six 

[00:13:42] months.

[00:13:43] Kyle Vamvouris: Very interesting point here, which is maybe things have are forced to change a little bit now cause we're seeing some pullback in the capital markets less that you're money being thrown around that leans people being more thoughtful about

[00:13:55] the different channels that they're using to generate customers and they're thinking a lot more [00:14:00] about their CAC payback. I talk to VCs pretty often. , and I can tell you a lot of 'em are like, I don't care about payback. Period. At the early stage. At the early stage when you're talking about like Series A, they're like, ah, 18 months at the high end, now it's probably gonna change a little bit and you gotta be a little bit more thoughtful.

[00:14:16] Nima Gardideh: I remember thinking 18 months was ridiculous, even like eight, 12 years ago or something when I first started. There's gotta be something that, some part of me is, because I'm like from the East, I'm from Iran, and so we don't actually, we don't have a lot of trust in institutions over there, so you just don't assume that you can get more money at all, on any numbers.

[00:14:34] So you, you want to break even as soon as possible. There are really good numbers. I think Craft Ventures has been putting up, putting out some really good content around now, what should you expect, maybe as a B2B company, for instance. And they have the burn multiple metric that I really that encompasses a lot of CAC and LTV under one singular metric you can look at.

[00:14:52] But yeah I agree with you. I think the capital markets are changing. I think we've seen this in our ad spend already. We maybe seen like a [00:15:00] decent decline from all our clients, which we're happy they're doing it. Sometimes people are willing to spend more even at the diminishing returns, just because they wanted more top line, not necessarily as focused on the bottom line.

[00:15:12] And I think this shift is very good for the companies themselves, honestly, right? And the mentality that goes behind building good businesses. 

[00:15:23] Dan McDermott: Good business is good processes. Yeah, for sure. That's a great way to look at it. Just on that point, I'm really interested in hearing, even if it's not scientific and not perfect, but is there a general couple of numbers that you would look at that are healthy when you're looking at specific metrics?

[00:15:36] So you mentioned the burn metric, but is there in a more simple level, let's say that I'm a founder and I've been running my own ads with sort of half decent results for the last year or so. , Are there numbers that I should be paying attention to. Specifically, let's say Facebook ads that that pop up and you say, Oh, this is a healthy campaign, or this is not.

[00:15:52] Nima Gardideh: Yeah, I think that's a good question. So basically what we look for there to see if there's potential for scale, right? Is how consolidated are [00:16:00] things. So if you have four or five campaigns and that's how much you're spending and you're still getting decent numbers, that's a good, very good sign, which is interesting, right?

[00:16:08] So the more campaigns and ad sets you have and you're still profitable, or let's say close to profitability, that's a really good sign because consolidation helps. Because you're now trusting the algorithm more to do the Sub- targeting for you. So you're consolidating your audiences and putting them under one big audience.

[00:16:28] It's called stacking audiences. And so when you stack audiences, what ends up happening is that you're creating a larger overall audience and giving the machine learning model more leeway to choose from that overall audience. So if you're not consolidated spending, let's say a hundred, And you're getting decent results.

[00:16:46] We look at it and we're like, Okay, immediately we're gonna have a huge impact here. We're gonna be able to scale you up immediately, right? If you already consolidated and like you have a handful of ad sets and you're barely making it, the question then goes to how big is [00:17:00] the size of your audiences currently in the existing ad sets?

[00:17:03] Are there spaces for us to broaden them up? So the broad strategy, really what it means. Let's say you're going after 10 million people right now. That's your general market. Even if you're b2b, let's say you're B2B and even smaller, 500,000 people is the size of the audience, and you're getting something like 50 leads a week right now and under one ad set.

[00:17:25] What I would tell you to do is actually be less targeted and let Facebook do the work of targeting for you, using the data, and then what your job becomes is what signal do I give Facebook. So then they learn that this is the type of person I'm looking for. One of the examples I have here is ramp.com is one of our clients, and they have this very long survey you have to fill up before you become a lead.

[00:17:52] And based on the answers of that survey, we send a separate event up to Facebook. And then, so if you have answered the questions correctly, [00:18:00] you're like considered MQL, we're trying to get you MQL as much as possible. And MQL being marketing qualified lead and the whole apparatus is set up to optimize towards that event.

[00:18:13] And the targeting is over time increased more and more in the size of it is increased at some point. Some of our even B2B brands, The targeting is all of the United States, even though they're a B2B brand.

[00:18:23] Kyle Vamvouris: Wow. 

[00:18:23] I have a follow up question here. I wanna get a little tactical. So in that example where you have a survey to fill out and you send the questions back How, are they using type form and then within, like it's integrated with Facebook.

[00:18:36] I've never heard of 

[00:18:37] that before.

[00:18:37] Nima Gardideh: Yeah. Typeform turns out to be too slow. It loads too slowly for ads cuz one of the other problems is you need subsecond load times on ads because otherwise people bounce. So generally speaking, for acquisition, people are using this tool called Form Short. We're big fans of it.

[00:18:51] I think almost every single one of our clients uses it. And then there's some custom JavaScript that we end up writing there based on the answers of the form sort, Let's say we send us [00:19:00] something up or we hit up clear bit, pull in the data for that user after they put in their email, and then use that information plus the answers of the form and then send it up.

[00:19:09] There's now a couple companies doing linear regression. On, on, on all of it. So they pull in all your pageview data, they pull in your clear bit data and then give you a score. And if it's above a certain score, then they send you send that event up. In our experience, that sometimes works for consumer specifically, not so much for b2b.

[00:19:26] It's been a I mess on the B2B side. 

[00:19:30] Kyle Vamvouris: Wow. Unbelievable. It's super sophisticated , right? Cause I'd never heard that before. 

[00:19:33] Wow. 

[00:19:33] Nima Gardideh: It's new. This stuff is like maybe two years old. I would say we've been doing it for maybe three or four. But not everyone's caught on this. It's hugely helpful.

[00:19:41] Kyle Vamvouris: Oh, I imagine. Yeah, cuz you know, I've always heard that the Facebook algorithm is really good, figure. Actually that leads me to another question for b2b. Do you are, do you do a lot of Facebook or I know people like LinkedIn stuff too. Does it matter? I'm just curious. 

[00:19:54] Nima Gardideh: Yeah, so we've been doing Facebook for B2B since we started.

[00:19:59] It's actually our [00:20:00] original niche was b2b because no one was doing Facebook spent at the time. This was about four years ago when we started. And no one believed that you could do it. And so at the time, a lot of products that out now exist didn't exist, and we did this manually. But effectively, there are companies like Clearbit metadata, Dot io or say primer.com.

[00:20:18] What they have done is they've gone there and bought all the lists that you could possibly buy. So they have the massive lists of everyone in the US, let's say for that works at X company and Y company. And you tell them, I want CEOs of SMBs between 25 people and 500. And instead of giving you the email address, what they've done is they've gone and bought data from Live Ramp and all these other places and enriched those work email addresses with the personal email addresses.

[00:20:45] And then they upload that list to Facebook for you. So you're effectively getting LinkedIn level targeting on Facebook. 

[00:20:52] Kyle Vamvouris: Wow. 

[00:20:52] Nima Gardideh: And so it's become quite useful to use it and for a while. And even now is cheap, much cheaper than LinkedIn to [00:21:00] get the same lead. Because people, even though they're, professionals, they're on Facebook and Instagram, it's just totally still people, , 

[00:21:05] Kyle Vamvouris: So I have two questions here. One is about the classic argument of, oh, when they're on Facebook, the, they're in leisure mode and they're less likely to take action when they're on LinkedIn. They're more in work mode. But before I get to that and I'll ask that better in a minute, I'm just curious.

[00:21:20] So if somebody wanted to do this themselves and they were gonna run Facebook ads, cuz one of the things you said earlier is it's better to learn it yourself? Do you recommend them get clear bit and set it up that way from the beginning, or should they first experiment with just like getting conversions through Facebook?

[00:21:37] Is Facebook where they start, that kind of 

[00:21:39] thing? 

[00:21:40] Nima Gardideh: Yeah, that's a good question. Generally speaking, I think it depends on what type of product you have. If your product is a, is the kind that people already know that they have the problem and people already know that there are solutions out there. You wanna go after high and 10 channels.

[00:21:55] So that really effectively means search, right? So people are searching for it, go get 'em [00:22:00] there. I see. Interesting. And that's the part of the, let's say if you were looking at a demand curve of sorts. The first 10% probably are gonna be on search because they are looking for it, they're in market.

[00:22:11] They're likely, sometimes even early adopters can capture it now. And that's all good news. There's a slight caveat to that is that if it, you're in a super competitive market, the keyboards might be so expensive that it just doesn't make sense for you to do it, and then you have to differentiate by not being on search and just going straight to some form of social discovery channel.

[00:22:27] And then that's when I would ask you maybe start going on LinkedIn. And the reason for this is because even though LinkedIn's gonna be more expensive, Clear bit, and all these other tools are actually also expensive, right? They're like in the tens of thousands of dollars per year to purchase.

[00:22:41] So when you're in the earlier stages, LinkedIn gives you the targeting you want, but then what you wanna do is just focus on clicks there and then retarget them on Google Display and Facebook.

[00:22:51] Kyle Vamvouris: I see. 

[00:22:51] Nima Gardideh: And so you're getting cheaper clicks in and then retargeting those folks. Eventually, as you're scaling up, you want to open up the Facebook channel.

[00:22:58] Then you go out there and [00:23:00] buy some of these data tools to then get the targeting apparatus. Yeah. Yeah. 

[00:23:04] Kyle Vamvouris: That 

[00:23:04] this was, that was my favorite part of this conversation so far. That was awesome. Yeah, cuz you know, there's so much. I don't know. It's tough in the marketing space. There's a lot of people who are very good at marketing, selling you how to do marketing, and they might be very, I don't know.

[00:23:20] I just think of all the internet entre internet advertisers out there, and you just get a lot of crap, and it's Oh, you gotta do it this way. This is the best way. This is the one seeker that it's gonna take to scale your ads, blah, blah, blah. 

[00:23:31] Nima Gardideh: Ugh. 

[00:23:32] Kyle Vamvouris: I know. I'm 

[00:23:33] sure it's

[00:23:33] Nima Gardideh: they make us look bad,

[00:23:34] it always get some of those ads sometimes from these like basically info SaaS companies trying to sell you the best next marketing thing. And it not only is the most like high, like super cringy, it also just devalues the whole industry. And it's a general reminder and we talk a lot about this, that there is like a historical bad faith generally in our industry, right?

[00:23:57] Kyle Vamvouris: Yeah. 

[00:23:57] Nima Gardideh: We're named after one of the first [00:24:00] ever advertising campaigns by Pear soap which has become the Unilever of today actually. And they had this artist, his name was John Millais, which is where the mill come mill comes from, right? So Pearmill and he had this painting and they used it, but that same company grabbed a series of John Millais painting

[00:24:22] and painted these babies black and had these soap commercials, Not commercials, but ads that the soap was getting rid of the black of their skin. 

[00:24:31] Kyle Vamvouris: Oh geez. 

[00:24:32] Nima Gardideh: And so we actually have this, we tell this story to everyone in our company that when they first joined, about where the name comes from, and we actually knew about this when we named the company this, because we wanted it to be a reminder that we can screw.

[00:24:47] And that the messaging we use the put things that we put out there millions of people see. And we should be very careful about that too. So I think that the industry generally has a lot of bad historical things sitting there, but also [00:25:00] there is still people that with bad faith in it, and they just, therefore the clicks and getting money out of people as soon as possible and fast as possible.

[00:25:08] Dan McDermott: This is something we, we talked about on Monday a little bit because we have a similar story on the sales world too. If you bring up sales to some people, they have a negative reaction, many people , and I probably was one of those people, . So it's really interesting to go in and see, okay, how do you add value to a space where you change something you really do try to act ethically and in a way that is a sort of win-win for the business and for the user.

[00:25:27] I'm curious is when you come into companies, do you work with them specifically around this topic? Like how do we maintain, for example, in our messaging, how do we keep it so that we don't hopefully dip into those those types of mistakes? Do you do any of that sort of training?

[00:25:41] Nima Gardideh: Yeah. Yeah. So we we have essentially like. Two levels of filters. One is that at the company level, we just will not work with people that we think there is questionable things to begin with, right? So I have this example I give because it always like, to some extent, hurts, but it's the right thing to have done for us is [00:26:00] that Juul came to us like a couple of years ago and we immediately rejected automatically, right?

[00:26:05] We're like you just don't fit. You're not ethically sound, we're just not gonna work with you. There was another weight loss company that came to us, they were spending millions. We just weren't willing to work with them because it felt like it was snake oil. So there was like that level of, hey, the founders of the company.

[00:26:19] And then actually when we are unsure about something, we share it with the whole company and say, What do you all think? Do you think it makes sense for us to work with this company because here is the areas in which we're unsure about. And then the second thing is, after we start working with them, We're currently working on this where we have, we want them to sign an ethical code, but what we do is there is one of our kickoffs, we explicitly discuss ethics but we also want to have people literally sign something, just more of as a gesture of, yes, we agree that we want to be ethical.

[00:26:48] Dan McDermott: I think that's a really special thing. That's something I, I've heard many people talk about, an intent to do such a well to just protect the ethics of a of a campaign. But to actually sit down and get them to, [00:27:00] name on paper the gesture is quite important.

[00:27:01] I think it's quite a, it's quite interesting, I think. Yeah I've, I have many questions on this, but I think one one more question that comes up for me that is a hot button topic within the little space of marketing that my foot is in is whether or not, for example you should have gated content.

[00:27:15] And, that leads back to should you have content, that you're targeting with ads. Do you have a particular stance. How much protection a a company should be giving its content when it comes to what they're 

[00:27:25] producing? 

[00:27:26] Nima Gardideh: Yeah. I think you're not gonna like what I'm gonna say about this one.

[00:27:28] I, we basically completely moved off of gated, gated content over the past year or so for all our B2B clients. And the philosophy is this, right? We were talking about on, on, when we talked about sales That everything, every touchpoint you have with the user should be net positive that you are trying to provide value when they click on your ad, when they see the email from the salesperson that comes in, that you're trying to make their day better and make their work easier and their job easier.

[00:27:59] And so [00:28:00] when you have gated content, you're effectively asking for something before you give them the value. And that feels philosophically wrong. And it also turns out performance wise, it's better not to gate it because what happens is your auctions because people are taking the action, you want them faster, which is to click, and they go around and they stay longer.

[00:28:19] These ad networks like Facebook and Google will think this is valuable. I wanna send more people down this route. And then so you get cheaper, clicks. And then, so you can retarget these folks that you now know are highly interested in what you're doing, and you create what we call we call these like creative funnel flows, where we just know that you are learning about this topic inside of the company.

[00:28:41] So let's say like one of your futures, there's a series of block posts, right? And then now after you pass two or three of them, we then ask for delete and say, Hey, you wanna schedule a demo? So we push you through this flow of creative before we ask something of you. And it turns out to work much better than gated content.[00:29:00] 

[00:29:00] Kyle Vamvouris: So let me ask, I want, I wanna dive a little deeper into this to make sure I fully understand. So let's say someone's scrolling through Facebook, they see an ad that's I'll just use my stuff cuz that's in my head right now. But we have this one article that Dan just updated for sales compensation guide.

[00:29:17] It works really well for us. We get a lot of leads from it. So let's say you're scrolling through Facebook, you get an ad saying what should your sales comp plan look like? You click on it, it goes right to a blog post. That blog post has a template for them to create their own sales comp plan.

[00:29:29] That's totally ungated. Then, is the plan from there to re-target them with adjacent content to some, what somebody may be looking for if they're working on a sales compensation plan, and then on the third time they have come to engage with the site, then we prompt them to book a demo. I'm just curious, tactically, is that what you're talking about?

[00:29:49] Nima Gardideh: Yeah, super close. So let's say they click on the first ad that goes to the piece of content. Now we can retarget them, right? For two different ways. One is that anyone that's visited that page [00:30:00] on your blog can be retargeted. The other thing is anyone interacting with an ad can also be retargeted.

[00:30:05] So that content could be fully on Facebook. You don't even need to them to leave if it's a tiny video, giving them value. And they're gonna listen to that 32nd video. That in itself is useful. Now they're in our universe right now. If we feel like, okay, the second step is they are looking for compensation.

[00:30:23] Maybe the second thing you wanna show them is recruiting. How to recruit good sales people. And then, so now you wanna show them that content. And the third thing to show them is effectively a lead form, direct to lead form. No more value. Hey Hey, we want to talk to you. We think we can solve your problem, and here's the solution we have.

[00:30:38] So you walk them through a series of steps. 

[00:30:40] Kyle Vamvouris: So do you try to stay on platform like that as much as possible? 

[00:30:43] Nima Gardideh: It's dependent on the specific product and solution, but ideally you're spend you're on platform for that first touch because it's just cheaper to get just your views than clicks.

[00:30:55] Kyle Vamvouris: . 

[00:30:55] Yeah. So super interesting. So now if somebody is much [00:31:00] earlier than, cuz one of the things you mentioned was search being the probably the best place to start. Do you take the same type of approach where it's hey, somebody types in sales compensation and you just send them straight to the blog post, that's ungated or do you then gate because you're early and that matters?

[00:31:16] Nima Gardideh: Yeah, it totally depends on the level of intent you think that set of keywords have. 

[00:31:21] Kyle Vamvouris: Oh, interesting. 

[00:31:22] Nima Gardideh: So if you think it's super high intent, then you go straight for a conversion, right? If you think it's a lower intent thing, then you go for more clicks and you want them to, either re targeted with maybe display or Facebook or other channels that you have, right?

[00:31:35] So it all does depend on the level of intent, and that's actually how you generally want to group your ad spend on Google is based on intent. It's actually new. Over the past year, they've gone very good at this on their. . 

[00:31:48] Kyle Vamvouris: Yeah, this is, yeah, this, I love this conversation. I, now that I think about it, Dan, this is the first time we've actually we've, but even myself, you've probably talked to people like this, that I've just talked to somebody who like, understands B2B marketing [00:32:00] very well.

[00:32:00] And I'll just tell you Nimah, a little bit of my background is, I grew up around the internet marketing world a lot. You when I was young, my dad owned a bunch of medical centers and he would go to a lot of these seminars to learn how to do marketing. And he's actually told me, he said, once I figured out how to do marketing for my my medical centers, it totally changed my life because I could now turn on.

[00:32:23] Get a bunch of customers. If I was too busy, I could shut 'em off. Like he, he said he had full, he felt like he had full control and he learned it all from Dan Kennedy. And Dan Kennedy I think for the most part he's held in high esteem. I'm not sure if that's true. Totally. But one of the old school copywriter, marketing folks I think mostly by Postcards and stuff.

[00:32:40] But anyway, from there, like I, it interacted with a whole bunch of internet marketers, so I'd been around that world for a really long time, like the Frank Kerns, like all of these people and a lot of them questionable ethics for sure. So when I speak to people, I get a lot of that a lot of that feedback of like

[00:32:56] quick conversions, just blast through and not [00:33:00] worry about the intent and or anything like that. And use really bold, audacious language in order to get a conversion quicker and this type of stuff. And I have a weird feeling around it, and it's nice to talk to somebody who, who knows the other side of it.

[00:33:16] Cause I feel like I talked to a lot of people who don't actually know and they say, Oh you should just do it this way cuz they're repeating talking points from some other person like yourself I'm sure who actually knows what they're talking about. But this is the first time I got to talk to somebody who like actually gets how the whole thing works.

[00:33:28] So this has been really refreshing. . 

[00:33:30] Nima Gardideh: Yeah. And I think you know what's cool? I think talking to Dan was very helpful to know that you all think similarly. Because ultimately the way to build some of these fast growing super valuable companies is to not do any of that. Not do they super aggressive things because you're trying to build very long term relationships with these customers.

[00:33:51] And so every touch point has to be super thoughtful. And yeah I agree with you that I'm glad this is interesting too, which is great. But also I think it seems like we have a very [00:34:00] similar philosophy on how to grow these companies, right? 

[00:34:02] Dan McDermott: We definitely 

[00:34:02] do. 

[00:34:03] Kyle Vamvouris: Go ahead Dan.

[00:34:04] Dan McDermott: I think just real quick on this the idea of how quickly things change and having, we all know where things are going is basically all the algorithms out there getting better and better.

[00:34:13] All the major ones getting better and better what they do, and understanding the nuance of human intent. I think that's the best word to come back to this. And I think it's really interesting to, again, why it's valuable to have a third party come in and be able to tell you what is happening in the industry, what are we seeing across the whole board than within your niche, than within, versus maybe a couple competitors.

[00:34:31] I'm curious, maybe could you give us a couple examples that you've touched on a couple, but what are some of the big changes that have happened maybe in the last six months or maybe the past year that have shifted the landscape a little bit? 

[00:34:42] Nima Gardideh: Yeah, I think it, it depends on the channel.

[00:34:44] So I would say, For Google, this intent grouping is the biggest thing. They've been talking about it for three years, but they cracked something in their machine learning models over the last six months, I would say, that have has made it so that we [00:35:00] actually can compete well with the old structures and beat them with it.

[00:35:04] And when I say these things, I actually have so much conviction because we run an experiment across like all of our clients, all of sudden. When we are getting close to high level of conviction on a new structure, and then we just prove it, right? Cause they're just saying, Oh, it worked not just across these three clients, but across millions of ad spend across different industries. And B2B and B2C and marketplaces. And look, it works like there's a fundamental truth here right now, right? So the intense thing on Google seems to be like the biggest change. Although it's been around for a couple years, it's finally getting to a point where it's more performant on the Facebook side.

[00:35:43] Unfortunately there has been like, it's been harder, right? So what's happened is, and if you have heard about this, is that their iOS 14.5 changes have made it such that they no longer have per user information. At best, they have aggregated information with a 48 [00:36:00] hour window. What they have done is they tried to do the same thing, which is like trying to get the data in, try to sub-target the bigger audience that they have on their end and match it with the people that they think bought within those days, and they've gotten better at it.

[00:36:13] So there was a huge hit. That if you were a mobile company you, you got but if your web, you still got ahead, but it wasn't as bad. The main problem with Facebook is the early stages. So you are effectively penalized as an early stage company right now because you cannot get enough data into the system for you to take advantage of these machine learning things.

[00:36:33] So as our audience says, as much as possible, and then relying on the optimization event, right? So let's say MQL is your optimization event if you're a B2B company, and then creative to sub-target the larger audience for you. And so the creative optimization problem is that the, I would say the most intellectually interesting part of this game where I want to go after CEOs and CFOs, let's say, for a product, [00:37:00] But the audience.

[00:37:02] Is best to be stacked. I want the CFOs, and CEOs on the same audience. So how do I make sure the CEOs see the CEO content and the CFOs see the CFO content? And so effectively you can do this more or less. You're gonna get some pollination, right? Cross pollination. But you can do this by producing enough different.

[00:37:24] Vibes of creative that then Facebook's algorithm fits to the sub-target of the audience for you automatically. So you'll literally, I can show you data that says , as long as the vibe is for like the CEO and it's just like maybe a.\ Cinematography approach or copywriting approach, or typography approach or something like that.

[00:37:45] You will see more CEO leads come out of those ads, even though some CFOs will come out of it. But mostly you will see the target audience that you want cuz Facebook reads all the creative content and feeds it into the ML model. 

[00:37:58] Kyle Vamvouris: That's 

[00:37:58] incredible.

[00:37:59] Dan McDermott: That's [00:38:00] ,that's fascinating. 

[00:38:00] Nima Gardideh: Cool thing. It's very fun.

[00:38:01] Kyle Vamvouris: It's 

[00:38:01] super cool. Yeah. And that's one of the things I really love about marketing is it's a really data, it seems, it seems like the further up the funnel you go, the more data driven you have to become. . It's actually one of the things that we're very good at is understanding this, the data, along this, the the sales process and the further you go up, once you go to SDRs who are doing cold outreach more data driven, And once you go to market, it gets even more data driven, which is just totally fascinating to me.

[00:38:26] I have a an actual, another question here based on something that you said earlier, and I'm trying to find my notes here. You mentioned that at a certain point you mentioned as you start running ads, you should see results pretty quickly. And I'm just curious, how do you define results in, in this world where there is no or where, since it's not gated, I would imagine the results technically come later.

[00:38:52] How do you know if you're on the right track, if you're not gating your content? 

[00:38:56] Nima Gardideh: Yeah. So if you're spending early this creating content and gating [00:39:00] people and all that sort of stuff, I think is too early for you, right? Because of the earlier demand curve conversation we had, there was people who are ready to buy now.

[00:39:08] Let's find those. So your goal in the beginning should be, am I generating leads? Are these leads converting immediately to MQL and SQLs and closed ones? Later on, then it becomes like a more nuanced conversation about am I trending towards the right costs for mql or cost per sql, or cost per pipeline, or whatever the metric it is that you are looking at, which all really means like how much is gonna cost.

[00:39:32] to close a deal. Yeah. And you bake that all in into your models. 

[00:39:36] Kyle Vamvouris: Yeah. And it was going back to what you said earlier about the higher the intent, the more focus you should have on conversion, and if you're just starting out, you should probably focus all your energy on the higher intent because that actually will 

[00:39:48] convert.

[00:39:48] Nima Gardideh: Yeah. Okay. The problem with it is, it's just like there, there's not a lot of scale in that. The keyboards that are super high intent, there's less people searching for the problem because maybe they already have another solution. You have to sell them on it, [00:40:00] or they don't even know that there is a solution out there for the problem that they have.

[00:40:03] Kyle Vamvouris: Yeah, 

[00:40:03] Yeah. Totally makes sense. So I wanna talk a little bit about attribution. So I get that if you're really early, focus on high intent, which means more conversion focus. As you start getting more mature in your ads. Ads you're gonna want to focus a little bit more on other ways of generating interest, which might be some less gated content.

[00:40:21] And it'll be a little bit more, a little bit longer before you start seeing conversions not as quick. So help me understand A bit about attribution. So one of the challenges we run into often, especially with companies that are a little bit larger, is they don't really know how to attribute their leads to specific campaigns.

[00:40:39] So an example of this, and I'll just make up is going back to what we talked about earlier. We ran an ad for sales compensation or, and it was native on Facebook, and then we sent him another video that was about hiring sales people. Then the third thing was actually a conversion. Now if you start adding events and you start adding off of [00:41:00] Facebook, maybe some LinkedIn, maybe some Google, There are a lot of things this person might have saw.

[00:41:05] What do you consider? What do you attribute the revenue of that deal, that closed to? Is it whatever actually converted them into meeting with somebody, or do you go all the way back to the first thing that they saw? 

[00:41:17] Nima Gardideh: Yeah. Great question. Can we extend this podcast by two hours? Yeah, clear our calendars.

[00:41:25] I think that there's two things to discuss here, especially in the world of b2b. One is basically the model, right? What is the model? We're looking at a first touch, last touch. Is it the multi-touch model? Is the weighted model. And we can talk about which ones that make sense for which companies.

[00:41:39] Maybe in a second. But the second thing is this is the hard part. Is that you have to attribute at the group level for b2b, which sucks effectively, right? Like I click on an ad and I am an engineering manager at Company X, and I'm like, Oh, this is a cool product. I call my VP and I'm like, Hey vp, we gotta buy this [00:42:00] product, right?

[00:42:00] And this VP is like, No, I don't think it's the right time. She sees an ad a week later. Then she clicks on it and then sees some more documentation directed for her because she's the vp. That is more interesting information for her. It's Ooh, this is interesting. We should actually start talking to them and then get converted.

[00:42:18] So now you have effectively two people that were involved 

[00:42:22] Kyle Vamvouris: Yeah.

[00:42:22] Nima Gardideh: In this purchasing cycle. So you have to attribute the group level. I don't know of a single software that does as well. We do it mostly by hand, as in like we write our own queries. We group at the company level and look at all the clicks and create a multitouch attribution model at the group level as opposed to the user level.

[00:42:40] Kyle Vamvouris: And what 

[00:42:41] do you mean by at the group level?

[00:42:43] Nima Gardideh: Like at the company level? 

[00:42:44] Kyle Vamvouris: Okay. So you're just saying you're attributing that person to the company? 

[00:42:47] Yeah. That has 

[00:42:48] conversion. 

[00:42:49] Nima Gardideh: I'm attaching it to a company. So like I'm grouping all the clicks. Not at the user level, but at the company 

[00:42:54] level. 

[00:42:54] Kyle Vamvouris: I see.

[00:42:55] So in that case you're almost less interested on the person who is responsible for the ultimate [00:43:00] conversion? 

[00:43:01] Nima Gardideh: Yeah. Cause what I'm actually interested in is the ultimate goal, and usually that takes multiple people , it usually takes a bunch of people and one of the audiences you can generate from Clearbit, for instance, and I love that they did this, is you get one person to click from one company.

[00:43:14] They autogenerate an audience for everyone else in that company for you. 

[00:43:18] Kyle Vamvouris: Wow. That's the, 

[00:43:19] Nima Gardideh: So you can start targeting everybody else with the messaging that makes sense for them based on their seniority, so on and 

[00:43:24] Kyle Vamvouris: Gosh, that's super interesting. So let's say you're doing it on the group level.

[00:43:28] Actually, let me, you said something interesting there. Gosh, I have a whole bunch of, We gotta do one of these again. 

[00:43:35] This is super 

[00:43:36] Nima Gardideh: We gotta do, adjust that 

[00:43:37] attribution, right? 

[00:43:38] Kyle Vamvouris: Yeah. Yeah. I love, Yeah, and I love nerding out about stuff in general, but, and this is just fun. So why don't you care about the, that individual I get you, like you said the ultimate goal is the conversion.

[00:43:47] Totally. Get you there. Does it matter? Like, how am I trying to ask this? I'm gonna ask this question in a really dumb way and just fix it for me after I ask it. Doesn't all that matter is that the VP of the sales was the, or the VP [00:44:00] that you mentioned, she's the one who converted like at the end of the day shouldn't it 

[00:44:03] just be attributed to her?

[00:44:04] Nima Gardideh: I think what you want is you wanna attribute to an ad, right? 

[00:44:09] Kyle Vamvouris: Okay.

[00:44:09] Nima Gardideh: And so they all work together to get the thing done. And, you guys are in B2B sales, it's just not one person making decision, effectively you have to think about the group that is involved in making a decision.

[00:44:20] And as, as a marketer and all these actually systems, Facebook and Google, ev, everything is set up not to think about groups and actually to think about people. So you need to solve for that on your end. You still, as a marketer are thinking, Okay, I want the VP to see this and I want the engineering manager to see this other one.

[00:44:38] But when you're trying to attribute and figure out how much it costs to get something done and you have to then think about the group level because you can't possibly count that twice, right? You can't count the engineering manager converting, and you cannot count the VP converting at the same time, you have to count that as one conversion.

[00:44:55] And then, so now I wanna group all the clicks that are associated with that together. So then I can think [00:45:00] about, Oh, you know what? There is value in the middle of the funnel clicks that I get, even though it's not the person that's the economic buyer that sees it, because in the data it shows me, and this is where weighted attribution can come in the data, I can see that if I get a bunch of clicks from low level people in the middle before the conversion happens, I'm able to convert at a higher rate, right?

[00:45:25] That's when you start doing all these like more nuanced and Ways of attributing. 

[00:45:31] Kyle Vamvouris: So 

[00:45:31] do you do the same thing with the ads then? So let's say there's 10 different ad ads that went to people at working at that company, at various different titles. Then if somebody's asking Hey, what ad is the most effective for us?

[00:45:43] Are you then saying it's actually which ads are most effective for you? Because we group them based on how, based on the conversion of conversion and be, 

[00:45:52] Nima Gardideh: We grouped them based on the persona you mean? Yeah. Not the conversion event. A conversion event is gonna be probably [00:46:00] similar, although in the beginning, if they're in the middle funnel, then it's not the conversion event.

[00:46:04] So you might Yeah, you're right about that. It's, we're grouping them based on the part of the funnel you're falling into as an overall company. So you're like, in the early stages, we're trying to warm you up. 

[00:46:13] Kyle Vamvouris: So if you're if you're looking at an effectiveness of Somebody's ad campaign and somebody as asks you like, What's the most effective ad?

[00:46:21] You're probably looking, I'm asking, are you looking at, hey, The VP she converted based on this ad and it has the highest convers converting for all VPs. Therefore, this is the effective ad. But in addition, all this other stuff was also taking place, which helps. It might but not be as directive a line, but does help.

[00:46:45] So you looking at almost in 

[00:46:47] pieces. 

[00:46:48] Nima Gardideh: In pieces and you give him weights, you're saying, You know what, maybe this is 80% important, but these other things should get some portion of the value of what we're doing. 

[00:46:56] Kyle Vamvouris: Yeah. Very smart way of doing it. Wow. Super interesting. [00:47:00] Alright, that's it for my side, Dan.

[00:47:01] I'm sorry, . I went down a bit of a rabbit hole there, but I just find this thing so fascinating and I don't know. I don't know if you see, I see a lot of this stuff, you just see so much content on LinkedIn and whatnot of people talking about like ungated advertising. And honestly, I've seen a lot around throwing out attribution all together, which has always just seemed like such a silly argument to me.

[00:47:22] And so it was like fun talking to you about attribution a bit. Cause I think it's such an important piece when you do anything. It's like you have to have the data to be able to tell you what's working, what's not working, and just, I don't know. I've heard so much about data that are not good. But anyway, we really appreciate the time today.

[00:47:38] Thank you so much. Let's do this again. 

[00:47:41] Nima Gardideh: Yeah, 

[00:47:41] absolutely. Thank you for having me, guys. 

[00:47:43] Kyle Vamvouris: Beautiful. Talk soon. 

[00:47:44] Dan McDermott: Speak soon.

‍