David Melamed 11:39
Um, it gets heated. And I basically eventually say, I’m trusting my instincts and continuing to spend. And I don’t need your guarantee.
Dr. Jeremy Weisz 11:52
Yeah, so you leave the safety net of, Okay, we’ll give you $15,000 test. And now you’re like, well, you’re seeing one thing they’re seeing another and that kind of goes back into the tracking? Is that why you think they’re saying, Okay, I think we need to shut this thing down.
David Melamed 12:13
So, I mean, there was a couple things I didn’t realize at the time. But I understand now. So there’s a sales cycle with any sort of lead generation. And let’s just assume for a second, someone searches on Google today and clicks on that. And maybe they work on that application and complete it today, even. But there’s probably some people that clicked on that today, and then fill out an application in a week from now. And so there’s also that lag. And then some of the applications were going through automated underwriting and submitted directly to the SBA, others had to go through manual underwriting, which could take three days or seven days. So at this stage, there wasn’t enough time to know that anything was going to turn into revenue. But it didn’t look like things were even getting to the stage of potentially becoming revenue. Um, at least not at the scale that made sense to them. But it didn’t add up to me. And so I personally was convinced that this was a tracking and attribution problem. And if I had the opportunity to audit the systems, so just give you a very simple example. They’re giving us an email address, and then ending up on another page. And that page asks them for an email address, again, if one of those two email addresses was a typo, yeah, they typed in for Gmail, GMIAL in one of them. And we’re just matching things with email addresses without going through like a ETL, or data normalization process first, where we clean it up and reconcile it, or even, you know, a data pen process where we can take an email address and find five other email addresses associated with that person or company and match it without that process, which is a very tedious process. So I felt like there’s a very decent chance that if this is a tracking problem, we could resolve it with a manual audit at some point. But without that tracking, but you know, it didn’t make sense to me that they were converting on the first page at the rate that they were converting and then just like nothing happening with anybody. It didn’t add up to me based on the type of search it was the type of behavior etc. But there’s, there’s another piece to this that comes from my experience in my career, which is very, very telling and interest interesting about search, which is my very first successes online we’re in car donations and in cash for cars. In the cash for car industry. We have pages converting at like 40% of visitors into leads. And that’s not just an email address, that’s a phone number that’s actually calling up. Because at the end of the day song, you go see Google and types in you know, you know, so I think cars or something well, junk car specifically because they think it’s garbage like and to call up a company and find out how much will you give me for my car, it’s really a very soft sell. Like you’re giving them free money, essentially, who’s not gonna pick up the phone, I make that phone call. So I work very early on that search engine specifically, we’re very, very good for converting those very easy conversion event type things like giving away free money. A couple of years later, I had a very big success with a homecare company in something called CDPAP, Consumer Directed Personal Care where which essentially, it’s a Medicaid waiver program where family members could get paid to care for their loved ones. So it’s the exact same scenario, again, someone’s going to Google and searching for get paid to care for loved ones, again, free money, they’re looking to get paid for something that they’re probably doing anyways. And that converted really well. And you know, did you know, we probably spent over a million dollars on those campaigns at one point. And so at this point, I’ve already had two experiences with search engines working really, really well generating leads for giving free money. And this was, you know, an SBA back loan, that was forgivable. I, myself, went to Google to search it and converted myself for my own behavior. I just instinctively knew that this is how people people were going to apply to multiple banks, they were going to apply a bunch of different ways. But they were going to go to the search engines, and they were going to apply and the first page conversion validated that for me. And so I had no doubt in my mind that these were the right people. So we’re bringing the right people at the right time when they’re interested. So some we’re something broken here. And so my instinct said it was tracking. But the the piece that I didn’t understand at this time that this company I worked with understood is that we didn’t even have enough time to know what the SBA was gonna approve or not, we didn’t realize at this point, that the first like that 1000s of applications, we’re gonna get gummed up within the SBA’s pipes, because of new anti fraud measures they put in place. And because there was technology involved, every time a small thing was changed in an application by the SBA, it broke a whole bunch of things that people have scrambled to fix it. So I was assuming that just because I found the right people and got the leads, that that would be enough for it to turn into revenue. And the company had the opposite assumption. They were like, We don’t know if anything’s gonna convert, this looks like a complete mess. And even if we bring in all the right people, maybe nothing’s going to get funded at the end of the day, and so they have very conservative perspective. And I had a very liberal perspective about what was working and what wasn’t that, you know, it was, who was right and who was wrong, I think, I think we were both wrong. And to be honest, I’m at that stage, but one of the things that drove me, and this is something that I think about constantly, and I think it’s often misunderstood. The person who’s searching on Google today is a different person than the person searching tomorrow. So let’s say what this company would have been fine with is after 10,000. So the full 50,000 saying, What’s wait a month, let’s wait three weeks, let’s give time to see what happens with SBA. And then we could spend money when it works. What was going through my head is, every competitor is probably going to say the same thing. We don’t know what’s going to work, we don’t know what’s gonna get funded. Let’s slow down, let’s not spend money, but search engines are an auction. And nowadays, where people often use like conversion bidding where they tell Google spend X amount per lead, and by Google go crazy with it, they could be spending a fortune on bids. I might my instincts told me that not only am I gonna miss the person today, that searching if I wait till tomorrow, but I’m going to be able to reach the person today for a fraction of the cost of reaching the person tomorrow. So I’m going to keep on spending to get these cheaper leads. And I’m going to be happy with a certain amount of, of, of, you know, I wasn’t confident that I wasn’t going to lose the money but I was I was confident that I was better served spending that money now that waiting till all the Competition got smart about what’s working and waited for the data to flow. And that’s really where our disagreement was, was, the prudent thing was to be conservative, and to wait. But I felt like that was gonna miss the boat. And sometimes you just have to take a risk and take a dive. And I had the benefit of two very large career successes. In a very similar pattern, I’m very good at recognizing patterns, this pattern of going to search engines to get free money. It just, I just knew it works. And I wasn’t willing to accept that it wasn’t working. What happened next, though, you know, could probably bankrupted me. So you know, 15k was all they risks. But I was probably another 30 $40,000. In what still without certainty,
Dr. Jeremy Weisz 20:58
You’re at a crossroads. You’re 15k. They’re like, stop, you’re like, no, all signals point to go. And so you go. So what happens next,
David Melamed 21:09
I continue spending money. Our relationship ends up very tenuous, because the nature of the deal, they’re not sure if they have to be guaranteeing above the 15,000 or not. It’s not like it was solidified, I officially had a certain amount of leeway with the decisions. That’s just what they were comfortable with. But they were getting very anxious about it and, and frustrated with me and upset. And I had to kind of reassure them that I’m the one taking the risk, not them here. But that I’m going to want to audit all their systems and all their data. And it was, it was quite a roller coaster.
Dr. Jeremy Weisz 21:54
So you spend it another how much before you see any signs of life.
David Melamed 22:02
So this is where it gets a little interesting, because we made a change. And that’s when things suddenly work. Bear in mind, at the $15,000 point, I had data of how much it costs per click, what keywords people are searching for what questions they were asking, I had a lot of information about where to point, what direction to go, et cetera. Um, and one thing I noticed, which is something I see consistently across other industries is a significant portion of the search volume was around various different banks. And this program. So people weren’t just searching generically. I mean, some people were searching generically for this business loan program, other people. So an example would be saw my search Google for business loans, and someone else might search for, you know, Wells Fargo.
Dr. Jeremy Weisz 23:00
Chase, like, you know, you’re like, Oh, what does Chase have? And, you know, let’s see Chase in business loans.
David Melamed 23:07
Right. And then there was also a lot of companies that were running all sorts of promotions around these services. So people are getting emails and hearing from their friends about it. So the demand was growing, the search has grown. But there was a different behavior and how people search, there are some people searching generically other people searching about their types of businesses. So business loans for escorts business loans for LLCs, business loans, for nonprofits, business loans for trucking companies, for restaurants. So like, there’s a lot of different ways people are searching. And there are different intents and I I started having this data probably before any competitors that because I was taking more risks. But I continued spending and continued hearing that it wasn’t working. I saw the leads coming through the email addresses, but eventually it just like, you know, something was broken, I couldn’t get access to the data. It’s obviously we’re dealing with banks, so they’re not legally can’t share their data. And so I wasn’t even sure certain I was going to be able to audit systems if I wanted to. And I was probably about $60,000 in and at this point, this is like money that I did not want to risk and I thought for sure it was profitable and making money. I started having small indications of a few $1,000 of revenue, maybe 10% of the money returned, but most people will pull the plug at the stage. And I basically just went to the company and I said I need a direct link to the bank. They were capturing email addresses on this landing page and then saying nothing was happening after that landing page. I basically said I need to send people directly to that landing page to that bank application page. I needed to remove that one step. And there was three reasons why I believe that was crucial one. This bank was a big bank with brand recognition, and I saw people searching for banks. And the second reason is, it’s one extra step in the process, that might actually be breaking people’s like the congruency of the moment, like they might. They might be–
Jeremy Weisz 25:32
You wanted the branding of the bank on your page, and then it sends them to the bank. And they’re like, what, is this the same thing?
David Melamed 25:39
Exactly. And then, you know, so that worried me too. And, oh, no, I just, they, they gave me a direct link,
Dr. Jeremy Weisz 25:53
Isn’t that a risk on your part, though, because they give you a direct link. Now, you don’t even have an email to prove that you got that.
David Melamed 25:59
So, yes, it’s a risk. But I, at this point, I was also able to add conversion tracking to the site, to the bank’s website, so I was gonna be able to track some conversion events, and I tracked the entire funnel. So there’s five stages in the application, I was able to track every stage. But more importantly, I was relying on direct tracking, I had to have that email address and that lead. But there are tons of people clicking on the website, not converting today that we’re coming back tomorrow, and then come back tomorrow could have been through a different link a different way. And maybe I wasn’t going to get any credit for it. But I basically got was a completely new instance of a link to the site. So not used by anybody else. So there was actually a Salesforce instance, that was impossible to come from any source. But what I was running, which essentially meant that if I sent somebody to the page, and then a week later, they came back to the page another way and build out or completed application, I still got credit for it.
Jeremy Weisz 27:05
And unless, unless they cleared their cookies on our computer or something like that.
David Melamed 27:13
No but that’s the thing. As long as they ended up on this URL, there was no one else that was using it anywhere. It was me.
Dr. Jeremy Weisz 27:21
Got it.
David Melamed 27:21
Their browser will recognize that URL, probably if they, I wasn’t relying on anything except for what actually made it and literally overnight attacks. And I started seeing millions of dollars of loan applications coming in every single day, on the same exact campaigns, same keywords, same strategies, just overnight panics. But I was in a huge hole, right, I’m 60,000 in the hole at this stage. And I don’t know if I’m ever going to be able to audit those systems. And so I have to start scaling my spend to dig myself out of that hole. And bit by bit, you know, maybe two, three weeks later, I was at the point where, you know, it was doing about $20,000 A Day in, in profits. And I even had a couple of days that broke $100,000 in profit. And it just like, I was able to optimize at this point, I was able to use automated bidding, I had a tremendous amount of data. But most importantly, I was now trafficking, a page branded by the bank that everybody recognized, right, the person that goes to Google and searches for Chase Bank business loans, that sees an ad for Chase Bank and lands on Chase Bank website, that person’s converting, and the person searching just for business loan, so sees Chase Bank is converting to much quicker on Chase Bank than on you know, Bob’s barbecue and bank outlet. So that that brand ability and that ability to not rely just on the last but for the conversion was the difference between profitability and not. I don’t think the extra step was the problem as much as the incongruency and the branding.
Dr. Jeremy Weisz 29:19
And by the way, we’re using Chase as example that is not the bank necessarily here. So–
David Melamed 29:27
Although if you work for Chase, and you do want a performance partner, on your marketing, feel free to reach out.
Dr. Jeremy Weisz 29:32
There you go. At what point did the partners you’re working with, turn around and in realize because you can’t keep I mean, even if the money’s coming in to that other company, you’re still spending the money on the campaign. So at some point, you need some kind of backing here. At what point do they say okay, David, yes. We’re giving you our blessing. Keep going.
David Melamed 29:59
Immediately, as soon as they saw the results. There was no bad faith happening here. At any point of time. Everybody was on the same page, it was just a matter of risk versus understanding the risk. And the earlier risks that were taken, maybe the things I ran the first month, were actually never worked, that maybe those people never filled in applications and auditing, the systems would show I wasted all that money. But had I given up earlier and not kept on trying things and saying, okay, you know, I see the search behavior, how do I create a better match experience for the user? If I didn’t push for that I would have never found the profitability, I would have just given up way too quickly. And I wouldn’t have been a massive seven figure payday.
Dr. Jeremy Weisz 30:59
So, David, first of all, one last question. I want you just to talk about some of the overarching big lessons from this, but I want to point people towards davidmelamed.com that is davidmelamed.com, and then they can email you if you have questions at David at tenfoldtraffic.com. So is a recap, what were some of the big overarching lessons.
David Melamed 31:25
The number one lesson is there’s lots of moving parts in marketing. And just because something doesn’t work doesn’t mean you should pull the plug on it, it means that you should deconstruct them put it back together, maybe there’s just a piece of the puzzle that plugging that changes something from losing money to making money. And making money at scale. I mean, we’re talking about at one point in time, we’re doing bringing in over 1000 applications a day, like finding an offer that can convert to the tune of 1000 conversions a day is like, I mean, those are the gold mines. Those are the holy grails that you’re chasing. So it has to have the potential Don’t Be Dumb, and don’t waste money stupidly, but also recognize that the tracking and the data is not the territory. It’s just a map. And that like the person searching today is, it’s a different person searching tomorrow. So you waiting two weeks, three weeks, four weeks, for more information to come in, to know what’s going to work. But your competitors are also waiting for that same information, you’re giving up a competitive edge that you’re missing out on which maybe it’s worthwhile to do that, maybe it’s not. But you have to evaluate it in that context. And also, don’t just assume, like, you really have to think about it from the user perspective. At the end of the day, the searches person searching for that business loan on, you know, whatever is going to make them convert the best as what was important. And also constantly learn and listen, like I took risks that I would not have taken have I had all the information. But I didn’t have all the information. And, and it worked out and it could have gone the other way. So it could have been, you know, God block whatever you want to karma? Maybe Maybe I’m a good person, I don’t know. But it definitely is, I mean, definitely take in other perspectives constantly learn and, and always put the user first and think about like, you know, their experience, ultimately, because that’s what that’s what actually reflects reality the most.
Dr. Jeremy Weisz 33:36
I said it was the last question I actually have I’ve a second. That was my second last question. My last question, actually is, were there any surprises of traffic sources? Because you said in the very beginning, you were testing a bunch of traffic sources? Were there any surprises there?
David Melamed 33:51
Yeah, Microsoft ads ended up actually being the most profitable channel by far outperformed Google, and a ton of volume. And a lot of people especially in the US under underestimate, you know, Microsoft ads, which formerly called being ads. They don’t realize that the search bar and every Windows device is built in to Cortana, which searches Microsoft, the Microsoft powers, other search engines and other sites that you might come across. And, I mean, the Microsoft Edge browser is in a bunch of places.
Dr. Jeremy Weisz 34:33
I am shocked at that.
David Melamed 34:35
It’s something like 25 30% of US market share and it’s a higher income demographic. What people don’t know about Microsoft that’s really interesting is because Microsoft owns LinkedIn. The LinkedIn targeting data is built into Microsoft search. So you could actually say, like, you know, let’s say you want to target restaurant owners, who also are looking for business loans on on search engines. You can actually target that using the data in your Microsoft ads account. So it has additional capabilities that Google doesn’t have. It’s also less competitive, you know, good, the competitors gonna get there as quickly. And there’s one other thing that is super useful. And this is my secret weapon. This is actually the real secret to the success I have with these campaigns. I’m not gonna reveal everything but Google, opposite. Google hides about 30% of the searches you show up for in your reporting. You don’t even know what’s on search, they show that for Microsoft not only reports on everything that you get clicks on the report on everything that gets searched that your ad showed up for so quick was impressions. So when Google would show me 3000, search terms, Microsoft was showing me 25,000 and mining that uncovered all sorts of keyword opportunities and search behavior that I would never imagine. And what I actually discovered is that there’s full URLs for like big applications that people were copying and pasting into the search engines. And when you do that, in Chrome, it opens the URL when you do that on Microsoft, that opens a search. So I was actually able to target people actually trying to access like an application for these loans that other banks with, you know, linked our application. I would never discover that without that data.
Dr. Jeremy Weisz 36:28
David, first want to thank you, everyone. Thanks for having me. Go to davidmelamed.com. Email him David at tenfoldtraffic.com.
David Melamed 36:37
Thanks. Thanks so much. Dr. Jeremy Weisz. This is awesome.
Outro 36:43
Thanks for listening to the Fixing Incentives podcast. We’ll see you again next time and be sure to click subscribe to get future episodes.
Pages: 1 2
About David Melamed
Click Images To Verify
