Meeting Title: Marketing Analytics Eden Data Sync Date: 2026-02-12 Meeting participants: Zoran Selinger, Mitesh Patel, Ryon, Mat Schwarz


WEBVTT

1 00:01:23.490 00:01:24.580 Zoran Selinger: I’mitesh.

2 00:01:29.410 00:01:30.880 Mitesh Patel: I don’t have,

3 00:01:31.010 00:01:37.899 Mitesh Patel: Wi-Fi right now, because I had to turn the power off, they’re working, so the Wi-Fi area is without power, so I’m on my phone.

4 00:01:38.230 00:01:39.020 Zoran Selinger: Okay.

5 00:01:39.020 00:01:39.740 Mitesh Patel: Okay, okay.

6 00:01:50.600 00:01:51.629 Mitesh Patel: How’s it going, guys?

7 00:01:51.630 00:01:57.020 Zoran Selinger: won’t be joining us today, he just focused on Omni, powering through that.

8 00:01:57.190 00:02:01.479 Zoran Selinger: He’ll give you any update that he has, during the day.

9 00:02:02.370 00:02:03.110 Mitesh Patel: Okay.

10 00:02:03.110 00:02:03.680 Zoran Selinger: Yeah.

11 00:02:08.000 00:02:14.340 Zoran Selinger: Okay, so let me see, guys… You do…

12 00:02:14.480 00:02:20.570 Zoran Selinger: Let me just go quickly through… through the slides, and then we… we can discuss,

13 00:02:20.900 00:02:25.040 Zoran Selinger: those critical things that we wanted. Ran, Matt, is that okay?

14 00:02:25.620 00:02:31.090 Zoran Selinger: Yep, I’ll be very quick. I’ll just… yeah, hmm.

15 00:02:31.340 00:02:32.920 Zoran Selinger: Share my screen…

16 00:02:38.120 00:02:39.090 Zoran Selinger: Okay.

17 00:02:41.150 00:02:52.200 Zoran Selinger: Alright, so… So this week, we activated, we activated Meta and TikTok. TikTok is done with Tesh.

18 00:02:52.360 00:03:03.079 Zoran Selinger: Yeah. Awesome. You’ll see that. We hold, we’ll have, we have those new, customer I.O. segments, in. We, improved

19 00:03:03.890 00:03:28.530 Zoran Selinger: trackability of catalysts on the edge worker as well, and obviously, we’re continuing to work on the KPI tracking. We do have a piece next week with Norbim, right? We have that meeting. But I was also have, if you look at gun charts from weeks before, we do have, like, our review scheduled anyway in February. That was initially, like, once a queue.

20 00:03:28.530 00:03:33.589 Zoran Selinger: We actually go and… Look at the… analyze the coverage itself.

21 00:03:33.760 00:03:37.140 Zoran Selinger: So that was on… on our deck zero way,

22 00:03:37.520 00:04:01.780 Zoran Selinger: Yeah, so those, those should be… those few things. So, you see, now, both for, so, both on Facebook and on TikTok, we have those purchase events, coming in. These 9 that you see in this screenshot, these are the real ones. The one for TikTok, obviously, campaigns hasn’t started, but this is just a test one that

23 00:04:01.780 00:04:04.480 Zoran Selinger: That I… that I sent, so…

24 00:04:04.520 00:04:09.070 Zoran Selinger: we are good there. I’m happy to see that.

25 00:04:09.070 00:04:13.429 Mitesh Patel: These are, hold on, go back to the meta numbers. Yeah, I was trying to zoom in.

26 00:04:14.040 00:04:15.120 Zoran Selinger: Sorry about that.

27 00:04:16.350 00:04:20.839 Mitesh Patel: So… I’m on the phone, that’s why.

28 00:04:21.250 00:04:33.139 Zoran Selinger: 9… Yeah, this is… performance is very low, Mitesh. We, we checked it so many times, it’s really,

29 00:04:33.680 00:04:37.910 Zoran Selinger: I sent you screenshots, like, comparison with other channels and all that.

30 00:04:38.150 00:04:41.150 Zoran Selinger: This looks like it is real performance.

31 00:04:41.550 00:04:43.770 Mitesh Patel: Okay, and this is Last Touch?

32 00:04:44.530 00:04:48.720 Zoran Selinger: No, this is first touch. So, I wanna…

33 00:04:48.890 00:05:03.949 Zoran Selinger: let… let me… let… so… and I mentioned this under, under kind of risks, for us, and that’s, like, using, so there are two types of attribution… attribution models. Like.

34 00:05:04.140 00:05:10.689 Zoran Selinger: I mean, for two different purposes. One is for reverse CTL. This is what report back to the platform.

35 00:05:10.840 00:05:18.380 Zoran Selinger: And one is for our decisions on budget allocation and all that, okay? So the second part, the budget allocation.

36 00:05:18.800 00:05:22.650 Zoran Selinger: We need to include… we need to have a multi-touch model. This is…

37 00:05:22.780 00:05:34.499 Zoran Selinger: something that Norbin, for example, solves. We don’t have to talk about that much, we can go into an attribution tool, change the attribution models, and make our decisions that way.

38 00:05:35.250 00:05:39.970 Zoran Selinger: The first issue is when we report back to the platforms.

39 00:05:40.250 00:05:43.590 Zoran Selinger: What? How do we report back?

40 00:05:43.770 00:05:48.259 Zoran Selinger: And this is where I think that the last touch is not appropriate.

41 00:05:48.880 00:05:54.429 Zoran Selinger: Because these are upper funnel, channels. We are focused on new customers.

42 00:05:54.430 00:05:57.760 Mitesh Patel: Hold on, there’s a couple of things. Yeah, I understand, Zora.

43 00:05:57.760 00:06:00.059 Zoran Selinger: I want to give them the right signal, Mitesh.

44 00:06:00.300 00:06:01.260 Mitesh Patel: Yeah, yeah, no, can…

45 00:06:01.260 00:06:03.490 Zoran Selinger: Signal is the first touch in this case.

46 00:06:03.490 00:06:07.029 Mitesh Patel: I get it. I get it. Can you hear me?

47 00:06:07.030 00:06:07.470 Zoran Selinger: Yeah, yeah, yeah.

48 00:06:07.470 00:06:08.010 Ryon: Yeah, we can hear you.

49 00:06:08.750 00:06:28.679 Mitesh Patel: Okay, I get it, Zoran, and I understand what you’re saying, and it’s a great point, right? My thinking about it is, you know, meta is gonna be both from a retargeting perspective, it’s the bottom of the funnel, but it’s also top and middle of the funnel, okay? What I don’t want to do…

50 00:06:29.090 00:06:37.340 Mitesh Patel: Is… here’s my concern, by feeding them top of the funnel or first touch attribution data.

51 00:06:39.430 00:06:52.749 Mitesh Patel: my concern is, then the platform is gonna think, hey, I am doing great, let me keep sending traffic and keep charging, right? My platform CPA looks fantastic.

52 00:06:53.070 00:07:12.010 Mitesh Patel: However, then, when we look at the CAC, our internal customer acquisition cost, based on last touch, it’s not going to line up, right? If it’s making decisions, targeting and budget and spend decisions on first touch.

53 00:07:12.010 00:07:25.609 Mitesh Patel: And if those numbers are wildly different than last touch, it will keep spending, thinking it’s performing well, but when we look at our internal CAC, we’re gonna be destroyed.

54 00:07:25.750 00:07:28.519 Mitesh Patel: This happened to us last summer.

55 00:07:28.920 00:07:30.250 Mitesh Patel: In July.

56 00:07:30.810 00:07:35.140 Mitesh Patel: Right? And I… we can’t, you know… I, I, I agree.

57 00:07:35.140 00:07:36.820 Zoran Selinger: Are they new customers only?

58 00:07:37.130 00:07:37.740 Zoran Selinger: In July.

59 00:07:37.740 00:07:40.850 Mitesh Patel: For new… yeah, for new customers, the cost, yeah.

60 00:07:41.130 00:07:49.410 Zoran Selinger: Because… no, so, I wanna, I wanna emphasize that. So the purchase that you see there is not any purchase, it’s just for new customers.

61 00:07:49.410 00:07:54.780 Mitesh Patel: No, no, I understand, I understand. But how does this look? What is… okay.

62 00:07:55.560 00:07:57.520 Mitesh Patel: First touch says 9.

63 00:07:57.700 00:08:03.249 Mitesh Patel: Right? What is the last touch number say? How many conversions, if we look at last touch?

64 00:08:04.830 00:08:12.199 Zoran Selinger: Yeah, I don’t have the number in front of me, but it’s gonna be… it’s gonna be low, very similar.

65 00:08:12.200 00:08:14.910 Mitesh Patel: So, here’s the thing.

66 00:08:15.270 00:08:22.989 Mitesh Patel: if… Let’s just say we’re at some point in the future, and first touch attribution says there’s 100 conversions.

67 00:08:23.370 00:08:24.470 Mitesh Patel: Okay?

68 00:08:24.600 00:08:30.189 Mitesh Patel: If last touch attribution says it’s 90 or 95 conversions, I’m okay.

69 00:08:30.780 00:08:37.209 Mitesh Patel: But, if first touch says 100, but last touch says, like, 10, or 20,

70 00:08:38.400 00:08:51.069 Mitesh Patel: then we are significantly off. That means… what I’m saying is that the spend decisions are being made on one metric, but our internal calculations are happening on another metric.

71 00:08:51.540 00:08:55.639 Mitesh Patel: If they’re both close, I don’t… it’s fine, doesn’t matter.

72 00:08:56.670 00:08:57.480 Mitesh Patel: Right?

73 00:08:57.740 00:08:58.920 Zoran Selinger: By that point.

74 00:08:59.100 00:09:10.019 Mitesh Patel: Right, so if the first touch is saying 100 conversions, and the CPA is, you know, really good, 200 CPA, right, on the platform, because it sees 100 conversions.

75 00:09:10.360 00:09:13.670 Mitesh Patel: we’re gonna keep wanting to spend and scale up Meta.

76 00:09:14.820 00:09:28.860 Mitesh Patel: However, if the last touch, which is what we measure from a budgeting perspective, an internal cat calculation, says, hey, I only got… I’m making up a wild number, I only got half of that, 50, right?

77 00:09:29.000 00:09:37.000 Mitesh Patel: Now, our… the CAC will be… Meta’s contribution towards CAC will be stupid, it’ll be really bad.

78 00:09:38.540 00:09:44.999 Mitesh Patel: And now we’re spending not based on our internal metrics.

79 00:09:46.240 00:09:48.699 Mitesh Patel: This is what happened last July.

80 00:09:50.180 00:09:54.010 Mitesh Patel: And we turned off Meta, we never went back to it.

81 00:09:57.570 00:10:06.899 Zoran Selinger: I think… so, my… my suspicion is that the first and last is going… are gonna be very similar. We can… we can look that up.

82 00:10:07.630 00:10:16.709 Zoran Selinger: My concern is that we are not going to credit enough to the platforms. Our goal is now aggressive expansion.

83 00:10:16.930 00:10:18.300 Zoran Selinger: aggressive growth.

84 00:10:18.570 00:10:21.040 Mitesh Patel: Yeah, what I… yeah, I understand session.

85 00:10:21.040 00:10:24.590 Zoran Selinger: Yeah, and John, the business decision we’ve made.

86 00:10:24.630 00:10:27.780 Mitesh Patel: Is… we’re going to…

87 00:10:28.490 00:10:43.510 Mitesh Patel: credit… attribute back to each platform what we use for internal metrics. I don’t want to grow on Meta if, you know, we’re lying to ourselves in terms of what Meta’s real contribution is to our CAC.

88 00:10:46.740 00:10:47.340 Zoran Selinger: Yeah.

89 00:10:47.950 00:10:53.790 Mitesh Patel: And like you said, if they’re both the same, or close, within 5%, I don’t care, right?

90 00:10:53.790 00:10:54.330 Zoran Selinger: Yeah.

91 00:10:54.330 00:11:03.999 Mitesh Patel: And I think this is a really, really… you’re bringing up a very, very important point, but the business rule that I’m saying is not movable, okay? It’s gonna be based on the CAC.

92 00:11:04.700 00:11:16.680 Mitesh Patel: Now, you and I can discuss… we, all of us, can discuss this with the meta agency that’s managing this. If they’re aligned, then on the last touch, then that’s what we go by.

93 00:11:20.460 00:11:25.759 Zoran Selinger: Yeah, general… I mean, obviously, I don’t have, I am absolutely…

94 00:11:25.860 00:11:35.299 Zoran Selinger: no problem with the decisions being made on any attribution model that you internally decide on. What I’m talking about here is

95 00:11:35.860 00:11:42.569 Zoran Selinger: I’m trying to give the best possible signal to the platform, so the algorithms can do their work.

96 00:11:42.570 00:11:42.920 Mitesh Patel: Ike.

97 00:11:42.920 00:11:45.409 Zoran Selinger: Thinking about what that signal is, right?

98 00:11:45.410 00:12:04.549 Mitesh Patel: I got it, Zoran, I appreciate that very much. I just don’t… and look, this is an internal versus Zoran. You are our part of our team, so this is us. This is our goals, okay? And what I don’t want… I understand and I appreciate giving the best signals to the platform.

99 00:12:04.550 00:12:14.220 Mitesh Patel: But we cannot give the wrong signals to the platform that causes us to overspend, then we’re gonna get killed on the internal cap and internal profitability.

100 00:12:14.220 00:12:19.339 Zoran Selinger: Yeah, it’s, it’s really, it’s really a, you know, a classic.

101 00:12:19.630 00:12:23.719 Zoran Selinger: Between… between efficiency and volume. Exactly.

102 00:12:23.720 00:12:35.560 Mitesh Patel: Exactly, and we don’t want to scale on a channel that we’re not profitable on. I want to scale, I want to blow up, but every channel has to be profitable on its own.

103 00:12:36.120 00:12:43.980 Zoran Selinger: Yeah, okay, let’s… so, what’s the definition that you want, exactly, then?

104 00:12:44.170 00:12:48.899 Zoran Selinger: And we can… we can put that in… into place next week.

105 00:12:49.180 00:12:59.880 Mitesh Patel: Yeah, I want it to be, you know, everything else about the new customers only lasts, whatever, 7 days, whatever we have, but I want it to be based on last touch.

106 00:13:02.370 00:13:03.250 Zoran Selinger: Okay.

107 00:13:03.250 00:13:06.110 Mitesh Patel: And John, I want us to monitor.

108 00:13:06.370 00:13:09.679 Mitesh Patel: What is the first touch versus last touch?

109 00:13:10.810 00:13:20.460 Mitesh Patel: And I want to review this with you on the phone, with Max and Stan, who are at Ghost Agency, who are managing this for us.

110 00:13:21.000 00:13:36.819 Zoran Selinger: Sure, I mean, we can leave the models that are… that are in there, we can leave them in place. We don’t… we’ll add new ones on top of the ones that we have, so we’ll have the numbers easily, easily there. So the models that… that are… that are currently in place.

111 00:13:36.980 00:13:41.279 Zoran Selinger: They can stay there, and we will have the comparison very clearly.

112 00:13:41.500 00:13:48.599 Mitesh Patel: Zarin, Matt and I… Matt, Ryan, and I need to discuss with all of you, sort of, the whole affiliate thing. How much more do you have? I want to make sure.

113 00:13:48.600 00:13:56.909 Zoran Selinger: No, it’s, it’s fine. I think we are talking about the most important points.

114 00:13:59.130 00:14:18.570 Zoran Selinger: the attribution for reverse ETL was one of the things that I wanted to discuss, and then the last… the last part that we touched on earlier today, I just added it to the slide, is the busk, but let’s… let’s talk about the affiliates first.

115 00:14:19.040 00:14:25.250 Mitesh Patel: Okay. This VASC thing, Ryan, have you helped escalate this?

116 00:14:26.600 00:14:37.550 Ryon: Which one, specifically? Are you talking about the, lack of support in terms of the product and item data, or the state router? Like, which one are we talking about here, Zoran?

117 00:14:37.550 00:14:41.559 Mitesh Patel: For this meeting, it’s this data, what he’s got on the chart here.

118 00:14:41.780 00:14:59.129 Ryon: This is the item one. Okay, this has not been escalated yet, but it needs to be. That’s one of the cult reasons I want to talk to you, is I’m going to start making a list of things right now that we’re not getting support from Bascon, and I need to chat with you and with others about this, because this is just one of a couple of things recently that are, like, a real big problem.

119 00:14:59.790 00:15:05.989 Ryon: This one is more recent, though. It’s not… Okay. Yeah, this one’s more recent. Alright, how many things are on this list?

120 00:15:05.990 00:15:07.639 Mitesh Patel: 3, 4, 5?

121 00:15:11.880 00:15:13.149 Ryon: At least 5.

122 00:15:13.420 00:15:14.470 Ryon: Okay.

123 00:15:14.470 00:15:19.539 Mitesh Patel: make that list today, let’s prioritize it with Brad, and Brad will escalate.

124 00:15:21.190 00:15:25.579 Zoran Selinger: Okay. Can you guys tell me, just for my understanding,

125 00:15:25.880 00:15:31.429 Zoran Selinger: the decisions to… the decision to go away from… from Basque, that’s final, that’s happening, right?

126 00:15:33.260 00:15:38.880 Mitesh Patel: That is final, that’s gonna happen, but it’s gonna take time to migrate completely off.

127 00:15:38.880 00:15:45.829 Zoran Selinger: Yeah, so this is… it’s worth doing all of this fixing in the meantime, because it’s going to be a while.

128 00:15:45.830 00:15:47.409 Mitesh Patel: Yeah, we gotta, we gotta do it, we can’t be hot.

129 00:15:47.410 00:15:49.699 Mat Schwarz: It’s broken right now, yeah.

130 00:15:49.700 00:15:52.639 Zoran Selinger: Yeah, I mean, he’s… okay, cool, okay, fine.

131 00:15:52.860 00:15:55.319 Mat Schwarz: We can run affiliates like this.

132 00:15:57.630 00:16:01.400 Mitesh Patel: Okay, let’s get to affiliates. I only have 10 more minutes, guys, let’s figure this out.

133 00:16:04.470 00:16:08.820 Mat Schwarz: So part of one… one of those things that Ryan was saying, Mitesh, is the nulls.

134 00:16:09.260 00:16:12.310 Mat Schwarz: We didn’t showing up as conversions and affiliates.

135 00:16:12.550 00:16:16.029 Mat Schwarz: There’s 300-something that we don’t know what to true with the payouts to.

136 00:16:16.300 00:16:35.790 Ryon: Yeah, so let me speak to a couple of things with regards to the nulls, and how we’re gonna fix that, because a lot of the fixes are in place, and we’ve seen a reduced number of nulls, but then other things have popped up. So, first things first. One of the nulls that we fixed was the bot issue. That… Azaran and I did everything we could on our side, and Calus had to come in and block those orders completely. So those are gone.

137 00:16:35.790 00:16:49.239 Ryon: The other one was that the server ping, the server API job, was pinging out to Catalyst too often. It was pinging on the existence of a cookie that was basically on every visit.

138 00:16:49.240 00:16:59.280 Ryon: So, that’s why you guys were getting conversions that had no offer ID, or publisher ID, or etc, no information there. Okay, so that’s… that’s the first set of null… that’s the second set of nulls that’s been resolved.

139 00:16:59.280 00:17:09.099 Ryon: The other now is gonna be a client-side solution that we’re gonna be deploying today, Zoran and myself. We’ll test it out today, probably deploy it tomorrow morning. And it basically is just gonna…

140 00:17:09.520 00:17:29.090 Ryon: further restrict when the conversion event fires to only catalyst conversions. Right now, it’s firing for a couple of other catalyst conversions, like maybe one or two that are non-catalyst conversions, and then it will fire only on catalyst conversions. The last thing is the BASC bug, where they’ve stopped reporting the item

141 00:17:29.190 00:17:31.690 Ryon: information in the data layer.

142 00:17:31.870 00:17:42.679 Ryon: Right, so that we’ve got to get support for. I think we can get that solved relatively quickly. It doesn’t seem like it’s that big of a bug or an issue for me, like, to solve, but I need to escalate that to others.

143 00:17:43.110 00:18:01.109 Mitesh Patel: Okay, who can you get on the call right now, Ryan? And I’m, you know, I’m… I’m trying to… I’m saying that it’s hyperbole, right? But who can you get on a call today? Can you, like, schedule yourself with Gal and get Nico on it, whoever the right people are, and just schedule it?

144 00:18:01.480 00:18:12.770 Ryon: I need to… so I spoke to Gaul this morning in a one-on-one about a couple of other things related to PLAs and cell theory. I will call a meeting with Nico, because Nico’s better for this kind of thing, and I’ll just let Gaul know how.

145 00:18:12.770 00:18:18.969 Mitesh Patel: Wherever you need, get on it today, don’t add it to a list to escalate, just you escalate it, get it fixed.

146 00:18:20.290 00:18:22.240 Ryon: Okay, I’ll talk to Nico today, then.

147 00:18:24.090 00:18:30.619 Mat Schwarz: So, I do want to talk about the timeframes that we’re dealing with, since we can’t… like, we’re gonna… we’re overdoing the payment.

148 00:18:31.620 00:18:32.800 Mat Schwarz: to catalyst.

149 00:18:33.340 00:18:36.930 Mat Schwarz: And it’s sort of like the second or third time that we’re there.

150 00:18:36.930 00:18:48.659 Ryon: We’ve identified the orders, Nick… sorry, Matt, we’ve identified the orders that we need to send through to them and say, hey, these are additional orders that we’re giving you guys credit for. That turns out to be…

151 00:18:48.660 00:18:51.959 Mitesh Patel: Ryan, Ryan, Ryan, Ryan, let me get back into…

152 00:18:53.390 00:18:58.729 Mitesh Patel: I hope we haven’t shared any information with Catalyst or anybody else right now, okay, our internal data.

153 00:18:58.850 00:19:04.149 Mitesh Patel: I hope it’s not on the sheet that… that was shared with them. Alright. How many…

154 00:19:04.150 00:19:06.490 Mat Schwarz: Right, Zora? You didn’t put it in there, right?

155 00:19:07.860 00:19:12.620 Zoran Selinger: I mean, I, I have no direct communication with,

156 00:19:12.880 00:19:16.549 Zoran Selinger: with… with Catalyst in any shape or form, I don’t know.

157 00:19:16.550 00:19:23.130 Mitesh Patel: No, no, Zoran, it’s not about… it’s not about communication directly with them. It is about.

158 00:19:24.310 00:19:28.760 Zoran Selinger: I shared it in a private message with Matt and Ryan.

159 00:19:28.990 00:19:30.060 Zoran Selinger: the file.

160 00:19:30.690 00:19:34.920 Mitesh Patel: Okay, is this meeting… is this meeting being recorded?

161 00:19:35.330 00:19:36.690 Zoran Selinger: It is, it is.

162 00:19:36.690 00:19:37.819 Mitesh Patel: Can you turn this off?

163 00:19:38.100 00:19:38.690 Zoran Selinger: Sure.

164 00:19:40.210 00:19:40.880 Mitesh Patel: Ryan, are you.