Meeting Title: Attribution and Edge Layer discussion Date: 2025-09-12 Meeting participants: cutter, Zoran Selinger, Henry Zhao, Stuart i phone
WEBVTT
1 00:02:38.290 ⇒ 00:02:39.270 Zoran Selinger: Hello?
2 00:02:41.230 ⇒ 00:02:42.300 Henry Zhao: Good morning, guys.
3 00:02:46.790 ⇒ 00:02:47.980 cutter: What’s up, squad?
4 00:03:00.620 ⇒ 00:03:03.129 Henry Zhao: Just wait a few minutes, for Stuart to come on.
5 00:03:03.630 ⇒ 00:03:04.850 cutter: Alright, let me text him.
6 00:03:33.120 ⇒ 00:03:35.100 cutter: I guess we can get going, Henry.
7 00:03:37.450 ⇒ 00:03:47.639 Henry Zhao: Yeah, I can start with my part, Cutter. Just wanted to give you an update on, you know, attribution, as well as the reporting that I’m working on for Devin and Judd.
8 00:03:48.300 ⇒ 00:04:03.730 Henry Zhao: So… so basically, you asked on the Wednesday meeting, like, who can be the attribution expert, and I’d like to be one of the people that you can count on for attribution. I just think right now, attribution is a little messy with kind of all the changes we’ve made in Northbeam.
9 00:04:03.940 ⇒ 00:04:10.149 Henry Zhao: As well as, you know, right now we’re sending attribution data to our data warehouse also, which is BigQuery.
10 00:04:10.470 ⇒ 00:04:18.209 Henry Zhao: And my problem is that these numbers don’t match, so I’m really digging deep into the data to understand why these numbers don’t match.
11 00:04:18.420 ⇒ 00:04:32.950 Henry Zhao: Do we have a true last touch, or do we have a true first touch? And I think we’re bringing on Zoran also to help us implement some of the technologies behind the attribution, so that the data that’s going into segment is the best data that we can possibly have.
12 00:04:33.350 ⇒ 00:04:38.280 Henry Zhao: So with that, and when I have some time to get a better understanding of our attribution data.
13 00:04:38.400 ⇒ 00:04:41.219 Henry Zhao: I think we’ll see things run a lot more smoothly.
14 00:04:41.540 ⇒ 00:04:46.709 Henry Zhao: So right now I’m getting some people on the team to just help me look at the numbers and figure out why they’re so different.
15 00:04:46.990 ⇒ 00:04:54.780 Henry Zhao: So for example, the numbers for JUD, right, we’re looking at customer I.O. campaigns that are either Twilio or email.
16 00:04:55.470 ⇒ 00:05:07.279 Henry Zhao: And I’ve pulled the numbers from 4 different sources, and they’re all very, very different. So that’s the issue. I’m trying to resolve that and figure out why they’re so different, and once we get a better understanding of that, the dashboard should be done.
17 00:05:07.600 ⇒ 00:05:10.689 Henry Zhao: And moving forward, we’ll have a better understanding of attribution.
18 00:05:10.970 ⇒ 00:05:18.059 Henry Zhao: But the purpose of today’s call is I wanted to introduce Zoran to you guys, who’s going to help us improve our attribution tracking.
19 00:05:18.400 ⇒ 00:05:23.700 Henry Zhao: And he wants to also get a better understanding of what you guys are looking for, and get to know you guys.
20 00:05:24.370 ⇒ 00:05:30.630 cutter: Alright, so let me, let me tell you, kind of, Deron, what you’re walking into.
21 00:05:31.480 ⇒ 00:05:41.369 cutter: Attributions for us have always been a battle. It’s been a battle before Brainforge came on, it’s been a battle after Brainforge came on.
22 00:05:41.530 ⇒ 00:05:45.389 cutter: it got to the point in July where
23 00:05:45.790 ⇒ 00:05:54.749 cutter: We have a totally useless channel, which is meta, because we literally have no idea how to track or
24 00:05:54.870 ⇒ 00:05:57.299 cutter: attribute sales to Meta, and…
25 00:05:57.740 ⇒ 00:06:07.640 cutter: Once you get in here and you understand the complexities of the data sources, that’s one of our biggest issues, is our data source is fast, and they do not do things cleanly.
26 00:06:08.190 ⇒ 00:06:17.750 cutter: So… One of the… one of the most recent things that we discovered is, beyond just meta, We are…
27 00:06:19.290 ⇒ 00:06:21.459 cutter: I’m gonna send you the screenshot of the chat.
28 00:06:22.730 ⇒ 00:06:23.450 Zoran Selinger: Okay.
29 00:06:24.760 ⇒ 00:06:31.170 cutter: We are… For the last 3 months, we’ve been overpaying affiliates
30 00:06:32.550 ⇒ 00:06:35.739 cutter: you know, between 200 and 300K.
31 00:06:35.960 ⇒ 00:06:41.459 cutter: A month overpayment for misattributed sales conversions.
32 00:06:42.000 ⇒ 00:06:44.319 cutter: These people should be paid on…
33 00:06:44.550 ⇒ 00:06:49.000 cutter: I put it in the chat with Amber in the analytics channel, which is, like.
34 00:06:49.700 ⇒ 00:06:58.600 cutter: the last place that I put it in there, but these… these people should only be paid where last click UTM equals affiliate name.
35 00:06:58.720 ⇒ 00:07:06.339 cutter: or where first click UTM plus 14 days occur. After that, they should not be paid.
36 00:07:06.500 ⇒ 00:07:09.119 Zoran Selinger: But in that screenshot I sent over to you.
37 00:07:09.180 ⇒ 00:07:16.190 cutter: We were paying for reorders, we were paying for cross-sells, we were paying conversions that never occurred.
38 00:07:16.540 ⇒ 00:07:29.839 cutter: So, it… it has been an expensive, expensive learning journey in here. That’s… that… the dollar amount that’s on your screen in that error log right there is half a million dollars.
39 00:07:31.130 ⇒ 00:07:31.910 Zoran Selinger: Yeah.
40 00:07:32.110 ⇒ 00:07:34.789 Zoran Selinger: a big, big problem, I understand.
41 00:07:35.100 ⇒ 00:07:45.219 Zoran Selinger: So in your view, is this a problem of… of, figuring out the attribution itself, or…
42 00:07:45.330 ⇒ 00:07:49.980 Zoran Selinger: We are collecting data incorrectly, or not collecting at all.
43 00:07:49.980 ⇒ 00:07:56.599 cutter: I think it’s… we are not collecting data at all, or we’re collecting it incorrectly.
44 00:07:56.790 ⇒ 00:07:57.380 cutter: And…
45 00:07:57.870 ⇒ 00:08:04.910 cutter: We have Stuart in here. Stuart, just wait till I’m done, and then talk in short sentences so we can stay on topic, but…
46 00:08:05.210 ⇒ 00:08:07.489 cutter: I think what’s happening is, like.
47 00:08:07.630 ⇒ 00:08:14.999 cutter: that screenshot I sent you, we know how much we overpaid because we have the right data in there.
48 00:08:15.450 ⇒ 00:08:24.370 cutter: But it’s not getting back to the channels correctly. Like, that report and that screenshot that I sent you would not exist if we didn’t have the data.
49 00:08:24.370 ⇒ 00:08:26.739 Zoran Selinger: So, we have the data at some…
50 00:08:27.100 ⇒ 00:08:37.910 cutter: level, at some layer. It’s just not getting back to the people. Like, we should have not had paid for that stuff. We have the report, Robert Dang ran it, and we can see what we missed.
51 00:08:38.140 ⇒ 00:08:49.199 cutter: But it’s not getting out correctly, it’s not set up in the dashboards correctly, it’s not going to the affiliates for payouts correctly, so we… we lose half a million dollars because we missed a step.
52 00:08:49.390 ⇒ 00:08:50.289 cutter: You know what I mean?
53 00:08:50.900 ⇒ 00:08:58.810 Zoran Selinger: Right, right. So, you’re saying it’s not getting back to the systems directly. We have, you’re meaning
54 00:08:59.140 ⇒ 00:09:02.690 Zoran Selinger: Reporting, but also ad platforms.
55 00:09:03.070 ⇒ 00:09:08.229 Stuart i phone: Yep, and yes, abso-fucking-loopy, absolutely. Some of them, yes, some of them. Yes, some of them.
56 00:09:09.190 ⇒ 00:09:09.570 cutter: Stuart.
57 00:09:09.570 ⇒ 00:09:10.220 Zoran Selinger: Okay.
58 00:09:10.450 ⇒ 00:09:14.819 cutter: Stuart, you’re gonna go at the end, so that I can… I need this thing fixed now, because…
59 00:09:14.820 ⇒ 00:09:15.610 Stuart i phone: Yeah.
60 00:09:15.610 ⇒ 00:09:18.730 cutter: This is what is messing up our NCAT calculations.
61 00:09:18.730 ⇒ 00:09:20.409 Stuart i phone: Right, no, I was just gonna say, just sometimes.
62 00:09:20.410 ⇒ 00:09:24.699 cutter: Cool, man, cool. So, Doren, so the thing that we’re looking at
63 00:09:25.070 ⇒ 00:09:31.500 cutter: like, ASAP to figure out is the new system that we’re using for affiliates is called Catalyst.
64 00:09:31.690 ⇒ 00:09:39.429 cutter: And I don’t want to keep overpaying for affiliate conversions in there. So that is the one that is supposed to go live today.
65 00:09:39.780 ⇒ 00:09:42.909 cutter: And with you on the case, hopefully.
66 00:09:43.550 ⇒ 00:09:49.989 cutter: We’ll not be doing what we see in here, where we’re literally half a million dollars overpaid for conversions
67 00:09:50.270 ⇒ 00:09:53.740 cutter: That are the incorrect conversion event, you know what I mean?
68 00:09:54.620 ⇒ 00:09:55.330 Zoran Selinger: Yep.
69 00:09:55.680 ⇒ 00:10:04.299 Zoran Selinger: So that… you… you would say that, Catalyst working correctly is the number one priority right now, is that correct?
70 00:10:04.300 ⇒ 00:10:12.110 cutter: It’s the number one priority because it’s brand new, and it’s like a new install that we can actually set up cleanly.
71 00:10:12.230 ⇒ 00:10:18.210 cutter: with no… with nothing outstanding, nobody doing any weird shit to it, it’s brand new, so I’m…
72 00:10:18.440 ⇒ 00:10:20.490 cutter: It’s top priority because
73 00:10:20.870 ⇒ 00:10:28.209 cutter: We are… we know that we’re overpaying right now to these people. Yesterday, we overpaid, the day before that, the day before that, the day before that.
74 00:10:31.330 ⇒ 00:10:40.459 Henry Zhao: Let me ask a question, because I’ll need to be involved in this as well. How does Catalyst get the data on how much we owe them? Do we provide that data, or…
75 00:10:40.460 ⇒ 00:10:56.359 cutter: Yes, we provide that data, and we’re providing the incorrect data. It’s not coming from segment, it’s not coming debugged from the data warehouse, it’s coming from, like, pixel fires and UTMs on the client side, which is what Stuart’s gonna talk about, probably, is server-side and edge layer.
76 00:10:56.560 ⇒ 00:11:01.859 cutter: But… Yeah, we’re sending them the data, and we know that we’re sending them incorrect data.
77 00:11:02.170 ⇒ 00:11:05.630 cutter: Or else Robert Tseng wouldn’t have been able to pull that report, you know what I mean?
78 00:11:06.480 ⇒ 00:11:08.179 Henry Zhao: Okay, so…
79 00:11:08.360 ⇒ 00:11:20.820 Henry Zhao: as soon as possible, we need to talk about how this data is going to Catalyst, so we can just make sure that even after what Zoran implements, that the correct data is going to Catalyst. So it’s not just implementing the server-side tracking, but it’s correctly attributing
80 00:11:20.970 ⇒ 00:11:22.289 Henry Zhao: The, the revenue.
81 00:11:22.740 ⇒ 00:11:29.760 cutter: Amen, bro. So, Kevin is gonna be your point of contact on that, but keep me in the loop, because…
82 00:11:30.200 ⇒ 00:11:35.389 cutter: This is our last… Peace, and then what’s gonna happen is you guys fix this shit.
83 00:11:35.860 ⇒ 00:11:43.999 cutter: And then we’ll actually be able to spend money and make money, because we won’t be double, triple paying for conversion, so our NCAP will be real.
84 00:11:44.160 ⇒ 00:11:51.050 cutter: And then we can move on to FB and see if we can make Facebook work, and see if we can make TikTok work, et cetera, et cetera.
85 00:11:51.860 ⇒ 00:11:53.080 Henry Zhao: Okay, perfect, got it.
86 00:11:53.360 ⇒ 00:11:54.920 cutter: Cool. Alright, Stuart.
87 00:11:55.100 ⇒ 00:12:05.040 cutter: It’s your turn. Stuart’s gonna talk around about, server-to-server and edge layer to try to help get the correct data into the system itself.
88 00:12:06.870 ⇒ 00:12:10.710 Stuart i phone: Yeah, so, yeah, Cutter’s, yeah, right about everything he said, then… then…
89 00:12:10.830 ⇒ 00:12:18.380 Stuart i phone: when we… when we were attempting, like, multi-touch attribution in Northbeam, we would see a lot of…
90 00:12:18.380 ⇒ 00:12:30.499 Stuart i phone: of not only missing data, but, like, first click equals… first click conversion equals brand search. So, we know that there’s massive chunks of data because they’re using client-side UTMs.
91 00:12:30.500 ⇒ 00:12:41.569 Stuart i phone: And you can tell us whether this can be set up in a way where we can capture at least, like, 75% of data. We know it’s never going to be 100%,
92 00:12:41.820 ⇒ 00:12:52.799 Stuart i phone: But, but because of, you know, privacy, all the privacy settings and browser changes that you’re… you’re all aware of, or that you’re aware of, Zorin, yeah.
93 00:12:52.900 ⇒ 00:13:06.130 Stuart i phone: we have… we have a lot of gaps in there, so we’re… we’re looking to either improve that multi-touch attribution tool, or get another one that will work. But it’s gotten to the point where
94 00:13:06.260 ⇒ 00:13:25.429 Stuart i phone: I’ve almost, at least for now, with all these challenges, given up on trying to use third-party multi-judge attribution and just using, like, an incrementality measurement tool that has been finicky, but it’s also uncovered some massive,
95 00:13:26.470 ⇒ 00:13:31.680 Stuart i phone: gaps, like, that’s how I was alerted to the issue with overpaying affiliates.
96 00:13:34.240 ⇒ 00:13:48.920 Stuart i phone: So, a big part of this project is just you telling us what is feasible, for… for how much data we can capture for… and what the best solution is for multi-touch attribution, third-party, unbiased.
97 00:13:50.390 ⇒ 00:13:54.590 Zoran Selinger: Yeah, of course,
98 00:13:57.010 ⇒ 00:14:08.600 Zoran Selinger: Listen, the best thing we can do, when it comes to all those, tracking prevention systems is to… to stay in a first-party context as much as possible.
99 00:14:08.840 ⇒ 00:14:18.050 Zoran Selinger: Obviously. So this will… this will kind of keep our cookies, live for as long as possible.
100 00:14:18.050 ⇒ 00:14:28.899 Zoran Selinger: We are also… we have, kind of, multiple systems in place, so it would be great if the stitching is happening well, right? Right. Between those systems.
101 00:14:28.900 ⇒ 00:14:38.460 Zoran Selinger: And we can achieve that as well. In terms of staying in the first-party context, listen, in terms of the tool itself, I believe
102 00:14:38.930 ⇒ 00:14:48.749 Zoran Selinger: there are multiple tools that can do that work. I don’t… I think an orbit is fine, it just needs to be implemented well, right?
103 00:14:48.760 ⇒ 00:14:56.129 Zoran Selinger: So, when it comes to tracking, yeah, server-side is the best thing we can do there.
104 00:14:56.130 ⇒ 00:15:08.849 Zoran Selinger: Is… so I’m still onboarding here. Do we have the server-side GTM? Because I see there’s reliance on Google Tag Manager here. Do we have… do we have server-side container?
105 00:15:10.050 ⇒ 00:15:18.270 Stuart i phone: For some of our channels, we have server-side tracking, but not integrated with our third-party, MTA.
106 00:15:19.740 ⇒ 00:15:23.510 Zoran Selinger: Yeah, so, I’ll have to look at,
107 00:15:23.870 ⇒ 00:15:35.130 Zoran Selinger: at the GTM, if possible, and kind of see, see, see what’s there, and what we can, what we can do, on the, on, on the server side.
108 00:15:35.600 ⇒ 00:15:43.730 Zoran Selinger: Ideally, we can have… we can kind of transfer everything to the server-side GTM,
109 00:15:43.850 ⇒ 00:15:46.809 Zoran Selinger: And, obviously, that needs to be…
110 00:15:47.130 ⇒ 00:15:55.439 Zoran Selinger: Correctly configure so that we have… that, kind of, we either host it in, on our own, network.
111 00:15:55.880 ⇒ 00:15:58.439 Zoran Selinger: Where… where the website is as well.
112 00:15:58.550 ⇒ 00:16:03.970 Zoran Selinger: Or we… or we proxy it through Cloudflare, DNS, right?
113 00:16:03.970 ⇒ 00:16:09.990 cutter: We have Cloudflare, and we have Cloudflare DNS records. The person that you want to talk to.
114 00:16:10.110 ⇒ 00:16:15.960 cutter: Henry, can you just open a chat with Q, B, and Ryan, and Doran?
115 00:16:16.290 ⇒ 00:16:20.599 cutter: Because Ryan has the GTM containers and all of that on our side.
116 00:16:21.520 ⇒ 00:16:22.940 Henry Zhao: Ryan, right? And us.
117 00:16:22.940 ⇒ 00:16:25.260 cutter: Yep. Yep. R-Y-O-N.
118 00:16:25.950 ⇒ 00:16:27.070 Henry Zhao: Yep, got it.
119 00:16:29.560 ⇒ 00:16:43.099 Zoran Selinger: Yeah, so, so that’s… that is one piece. So we can… we can kind of… we can switch to… to server-side tracking for basically everything, that is… that is there.
120 00:16:43.100 ⇒ 00:16:57.800 Zoran Selinger: So, that’s… that’s a really good thing. We are going to stay in a first-party context, our cookies will… will be live for… for a long time. We won’t have any… any restrictions there. We will have a… we need to have, kind of.
121 00:16:57.880 ⇒ 00:17:02.190 Zoran Selinger: a tracking URL for on our own domain.
122 00:17:02.410 ⇒ 00:17:14.950 Zoran Selinger: for that, that’s kind of the part of the process, but that’s… that’s a technicality, right? So I don’t want to go into detail… into detail there. And then, obviously.
123 00:17:15.450 ⇒ 00:17:18.369 Zoran Selinger: We do have…
124 00:17:18.750 ⇒ 00:17:26.570 Zoran Selinger: So that still requires that the web page is loading, and that we have tags firing, right?
125 00:17:27.089 ⇒ 00:17:30.790 Zoran Selinger: Those things can happen even…
126 00:17:30.790 ⇒ 00:17:40.630 cutter: We have a… we have in Cloudflare, the last CPO had… Had it so that… There is,
127 00:17:41.080 ⇒ 00:17:49.629 cutter: I think there’s a space between where the website loads and where Cloudflare loads, and in that space, it’s just pure HTT.
128 00:17:51.030 ⇒ 00:17:56.679 cutter: That’s what I was using to… that’s how I set up our conversion rate… I mean, our…
129 00:17:56.800 ⇒ 00:18:06.539 cutter: intake questionnaires, that’s how I set them up. I loaded them on that. Instead of loading them inside of Webflow, where that whole giant architecture has to load.
130 00:18:07.280 ⇒ 00:18:16.799 cutter: Ryan has that access into Cloudflare as well, so maybe there… you can just add a tracking domain, especially from, like, email or something like that, you know what I mean? I’ve seen it done
131 00:18:16.910 ⇒ 00:18:19.450 cutter: With the tracking domain, with email, with…
132 00:18:19.620 ⇒ 00:18:22.929 cutter: Community Link, with all that random shit, so…
133 00:18:23.400 ⇒ 00:18:32.459 cutter: Stuart and I are just gonna rely on you to tell us, like, this is the best way to stitch the data, but for me, I know 100% I’m triple paying
134 00:18:32.800 ⇒ 00:18:34.519 cutter: for affiliate conversion.
135 00:18:34.780 ⇒ 00:18:40.440 cutter: Half of the data is in there. I know when we were running Meta, we were attributing, like.
136 00:18:40.770 ⇒ 00:18:41.450 cutter: Excuse me.
137 00:18:43.220 ⇒ 00:18:45.010 cutter: For women.
138 00:18:45.610 ⇒ 00:18:50.979 cutter: That’s the kind of shit that’s happening here, and it’s, like, the data on our side, but also the data getting…
139 00:18:57.520 ⇒ 00:18:58.799 Zoran Selinger: You’re muted.
140 00:19:00.060 ⇒ 00:19:02.230 cutter: Yeah, that’s it, that’s all I gotta say, bro, come…
141 00:19:03.420 ⇒ 00:19:05.080 Stuart i phone: Okay, cool. You cut out for a minute.
142 00:19:05.280 ⇒ 00:19:20.800 Zoran Selinger: Yeah, yeah. It was hard to understand, but we can catch up on… we can catch up on that. So, your suggestion about the… doing a little bit of stuff on the edge, in the edge, that… that’s a good suggestion.
143 00:19:21.170 ⇒ 00:19:25.999 Zoran Selinger: So that’s absolutely doable. We can,
144 00:19:26.250 ⇒ 00:19:31.739 Zoran Selinger: We can absolutely grab data before we attempt to load trackers, using.
145 00:19:31.740 ⇒ 00:19:49.279 Stuart i phone: Because that’s because so much gets stripped by, like, by Meta strips the UTMs if we don’t have the UTMs, and then the more identifiers that we have, the better the identity resolution is, but we can’t get those if they get stripped so fast. But with Edge, we capture them instantaneously.
146 00:19:49.630 ⇒ 00:19:52.610 Zoran Selinger: Absolutely, so we can, we can, we can do that.
147 00:19:52.630 ⇒ 00:20:06.450 Zoran Selinger: we can… I was looking into… I’ve done a little bit of collateral workers before, but it’s… I mean, our use case is going to be fairly simple, so it’s not a big problem.
148 00:20:06.450 ⇒ 00:20:15.320 Zoran Selinger: we can grab UTMs or whatever other parameters we need. We save them into whatever destination we want. Obviously, we will…
149 00:20:15.320 ⇒ 00:20:19.599 Zoran Selinger: So, there will be no impact on, on, on performance.
150 00:20:19.600 ⇒ 00:20:29.120 Zoran Selinger: Because we can do that work asynchronously. So we can let the page load for the user, but we, in the backend.
151 00:20:30.190 ⇒ 00:20:43.710 Zoran Selinger: we can do, whatever, saving to BigQuery and, you know, using NordBeam API, whatever we need to do there. In terms of the stitching, this is actually a good opportunity for us to
152 00:20:43.980 ⇒ 00:20:52.200 Zoran Selinger: to control the identifiers. Because what we can do, what we can do here, we can create a new hash.
153 00:20:52.410 ⇒ 00:21:08.900 Zoran Selinger: let’s say this is a new user. We can create a hash that we will initialize Google Analytics, NordBeam, whatever, with it, because all of those trackers, you can create a client, user ID, device ID, you want.
154 00:21:10.550 ⇒ 00:21:17.039 Zoran Selinger: So we can actually control what the ID is, and maybe we can use the same ID on multiple… on multiple…
155 00:21:17.040 ⇒ 00:21:22.219 cutter: Yeah, dude! Exactly. That would be fucking cool if we could do…
156 00:21:22.220 ⇒ 00:21:33.179 Zoran Selinger: Yeah, I think so too. That would be cool, to create. So, basically, in the worker, we create the ID, and we add it to the… we add it to the URL, and just
157 00:21:33.340 ⇒ 00:21:36.199 Zoran Selinger: let the user go to that URL,
158 00:21:36.480 ⇒ 00:21:48.630 Zoran Selinger: It will now have a new identifier, and we can initialize the GA4 tracker, Norbin tracker, with that ID. And every system will have the same ID.
159 00:21:50.220 ⇒ 00:21:58.830 cutter: Dude, that would be so clean if we could do that, because then we could track what happens on email, what happens on SMS, what happens on blah blah blah.
160 00:22:00.270 ⇒ 00:22:15.389 Zoran Selinger: Yeah, yeah, so basically we don’t have to do a lot of work for stitching, right? We don’t have to, you know, cross-reference, I don’t know, IPs, because, for example, when we are on the edge there, in the edge.
161 00:22:15.520 ⇒ 00:22:23.169 Zoran Selinger: we really, out of all the identifiers, I mean, we did not initialize… we haven’t initialized anything yet.
162 00:22:24.020 ⇒ 00:22:30.700 Zoran Selinger: So we don’t have an ID. The only identifier, really, at that point that we have is the IP address.
163 00:22:31.450 ⇒ 00:22:32.260 Zoran Selinger: Right?
164 00:22:32.430 ⇒ 00:22:33.320 Zoran Selinger: So…
165 00:22:33.660 ⇒ 00:22:51.300 Zoran Selinger: it is good that at that point, that we create the identifier, and then when the page loads, we actually load pixels with that identifier. So, that’s a… that’s an opportunity that we can start using.
166 00:22:51.610 ⇒ 00:22:55.939 cutter: That’s awesome, because CTV only comes with an IP address.
167 00:22:58.340 ⇒ 00:22:59.810 Zoran Selinger: Yeah, yep.
168 00:23:00.480 ⇒ 00:23:01.350 Zoran Selinger: And.
169 00:23:01.350 ⇒ 00:23:01.970 Stuart i phone: would…
170 00:23:03.460 ⇒ 00:23:04.030 Zoran Selinger: Go ahead, history.
171 00:23:04.500 ⇒ 00:23:11.610 Stuart i phone: Would this, work? Have you done this before with Northbeam or with other, like, MTA tools?
172 00:23:12.660 ⇒ 00:23:20.830 Zoran Selinger: So, so, listen, NordBeam, I haven’t done anything with NordBeam, but this is, this is,
173 00:23:21.330 ⇒ 00:23:27.520 Zoran Selinger: Kind of the… the tracker, the pixels are pretty much similar for… for… for…
174 00:23:27.690 ⇒ 00:23:35.370 Zoran Selinger: many different tracking platforms, so I… I’m pretty sure we can init… if we can initialize IP in the pixel.
175 00:23:36.830 ⇒ 00:23:38.320 Zoran Selinger: That’s it, we can do that.
176 00:23:39.190 ⇒ 00:23:40.240 Zoran Selinger: If you’re gonna sit…
177 00:23:40.240 ⇒ 00:23:40.690 Stuart i phone: I don’t know.
178 00:23:40.690 ⇒ 00:23:43.020 Zoran Selinger: Send the identifiers ourselves.
179 00:23:44.130 ⇒ 00:23:45.449 Stuart i phone: Cindy, okay.
180 00:23:45.450 ⇒ 00:23:47.630 Zoran Selinger: Yeah, we’ll be able to do it.
181 00:23:47.880 ⇒ 00:23:58.120 Zoran Selinger: In any case, even if that’s not possible, we will still save an IP address with the UTMs, which will allow us to stitch.
182 00:23:58.860 ⇒ 00:24:01.010 Stuart i phone: Awesome. Awesome.
183 00:24:01.920 ⇒ 00:24:03.310 cutter: Coop! Bye!
184 00:24:05.280 ⇒ 00:24:15.479 cutter: Alright, Henry and Zoran, it’s great to meet you guys. Henry, just open that chat with you, me, Ryan, and Zoran, so we can get the GPM.
185 00:24:15.670 ⇒ 00:24:18.040 cutter: Not the easiest.
186 00:24:18.510 ⇒ 00:24:21.380 cutter: Tegron also has Cloudflare access.
187 00:24:22.570 ⇒ 00:24:23.770 Henry Zhao: Yeah. Yeah, yeah.
188 00:24:23.770 ⇒ 00:24:27.360 cutter: word in there, and then… He’s gone as well, because he’s got…
189 00:24:29.630 ⇒ 00:24:30.750 Henry Zhao: And who as well?
190 00:24:33.540 ⇒ 00:24:34.500 cutter: Keep grind.
191 00:24:40.730 ⇒ 00:24:42.130 Zoran Selinger: Bring on, okay?
192 00:24:45.190 ⇒ 00:24:47.829 cutter: Duran, pleasure to meet you, looking forward to.
193 00:24:47.830 ⇒ 00:24:48.409 Zoran Selinger: You too.
194 00:24:48.530 ⇒ 00:24:50.259 cutter: Couple weeks, man.
195 00:24:50.260 ⇒ 00:24:50.630 Stuart i phone: Yeah.
196 00:24:50.630 ⇒ 00:24:53.869 cutter: Stitch some data and save some hashing and fucking…
197 00:24:54.340 ⇒ 00:24:55.110 Zoran Selinger: Makeup!
198 00:24:55.270 ⇒ 00:24:57.380 cutter: Make us make money, bro, that’s it!
199 00:24:58.550 ⇒ 00:25:09.029 Zoran Selinger: Sure, sure, let’s do it. I’m also excited, I think it’s gonna be a really interesting one, and I think the solution also will be, like, cool as well.
200 00:25:09.030 ⇒ 00:25:11.849 cutter: Elegant as fuck, it’s gonna be elegant, dude.
201 00:25:12.080 ⇒ 00:25:13.890 Zoran Selinger: Oh, yeah Exactly.
202 00:25:13.890 ⇒ 00:25:15.070 Stuart i phone: We’re gonna make it.
203 00:25:16.500 ⇒ 00:25:18.210 cutter: Alright, thank you guys so much.
204 00:25:18.390 ⇒ 00:25:19.160 Zoran Selinger: Appreciate it.
205 00:25:19.370 ⇒ 00:25:20.030 Zoran Selinger: Thanks.
206 00:25:20.030 ⇒ 00:25:20.939 cutter: Thank you. Bye.
207 00:25:20.980 ⇒ 00:25:21.980 Zoran Selinger: Take care.