Meeting Title: Omni Migration and Attribution Sync Date: 2026-03-02 Meeting participants: Greg Stoutenburg, Zoran Selinger


WEBVTT

1 00:01:05.330 00:01:06.229 Greg Stoutenburg: Is it wrong.

2 00:01:06.750 00:01:08.120 Zoran Selinger: Oh, you’re here after all!

3 00:01:08.120 00:01:13.380 Greg Stoutenburg: Yeah, I said I wouldn’t be here, and then I call in it early, so I was like, well, okay. I’ll just surprise everybody.

4 00:01:13.650 00:01:14.169 Greg Stoutenburg: That’s true.

5 00:01:14.170 00:01:24.950 Zoran Selinger: Sure. Yeah, I think, I don’t know, we might, we might just comment on, changes that you made on Friday to conversion tracking.

6 00:01:26.920 00:01:33.140 Zoran Selinger: Yeah, we, we had a… we kind of missed the weekly on Thursday as well.

7 00:01:33.140 00:01:33.900 Greg Stoutenburg: Yeah.

8 00:01:35.950 00:01:41.530 Greg Stoutenburg: How’s that going with the, with the, changes to attribution?

9 00:01:43.160 00:01:50.929 Zoran Selinger: I’ve just asked the guys if I can consider them officially unblocked. They should be.

10 00:01:50.930 00:01:54.620 Greg Stoutenburg: Okay. We are now sending basically all purchases to them.

11 00:01:54.870 00:01:55.310 Greg Stoutenburg: Okay.

12 00:01:55.310 00:02:05.969 Zoran Selinger: into the platforms, they’re still going to be judged as… on the same criteria, and they just have more… more data in the platforms, in TikTok and Meta.

13 00:02:06.070 00:02:06.880 Greg Stoutenburg: Okay.

14 00:02:06.990 00:02:07.700 Zoran Selinger: Yeah.

15 00:02:07.950 00:02:08.340 Zoran Selinger: Alright.

16 00:02:08.850 00:02:13.399 Zoran Selinger: I like that approach, and I pushed for that approach.

17 00:02:13.990 00:02:20.210 Zoran Selinger: As soon as they start And we’ve seen that with Catalyst, right?

18 00:02:20.210 00:02:20.850 Greg Stoutenburg: Hmm.

19 00:02:21.510 00:02:25.250 Zoran Selinger: They will never trust you. If you just don’t give them simple,

20 00:02:25.250 00:02:25.910 Greg Stoutenburg: Yeah.

21 00:02:26.230 00:02:34.220 Zoran Selinger: Pixel tracking, at least at the beginning, and then we can make… if we want, we can start demonstrating inadequacy of that.

22 00:02:34.460 00:02:42.510 Zoran Selinger: Yeah, right. But if you don’t start with that, it’s gonna be really hard. They’re gonna, they’re gonna complain all the way through.

23 00:02:42.780 00:02:51.430 Zoran Selinger: Especially, here, where they have the performance So…

24 00:02:51.990 00:03:05.469 Zoran Selinger: The lack of conversions that they see is very likely not a mistake, but the performance is that bad, because we also see… we are confirming the same from 3 different systems.

25 00:03:05.600 00:03:06.790 Greg Stoutenburg: Right. And…

26 00:03:06.790 00:03:12.430 Zoran Selinger: We also looked at, engagement on the website metrics, which are also terrible.

27 00:03:13.070 00:03:16.830 Zoran Selinger: So I think they need… they just need support.

28 00:03:17.240 00:03:25.059 Zoran Selinger: To me, it also seems weird that they do claim that they worked on similar projects, and that they’re done well.

29 00:03:25.380 00:03:29.510 Zoran Selinger: Why are we different?

30 00:03:29.630 00:03:38.280 Zoran Selinger: especially so much different, I don’t know, I’m really curious about that. Yeah. Is it us, or is it them? I’m not sure.

31 00:03:38.280 00:03:38.920 Greg Stoutenburg: Yeah.

32 00:03:39.240 00:03:40.260 Zoran Selinger: He’s…

33 00:03:40.310 00:03:41.349 Greg Stoutenburg: Which firm is it?

34 00:03:41.870 00:03:42.760 Zoran Selinger: Sorry?

35 00:03:42.760 00:03:46.090 Greg Stoutenburg: Which group is it that’s got the different numbers?

36 00:03:47.600 00:03:52.039 Zoran Selinger: No, so there’s, like, there’s Meta and TikTok agencies, right?

37 00:03:52.040 00:03:52.640 Greg Stoutenburg: Yeah, especially.

38 00:03:52.640 00:04:12.469 Zoran Selinger: at the agency, we do have their numbers from the beginning of February, and everything points to completely missing the mark on, I guess, the audience. They… they generate, they spend 25, 25, 27K, generated

39 00:04:13.130 00:04:15.200 Zoran Selinger: 50 new customers.

40 00:04:16.079 00:04:19.180 Zoran Selinger: 250 repeat purchases.

41 00:04:19.300 00:04:29.989 Zoran Selinger: even if all of those are new, it’s not amazing, right? Right. 50s… Like, what are we doing?

42 00:04:29.990 00:04:32.299 Greg Stoutenburg: Right, shut it off. Just turn it off.

43 00:04:32.300 00:04:38.289 Zoran Selinger: Yeah, and they’ve done that before, by the way. This is Eden’s second attempt.

44 00:04:38.750 00:04:44.480 Zoran Selinger: at, at Meta, in, like, what, year and a half?

45 00:04:44.920 00:04:45.620 Greg Stoutenburg: Yeah.

46 00:04:46.340 00:05:04.389 Zoran Selinger: So, I don’t know. We’ll see. Now, I just want to give them all the data they can ingest, and they can try to do their magic, whatever. Ryan will try to support them as well this week. He said, just looks like they are generating

47 00:05:04.390 00:05:13.810 Zoran Selinger: terrible audience, and he just wants to… okay, guys, how can I help? How can I give you the audiences that you can actually use, and…

48 00:05:13.810 00:05:14.190 Greg Stoutenburg: Yeah.

49 00:05:14.190 00:05:16.770 Zoran Selinger: contraction, because this is not working.

50 00:05:16.770 00:05:27.090 Greg Stoutenburg: I wonder what the deal is. You know, it’s funny, like, so just on my own phone, I went to Eden’s website, and then on Facebook, I started getting ads for all sorts of other peptides.

51 00:05:27.300 00:05:28.000 Zoran Selinger: Huh.

52 00:05:28.000 00:05:29.130 Greg Stoutenburg: but not eaten.

53 00:05:29.810 00:05:30.950 Greg Stoutenburg: I’m like.

54 00:05:30.950 00:05:31.430 Zoran Selinger: Interesting.

55 00:05:31.430 00:05:32.690 Greg Stoutenburg: Why?

56 00:05:32.690 00:05:33.220 Zoran Selinger: Yeah.

57 00:05:33.220 00:05:37.650 Greg Stoutenburg: I’m like, I didn’t even know these other ones existed. And I went to Eden’s page, none of the other ones.

58 00:05:37.650 00:05:39.249 Zoran Selinger: And now you do know.

59 00:05:39.250 00:05:42.230 Greg Stoutenburg: And Eden doesn’t show up. Yeah, right. Yeah. It’s like…

60 00:05:43.080 00:05:44.420 Greg Stoutenburg: Apparently, it’s not that hard to escape the.

61 00:05:44.420 00:05:45.340 Zoran Selinger: I don’t know, it’s…

62 00:05:45.340 00:05:46.280 Greg Stoutenburg: Mmm.

63 00:05:46.480 00:05:50.580 Zoran Selinger: Robert said an interesting thing. He says, he said.

64 00:05:51.750 00:05:58.350 Zoran Selinger: Feels like they just copy and pasted from other similar clients that I have.

65 00:05:58.910 00:06:02.370 Zoran Selinger: And that’s not working for whatever reason.

66 00:06:02.820 00:06:04.180 Greg Stoutenburg: Yep, yep.

67 00:06:05.060 00:06:06.610 Zoran Selinger: But we’ll see, we’ll see.

68 00:06:06.610 00:06:07.300 Greg Stoutenburg: Yep.

69 00:06:07.300 00:06:08.220 Zoran Selinger: But it’s…

70 00:06:09.470 00:06:11.040 Greg Stoutenburg: Log, do you think we should give Mitesh?

71 00:06:11.800 00:06:16.520 Zoran Selinger: No, no, I mean, it’s been 5 minutes, I think we’re… We’re going out.

72 00:06:16.740 00:06:25.679 Zoran Selinger: close it for… for now, if they want to join. They’ll ping me, feel free, like, you… if you want to give…

73 00:06:27.520 00:06:29.420 Zoran Selinger: Give an update later.

74 00:06:29.740 00:06:34.519 Greg Stoutenburg: Sure. Yeah, sounds good, yeah. Cool. All right.

75 00:06:34.520 00:06:43.510 Zoran Selinger: just out of curiosity, before we go. So, Omni went well. I mean, sounds like you got everything done.

76 00:06:44.300 00:06:45.600 Greg Stoutenburg: It seems that way, like, really.

77 00:06:45.600 00:06:46.430 Zoran Selinger: project.

78 00:06:46.740 00:06:48.100 Greg Stoutenburg: Yeah, really big part.

79 00:06:48.100 00:06:48.659 Zoran Selinger: That’s incredible.

80 00:06:48.660 00:06:49.210 Greg Stoutenburg: quick.

81 00:06:49.210 00:06:50.320 Zoran Selinger: That… that’s…

82 00:06:50.720 00:06:51.440 Greg Stoutenburg: Yeah.

83 00:06:51.440 00:06:53.460 Zoran Selinger: That’s actually done.

84 00:06:53.670 00:07:05.020 Greg Stoutenburg: Yeah, it’s… everything has been QA’d. There’s some… there’s, like, a small handful of revisions that need to happen, and a little bit of dbt work for a couple of charts.

85 00:07:05.030 00:07:16.969 Greg Stoutenburg: But yeah, the last thing I need to do right now is just… I need to schedule the mobile snapshots, which is most of what it seems like ELT relies on anyway. They don’t even… they don’t even log in, they just take the email snapshot.

86 00:07:17.280 00:07:28.500 Greg Stoutenburg: Yeah, but between, me, and then especially Mustafa, and where needed, Demolade and Awash, where Mustafa needed some extra, back-end support.

87 00:07:28.600 00:07:33.469 Greg Stoutenburg: Yeah, I mean, it was, like, 90% of it was done in… in…

88 00:07:33.660 00:07:39.239 Greg Stoutenburg: 5 days, and then the hardest remaining 10% was another 5 days.

89 00:07:40.610 00:07:43.570 Zoran Selinger: So… Do you have a few minutes?

90 00:07:43.570 00:07:43.920 Greg Stoutenburg: Yeah.

91 00:07:43.920 00:07:44.739 Zoran Selinger: I, I want…

92 00:07:44.740 00:07:49.520 Greg Stoutenburg: Yeah, totally. I don’t have anything until 2. And for me, it’s before 1, so…

93 00:07:50.370 00:07:56.640 Zoran Selinger: So I want to figure out… so if I want to find… figure out…

94 00:07:56.990 00:07:59.990 Zoran Selinger: Just an ad hoc analysis of whatever.

95 00:08:00.130 00:08:01.110 Greg Stoutenburg: Yeah.

96 00:08:01.110 00:08:11.890 Zoran Selinger: instead of writing SQL into the BigQuery console, can I now use… can I go to Omni and try to use Omni for exploratory data analysis?

97 00:08:11.890 00:08:12.500 Greg Stoutenburg: Yes.

98 00:08:12.660 00:08:15.950 Zoran Selinger: Not necessarily, right, for… Nope.

99 00:08:16.270 00:08:20.549 Zoran Selinger: reporting and dashboarding. Can I actually explore data there?

100 00:08:20.550 00:08:21.190 Greg Stoutenburg: Yep.

101 00:08:21.850 00:08:22.540 Greg Stoutenburg: It’s pretty cool.

102 00:08:22.540 00:08:27.460 Zoran Selinger: Is that a normal use for Omni? It seems to me like it is.

103 00:08:27.740 00:08:28.060 Greg Stoutenburg: Yeah.

104 00:08:28.060 00:08:29.599 Zoran Selinger: We use only for that.

105 00:08:29.600 00:08:36.190 Greg Stoutenburg: Yeah, I think… I mean, that’s a good question. Yeah, maybe we should… maybe we should just be relying on Omni ourselves. That’s a good question.

106 00:08:36.500 00:08:52.969 Greg Stoutenburg: So, that is… I think that’s going to be the primary use for it, is, you know, exploratory is the right word. You… we did rely on the AI tool to also build some charts and dashboards, but its ability to do that is… is slow and somewhat unreliable.

107 00:08:52.970 00:08:53.290 Zoran Selinger: Okay.

108 00:08:53.290 00:09:04.700 Greg Stoutenburg: Part of the reason why we were able to stand it up so quickly was by relying on the AI tool. But that said, even… it was just way more difficult and time-consuming than you’d think it should be.

109 00:09:04.960 00:09:05.510 Zoran Selinger: Okay.

110 00:09:05.510 00:09:17.280 Greg Stoutenburg: AI tool. And that was part of the feedback that I fed back Tommy. I was like, we kinda, you know, like, here’s what we thought it could do, and here’s what it could do, and it was a lot of hours to close the gap between those.

111 00:09:17.470 00:09:19.369 Greg Stoutenburg: But yeah, I mean, as far as, like.

112 00:09:20.750 00:09:24.870 Greg Stoutenburg: If you wanted to go in and just explore a little bit,

113 00:09:25.250 00:09:32.789 Greg Stoutenburg: the AI is great for that. I wouldn’t… I wouldn’t rely on it overly much to build some, you know, like, a new executive dashboard, at least not without.

114 00:09:32.790 00:09:33.900 Zoran Selinger: Oh, no, Ruben.

115 00:09:33.900 00:09:34.399 Greg Stoutenburg: And, you know.

116 00:09:34.400 00:09:35.299 Zoran Selinger: Not at all.

117 00:09:35.300 00:09:36.190 Greg Stoutenburg: Yeah, you know.

118 00:09:36.190 00:09:43.699 Zoran Selinger: In most cases, I need… I need… I need it for QA purposes right now, in most cases, right? Yeah.

119 00:09:43.960 00:09:44.780 Greg Stoutenburg: Yep.

120 00:09:45.720 00:09:51.899 Zoran Selinger: So, in… in building the… the current dashboards,

121 00:09:52.290 00:09:59.629 Zoran Selinger: Do you think we have, like, a comprehensive list of keys between tables for merging?

122 00:09:59.870 00:10:06.129 Zoran Selinger: Is that… Or you think there’s probably not everything is linked as much as it could be?

123 00:10:06.650 00:10:21.950 Greg Stoutenburg: Well, so… I mean, I think I would want to… I’d want to defer to Mustaf on that, since he had to touch more of the stuff in the back, but the reason why we were able to do this… there’s basically two reasons why we were able to do this so quickly. One is because BigQuery

124 00:10:22.050 00:10:39.940 Greg Stoutenburg: is already in pretty good shape. So, you just… we just connected BigQuery, so the data sources are now in Omni, and then, and then, of course, the fact that we were relying on already existing dashboards that were in Tableau. So, it was mostly just a matter of, like, copy this over and make sure it still does the same thing.

125 00:10:40.430 00:10:42.620 Greg Stoutenburg: Yep.

126 00:10:43.120 00:10:44.150 Zoran Selinger: Okay, okay.

127 00:10:44.150 00:10:44.520 Greg Stoutenburg: Yup.

128 00:10:44.520 00:10:57.490 Zoran Selinger: Oh, cool, cool. I’ll… I’ll start playing in there. I’ll try to… the next time I have a, like, a QA data task, I actually want to jump in there and try to do it there.

129 00:10:57.490 00:11:00.189 Greg Stoutenburg: Yeah. Yeah, I… Give it a try. Yeah.

130 00:11:00.190 00:11:17.160 Zoran Selinger: Yeah, I’m comfortable with spreadsheets, so that’s one huge advantage for me in Omni, because I can do some… it might be easier for me to do calculated fields, if I need to do it manually, to just do it in a spreadsheet way.

131 00:11:17.160 00:11:17.510 Greg Stoutenburg: Yeah.

132 00:11:17.510 00:11:28.310 Zoran Selinger: phenomenal. Yeah. That feature was… I was mind-blown by seeing these applicants in all those different ways, which is… wow. Yeah.

133 00:11:28.520 00:11:32.819 Greg Stoutenburg: Everybody’s excited about AI, you’re like, I get to use a spreadsheet here.

134 00:11:32.820 00:11:48.380 Zoran Selinger: And it’s still, like, you’d be surprised, maybe not, maybe you know, so many huge companies still rely on spreadsheets, and they just love it to this, like, they just wanna stay there. Spreadsheets are still super legit.

135 00:11:48.450 00:11:53.220 Zoran Selinger: For everything, basically. Yeah, so I’ll,

136 00:11:54.220 00:11:58.380 Zoran Selinger: Yeah, I want to do an analysis that is very complicated, though.

137 00:11:58.380 00:11:59.110 Greg Stoutenburg: Okay.

138 00:12:00.060 00:12:04.890 Zoran Selinger: Yeah, I wanna see the paths of, like, channels before the conversion.

139 00:12:05.230 00:12:11.730 Zoran Selinger: Okay. I don’t know if… if you touched Google Analytics, In your career?

140 00:12:12.040 00:12:21.459 Greg Stoutenburg: I mean, yeah, I’ve looked at it, I’ve looked at it a little, like, basic reporting, how are things going this week, this month, but not really done that much work in it.

141 00:12:21.660 00:12:32.000 Zoran Selinger: There is a really cool… there was a really cool analysis in there that shows you, the most typical paths to conversion of a single user.

142 00:12:32.000 00:12:42.989 Zoran Selinger: So, if they… so they first visited via email, then they visited via Google Ads, then direct, then organic, and then finally converted.

143 00:12:42.990 00:12:43.620 Zoran Selinger: Right?

144 00:12:43.620 00:12:44.440 Greg Stoutenburg: Yeah.

145 00:12:45.300 00:12:47.869 Zoran Selinger: It shows you the most common

146 00:12:48.390 00:13:00.009 Zoran Selinger: paths to a conversion, and, like, gives you stats against it. It’s a super, super useful tool to understand how your channels interact together.

147 00:13:00.010 00:13:00.430 Greg Stoutenburg: Right.

148 00:13:00.430 00:13:17.229 Zoran Selinger: whether they are mostly starting the interaction, or ending the interaction before a conversion, or there are mostly channels in the middle, sometimes you get surprised. You don’t necessarily know exactly what’s happening, so sometimes you get surprised there, and you can

149 00:13:17.440 00:13:23.379 Zoran Selinger: learn a lot from it, and I’ve never seen that report. Maybe you have.

150 00:13:23.550 00:13:25.669 Greg Stoutenburg: In any other tool.

151 00:13:27.140 00:13:34.489 Zoran Selinger: This is an attribution tool, like, normally it’s attribution tool, and it’s essentially an industry standard now, and…

152 00:13:35.050 00:13:36.860 Zoran Selinger: I don’t have that kind of report.

153 00:13:38.090 00:13:38.590 Greg Stoutenburg: Aww.

154 00:13:38.590 00:13:46.809 Zoran Selinger: I don’t have a path to a conversion anywhere. Right. So I’d like to build it, and our edge? Yeah.

155 00:13:47.480 00:13:55.449 Zoran Selinger: October of a… Never is high-quality work.

156 00:13:55.580 00:14:03.100 Zoran Selinger: My brother-in-law, he’s going. So…

157 00:14:04.250 00:14:09.800 Zoran Selinger: it’s a… it’s a pretty complicated one to do. I know it’s… it is complicated.

158 00:14:10.000 00:14:10.330 Greg Stoutenburg: Yeah.

159 00:14:10.330 00:14:14.289 Zoran Selinger: Our edge data is perfect for it. We have exactly

160 00:14:15.060 00:14:18.520 Zoran Selinger: exactly what we need, in Edge.

161 00:14:18.780 00:14:22.410 Greg Stoutenburg: Yeah, hmm… I am…

162 00:14:22.890 00:14:31.159 Greg Stoutenburg: I mean, well, I guess two thoughts. One is I’m trying to see if Amplitude will do that for you, because I think it can now.

163 00:14:31.550 00:14:35.180 Greg Stoutenburg: Traffic, traffic, ad performance.

164 00:14:35.690 00:14:42.429 Greg Stoutenburg: Version, segment, breakdown by channel… that’s not exactly what you just said, though.

165 00:14:42.650 00:14:45.499 Greg Stoutenburg: Yeah, I don’t know, I’d have to look. I know that I’ve…

166 00:14:45.710 00:14:52.610 Greg Stoutenburg: I’ve done things like that for, like, product analytics, like, once someone’s in an app to look at conversion there, but that’s not exactly.

167 00:14:52.610 00:14:53.980 Zoran Selinger: Oh yeah, of course, of course.

168 00:14:53.980 00:14:55.429 Greg Stoutenburg: Yeah,

169 00:14:56.480 00:15:06.369 Greg Stoutenburg: Yeah. Separately, so how does the edge layer tracking work? I had a question a while ago, and I was like… I was like, I don’t… I don’t… I don’t know what edge attribution is.

170 00:15:06.990 00:15:15.340 Zoran Selinger: Yeah, yeah, so, essentially, so, the idea of the edge layer is that

171 00:15:16.040 00:15:19.210 Zoran Selinger: Have we talked about this before? I’m not sure how…

172 00:15:19.210 00:15:21.220 Greg Stoutenburg: I think I sent you a message about it one time, and I think.

173 00:15:21.220 00:15:23.629 Zoran Selinger: Maybe we sat by the talking.

174 00:15:24.630 00:15:28.720 Zoran Selinger: Essentially, we are not relying on your browser loading the

175 00:15:29.480 00:15:41.910 Zoran Selinger: loading the website in any kind of way, we are doing that before your browser. We are collecting data before your browser loads the data. Loads the web… the website.

176 00:15:42.060 00:15:52.699 Zoran Selinger: Because when you are loading the website, then there’s tracking prevention that are coming into place, and your browser can do a lot of that work.

177 00:15:52.970 00:15:58.310 Zoran Selinger: Just preventing… Preventing, no…

178 00:15:58.510 00:16:11.539 Zoran Selinger: any tracking event is being sent via a request to a server that gets blocked, and there’s all kinds of different ways. And with Edge, we are jumping in front of that.

179 00:16:13.130 00:16:13.530 Greg Stoutenburg: house.

180 00:16:13.530 00:16:15.620 Zoran Selinger: Start loading the website.

181 00:16:15.620 00:16:17.750 Greg Stoutenburg: to the very beginning. So I’ve clicked a button.

182 00:16:17.750 00:16:19.380 Zoran Selinger: You click the button.

183 00:16:19.380 00:16:19.710 Greg Stoutenburg: Right.

184 00:16:19.710 00:16:26.180 Zoran Selinger: You request it, and before your browser gets a response, any response, We are there.

185 00:16:26.290 00:16:29.959 Zoran Selinger: We are there. We see that you requested a page.

186 00:16:29.960 00:16:30.340 Greg Stoutenburg: select.

187 00:16:30.340 00:16:35.169 Zoran Selinger: we can read, you know, your attributes like IP and all that stuff.

188 00:16:35.170 00:16:35.660 Greg Stoutenburg: Yeah.

189 00:16:35.660 00:16:36.580 Zoran Selinger: we can…

190 00:16:37.390 00:16:47.100 Zoran Selinger: change the request, essentially setting cookies, so you don’t see, basically, any impact on the performance of the website.

191 00:16:47.100 00:16:47.630 Greg Stoutenburg: Yup.

192 00:16:48.290 00:16:59.669 Zoran Selinger: work normally, but we just change your request to include two cookies, for example, and they’re already in the request. And obviously, we collect

193 00:16:59.840 00:17:07.590 Zoran Selinger: like, UTMs, click IDs, parameters, we have all of that. So, you have absolutely no chance to.

194 00:17:07.599 00:17:08.059 Greg Stoutenburg: Right?

195 00:17:08.200 00:17:11.889 Zoran Selinger: Because it happens before your browser can do that stuff.

196 00:17:12.020 00:17:13.780 Zoran Selinger: Yeah.

197 00:17:13.910 00:17:15.590 Greg Stoutenburg: What makes that possible?

198 00:17:16.390 00:17:24.220 Zoran Selinger: Oh, yeah, so, what… essentially, it’s called… you know what CDN is, maybe? CDN…

199 00:17:24.220 00:17:25.200 Greg Stoutenburg: Yeah. Yep.

200 00:17:25.200 00:17:34.469 Zoran Selinger: Yeah, Content Delivery Network. So it’s essentially, your website is a bunch of files that are together, right? And there, you…

201 00:17:34.760 00:17:46.100 Zoran Selinger: put that website onto a New York server, right? But you also joined a CDN, a Content Delivery Network. Cloudflare is one of them.

202 00:17:46.100 00:17:47.089 Greg Stoutenburg: No. Okay.

203 00:17:48.030 00:17:54.349 Zoran Selinger: they have servers across the world, so when someone wants to load your website from Tokyo.

204 00:17:55.070 00:18:02.300 Zoran Selinger: what you… what… what they will do, they will copy your files in the Tokyo, server.

205 00:18:03.160 00:18:05.300 Zoran Selinger: So, Tokyo users.

206 00:18:05.350 00:18:08.110 Greg Stoutenburg: You’ll load your website from the Tokyo server.

207 00:18:08.360 00:18:08.890 Greg Stoutenburg: Got it.

208 00:18:08.890 00:18:10.879 Zoran Selinger: It’ll be, like, instantaneous.

209 00:18:11.020 00:18:11.759 Greg Stoutenburg: Got it.

210 00:18:12.010 00:18:15.489 Zoran Selinger: That’s where we are. We are in this CDN system.

211 00:18:15.490 00:18:16.120 Greg Stoutenburg: Oh, dang.

212 00:18:16.120 00:18:19.000 Zoran Selinger: are on the server.

213 00:18:19.000 00:18:19.590 Greg Stoutenburg: Right.

214 00:18:19.590 00:18:23.449 Zoran Selinger: On the request, before your… before you get a response.

215 00:18:23.450 00:18:24.030 Greg Stoutenburg: Okay.

216 00:18:24.030 00:18:29.229 Zoran Selinger: from the server, that’s where we are. This is… that’s why it’s called Edge.

217 00:18:29.490 00:18:30.630 Zoran Selinger: Got it.

218 00:18:30.630 00:18:37.689 Greg Stoutenburg: That’s why it’s called Edge. Okay, so… so do I have it right then? So, I’m… In Tokyo.

219 00:18:38.140 00:18:47.520 Greg Stoutenburg: But the… is the webpage hosted in, like, New York? And so I click a button, but I connect to the Tokyo server, it modifies my request before

220 00:18:47.630 00:18:48.540 Greg Stoutenburg: It loads…

221 00:18:48.540 00:18:50.100 Zoran Selinger: No, no, it stays in the.

222 00:18:50.100 00:18:51.560 Greg Stoutenburg: It’s all in the Tokyo server.

223 00:18:51.560 00:19:02.479 Zoran Selinger: It’s literally copied, files are copied across the servers, across the world. Obviously, the way it works is it’s not gonna copy your files to Moscow automatically.

224 00:19:02.480 00:19:03.750 Greg Stoutenburg: Right. But first.

225 00:19:03.750 00:19:11.260 Zoran Selinger: Russian person that loads your website will actually… their page load won’t be fast.

226 00:19:11.450 00:19:12.270 Greg Stoutenburg: Right.

227 00:19:12.270 00:19:27.390 Zoran Selinger: Upon that request, for the first time, those files will be copied, but then every other person from… from similar area, from that region, will get super fast, because your website is now available on that server.

228 00:19:27.630 00:19:28.510 Greg Stoutenburg: Right.

229 00:19:28.660 00:19:33.820 Greg Stoutenburg: And this is how we’re gonna end up with thousands and thousands of copies of the entire internet all across the world.

230 00:19:33.820 00:19:41.500 Zoran Selinger: Yeah, we do. Essentially, the way, and it’s not just website, but literally any file that loads fast.

231 00:19:42.070 00:19:48.329 Zoran Selinger: is served through the CDN. So, copies are across the world, and obviously, you have…

232 00:19:48.510 00:20:05.370 Zoran Selinger: and options that will expire that document and setup. It just expires after 10 minutes, and the local server from Moscow, for example, has to load a new… a fresh version and all of that.

233 00:20:05.370 00:20:05.700 Greg Stoutenburg: Yeah.

234 00:20:05.700 00:20:08.219 Zoran Selinger: Your cash and all of that stuff.

235 00:20:08.430 00:20:12.850 Zoran Selinger: In any case, that gives us an opportunity to jump in there.

236 00:20:12.980 00:20:13.300 Greg Stoutenburg: Yeah.

237 00:20:13.300 00:20:16.979 Zoran Selinger: jump in there and do some work. What’s really cool.

238 00:20:17.440 00:20:24.669 Zoran Selinger: is we can modify your request, so we add, identifier cookies. Right. Just for Edge, in particular.

239 00:20:24.670 00:20:24.990 Greg Stoutenburg: Cool.

240 00:20:24.990 00:20:29.170 Zoran Selinger: We can collect the cookies that you already have, so what we’re doing…

241 00:20:29.500 00:20:36.909 Zoran Selinger: is we’re collecting identifiers from Google Analytics, from Mixpanel, from all the systems that you have.

242 00:20:36.910 00:20:37.310 Greg Stoutenburg: Yeah.

243 00:20:37.310 00:20:52.290 Zoran Selinger: pulling those identifiers as well. So, when we import data from Google Analytics directly to BigQuery, and we have that identifier also on the edge table, obviously we can merge them and do analysis. So.

244 00:20:52.500 00:20:58.589 Zoran Selinger: it’s just so many possibilities, right? And saving to a,

245 00:20:58.710 00:21:03.699 Zoran Selinger: warehouse happens asynchronously. So, your website loads.

246 00:21:03.840 00:21:06.440 Zoran Selinger: And saving happens in the background.

247 00:21:07.490 00:21:13.610 Zoran Selinger: In parallel, so you… you see no… no impact, page load speeds.

248 00:21:14.350 00:21:16.169 Greg Stoutenburg: That’s impressive. That’s amazing.

249 00:21:16.290 00:21:23.200 Zoran Selinger: Yeah, it, it really is, and it’s more precise and, and all of that stuff, so… Yeah.

250 00:21:23.630 00:21:32.750 Zoran Selinger: And it’s a really good, it’s a really good opportunity if we… if this is our first service that we do for the client.

251 00:21:33.130 00:21:33.520 Greg Stoutenburg: Yeah.

252 00:21:33.520 00:21:45.189 Zoran Selinger: Even if you look at marketing research for 26th, the main… the main obstacle for AI adoption is bad data, literally the number one.

253 00:21:45.310 00:22:01.019 Zoran Selinger: the whole point of EDGE is that we get that 95 plus percent accuracy with Edge data. So we have super accurate, and it’s, right, for now, it’s not 95%, it’s 99%.

254 00:22:01.350 00:22:04.409 Zoran Selinger: Right. So we are… that’s amazing.

255 00:22:04.860 00:22:12.950 Greg Stoutenburg: I don’t know, what do you think? You know, honest, experienced opinion, how long do you think it’ll take before, like, the EU prevents this?

256 00:22:14.130 00:22:16.660 Zoran Selinger: Before EU prevents this.

257 00:22:17.450 00:22:26.370 Greg Stoutenburg: Because it seems like, essentially, at least as far as the tracking part of it goes, it seems like, essentially, a way to circumvent ad blockers.

258 00:22:26.370 00:22:33.920 Zoran Selinger: It is. We… I think, what they will… they might do as a first step

259 00:22:34.410 00:22:39.440 Zoran Selinger: Is we will also have to listen for consent signals.

260 00:22:39.440 00:22:39.970 Greg Stoutenburg: Yeah.

261 00:22:39.970 00:22:45.360 Zoran Selinger: So, if there is no consent signal cookie already in the request.

262 00:22:45.710 00:22:46.410 Greg Stoutenburg: Right.

263 00:22:46.710 00:22:57.609 Zoran Selinger: there might be, an ordinance to… we cannot save any data. But if there… if there is a cookie already, we can do that work in… in there.

264 00:22:57.610 00:22:58.080 Greg Stoutenburg: Yeah.

265 00:22:58.080 00:23:05.649 Zoran Selinger: So, some… some tracking… we will still avoid some tracking prevention, doing… doing it that way.

266 00:23:05.860 00:23:13.660 Zoran Selinger: But, yeah, we’ll see. We’ll see. It’s very new. There is only one other company that does something like this.

267 00:23:13.900 00:23:21.919 Zoran Selinger: There’s already one company that does something like this. So, yeah, we’re not on the radar as much.

268 00:23:21.920 00:23:27.090 Greg Stoutenburg: Yeah. Yeah, yeah, yeah. Okay. Cool. Alright. Very clever.

269 00:23:27.300 00:23:33.589 Zoran Selinger: Thanks for staying a little longer and explaining the stuff to me. Yeah.

270 00:23:33.590 00:23:38.600 Greg Stoutenburg: I’ll definitely try and go in and try to get that report.

271 00:23:38.700 00:23:39.799 Zoran Selinger: In there.

272 00:23:39.800 00:23:54.820 Greg Stoutenburg: Yeah, perfect. And then, as far as, as far as the, the walkthrough that Jasmine organized for tomorrow, if you have, like, if you have what you know are Mitesh’s KPIs, or any benchmarks that you’d want to include some

273 00:23:55.110 00:24:09.329 Greg Stoutenburg: notes on that, I think that’d be good. I think one… one thing I think will be important is, you know, I’m… I just want to make sure that our dashboards are as usable for the stakeholders as they can be, so that also means, like.

274 00:24:09.370 00:24:15.209 Greg Stoutenburg: not adding things they wouldn’t necessarily want, or would quickly go out of date, or things like that, so…

275 00:24:15.210 00:24:17.570 Zoran Selinger: I told this to Yasmin already.

276 00:24:17.820 00:24:18.160 Greg Stoutenburg: Okay.

277 00:24:19.040 00:24:20.030 Zoran Selinger: what’s…

278 00:24:20.700 00:24:33.130 Zoran Selinger: seems to me, from the conversations I had with Mitesh, he feels like he doesn’t understand why a certain metric, or how it’s calculated, or how he got to the number.

279 00:24:33.250 00:24:36.670 Zoran Selinger: He just rejects it. So this is very important.

280 00:24:36.820 00:24:45.519 Zoran Selinger: It’s not even if we… if we’re giving him things that he doesn’t care about. Even if the thing that he cares about is there.

281 00:24:45.650 00:24:46.679 Zoran Selinger: call it…

282 00:24:46.860 00:24:47.770 Greg Stoutenburg: Yeah.

283 00:24:47.770 00:24:49.160 Zoran Selinger: like, NCAC.

284 00:24:49.560 00:24:54.350 Zoran Selinger: Yeah. And he’s… he’s doubtful about the way it’s calculated.

285 00:24:55.060 00:24:58.599 Greg Stoutenburg: So does he just want to see the formula? Like, MCAT…

286 00:24:58.600 00:25:10.149 Zoran Selinger: No, not necessarily the formal, but we have to be able to explain, for example, we’re using Norbin. Norbin has few attribution models, right? Yeah. And…

287 00:25:10.410 00:25:15.570 Zoran Selinger: We’re switching through models, and he sees things that don’t make sense to him, right?

288 00:25:15.570 00:25:15.970 Greg Stoutenburg: Right?

289 00:25:15.970 00:25:21.510 Zoran Selinger: The numbers between two models are significantly different, right?

290 00:25:21.510 00:25:22.930 Greg Stoutenburg: Why is that?

291 00:25:22.930 00:25:26.949 Zoran Selinger: And he’s not gonna look at the number if he doesn’t understand why.

292 00:25:27.060 00:25:33.429 Zoran Selinger: And we literally… we had to go into the documentation, really understand the documentation, and then…

293 00:25:33.620 00:25:52.199 Zoran Selinger: tell them, explain, why is… why is there such a difference? So even in the tools that we use, a lot of it will be proprietary, right? So obviously, we don’t have the full algorithm of NordBeam. Right. We have a general explanation of the model.

294 00:25:52.200 00:25:52.890 Greg Stoutenburg: Right.

295 00:25:52.890 00:25:57.190 Zoran Selinger: And then, if he understands, okay, that is fine, that is fine.

296 00:25:57.620 00:26:04.499 Zoran Selinger: he takes the number. That’s fine. Okay. So if you’re doing any, any type of stuff that, that is not, like.

297 00:26:04.980 00:26:14.179 Zoran Selinger: there’s some model data, or whatever, we’ll… we need to be able to explain, otherwise he’s going to reject it. He really needs to understand.

298 00:26:14.450 00:26:15.540 Greg Stoutenburg: Yeah, yeah.

299 00:26:15.540 00:26:16.560 Zoran Selinger: Anxiety.

300 00:26:16.560 00:26:25.119 Greg Stoutenburg: Okay, and then I guess we’ll just need to think about what the best way to present that is, whether that’s, you know, a lot of annotations on dashboards, or if that’s.

301 00:26:25.120 00:26:26.600 Zoran Selinger: I think that’s always helpful.

302 00:26:26.600 00:26:27.250 Greg Stoutenburg: Other documentation.

303 00:26:27.250 00:26:28.590 Zoran Selinger: Always. Yeah.

304 00:26:28.750 00:26:32.420 Zoran Selinger: Do we have a tool, like, tooltips, like, if you hover over, then…

305 00:26:32.970 00:26:34.219 Zoran Selinger: Do we have that in Omni?

306 00:26:34.220 00:26:35.929 Greg Stoutenburg: Question. I haven’t looked at that. Let’s see.

307 00:26:35.930 00:26:40.920 Zoran Selinger: Yeah, that could be a really nice thing, if there is that option.

308 00:26:41.290 00:26:42.120 Greg Stoutenburg: Yeah.

309 00:26:42.680 00:26:43.970 Zoran Selinger: That’s always nice.

310 00:26:44.400 00:26:47.669 Zoran Selinger: That will definitely help, at least Mitesh.

311 00:26:47.870 00:26:51.530 Zoran Selinger: To use the dashboard, and to trust it, and…

312 00:26:51.530 00:26:52.400 Greg Stoutenburg: Yeah.

313 00:26:53.120 00:26:56.610 Greg Stoutenburg: Yeah, I think one of the things that’ll come out of this Omni migration is just…

314 00:26:56.800 00:27:01.850 Greg Stoutenburg: Getting a better understanding from users what they want to see and what makes it most usable for them.

315 00:27:03.000 00:27:05.050 Zoran Selinger: Yeah, I mean, listen,

316 00:27:05.240 00:27:16.060 Zoran Selinger: Mitesh, it’s actually pretty simple for him. He wants to, he has a really like, logical…

317 00:27:16.240 00:27:17.750 Zoran Selinger: top-down approach.

318 00:27:17.900 00:27:19.709 Greg Stoutenburg: He will go from…

319 00:27:19.880 00:27:20.870 Zoran Selinger: Okay.

320 00:27:22.830 00:27:29.589 Zoran Selinger: overall metric, our NCAC in the last 7 days was way, way too high.

321 00:27:29.640 00:27:30.510 Greg Stoutenburg: Yeah.

322 00:27:30.830 00:27:33.229 Zoran Selinger: Okay, that’s top level.

323 00:27:33.230 00:27:33.670 Greg Stoutenburg: Yep.

324 00:27:33.670 00:27:44.269 Zoran Selinger: Now, let’s dig into which channels are down. Yeah. Okay? So now, we know that paid search was significantly down.

325 00:27:44.500 00:28:07.239 Zoran Selinger: And it’s our biggest channel, so that’s probably what caused it. Okay, now we click into the paid search, and then it’s the granular metrics for each channel that he’s insisting so much about. Literally, clicks, impressions, click-through rates, impression share, whatever’s relevant for that particular, he won’t literally go down to the metric.

326 00:28:07.460 00:28:22.400 Zoran Selinger: Keep drilling down. He wants a very specific metric for a specific channel, like impression share, which is not something that exists on email, or affiliate, or whatever, right? It’s specific for… and he wants to go down that

327 00:28:22.960 00:28:30.710 Zoran Selinger: into that much detail. That’s what he wants to achieve. He wants to achieve that approach.

328 00:28:30.710 00:28:31.330 Greg Stoutenburg: Yeah, yeah, yeah.

329 00:28:31.330 00:28:32.020 Zoran Selinger: bombs.

330 00:28:32.280 00:28:33.890 Greg Stoutenburg: Yep, drill down, and then next level.

331 00:28:33.890 00:28:44.260 Zoran Selinger: It doesn’t have to be one click away. It can be literally 5 separate documents, and he’s fine with that. He’s fine with having 5 separate spreadsheets.

332 00:28:44.260 00:28:44.920 Greg Stoutenburg: Yeah.

333 00:28:44.920 00:28:45.370 Zoran Selinger: That will…

334 00:28:45.370 00:28:53.750 Greg Stoutenburg: Yeah, I was actually just gonna say that this kind of explains why we’ve seen him make some very detailed spreadsheet requests. Yes. That’s the way he thinks about it.

335 00:28:54.260 00:28:56.520 Zoran Selinger: That’s exactly what he wants to achieve.

336 00:28:56.680 00:28:57.430 Greg Stoutenburg: Yeah, okay.

337 00:28:57.430 00:29:09.900 Zoran Selinger: And if he trusts it, he will be in there every day, maybe micromanage, I don’t know. We’ll find out. Maybe, I don’t know. In earlier conversations with him.

338 00:29:10.110 00:29:18.750 Zoran Selinger: He told me he’s done… he’s done similar things before, and he… he’s able to produce results that way.

339 00:29:18.750 00:29:19.210 Greg Stoutenburg: Okay.

340 00:29:19.210 00:29:29.929 Zoran Selinger: It’s how he likes to work. He had success before with those kinds of systems, and to be able to drill down that much, and he really trusts that he will be able to deliver

341 00:29:30.190 00:29:37.210 Zoran Selinger: The road that is very ambitious for this year, is if he has that kind of visibility.

342 00:29:37.210 00:29:41.520 Greg Stoutenburg: If you can see it. Okay. Okay. Alright, that’s helpful.

343 00:29:41.980 00:29:42.540 Zoran Selinger: Yeah.

344 00:29:42.540 00:29:44.010 Greg Stoutenburg: Very helpful. Cool.

345 00:29:45.130 00:29:45.990 Zoran Selinger: Okay.

346 00:29:45.990 00:29:46.870 Greg Stoutenburg: Alright.

347 00:29:46.870 00:29:47.530 Zoran Selinger: Thank you for your time.

348 00:29:47.530 00:29:49.820 Greg Stoutenburg: Good luck. Yep, you got it, you too. See ya.