Meeting Title: Tagging-Tracking Sync Date: 2025-07-28 Meeting participants: Robert Tseng, Andrew O’Neil
WEBVTT
1 00:00:23.390 ⇒ 00:00:24.350 Andrew O’Neil: Hello!
2 00:00:29.750 ⇒ 00:00:38.449 Robert Tseng: Okay? Well, hopefully, that was helpful context. Just to see broadly, like what our team is like working on for the client.
3 00:00:39.556 ⇒ 00:00:47.450 Andrew O’Neil: Yeah, yeah, perfect. Yes. And I I did. I got the tag manager invite. So this is great.
4 00:00:48.300 ⇒ 00:00:57.370 Robert Tseng: Okay, I like to just hear kind of like your understanding of, like the situation. I know you met with Henry last week. He briefly caught me up on like what he showed you, but.
5 00:00:58.930 ⇒ 00:01:05.769 Robert Tseng: Yeah, and then we’ll I guess I’ll be kind of getting organized as you’re as as we’re talking.
6 00:01:06.420 ⇒ 00:01:33.699 Andrew O’Neil: Yeah. Yeah. So it seems like, you know, the priorities include you know, some like, look into Gtm, giving your you all more insight into what’s in the container. You know, auditing it for for initial tracking, perhaps. Cause I I believe that there’s a ticket that core. Yeah. Ticket number 4, 90 cleanup. Gtm, container.
7 00:01:34.118 ⇒ 00:01:47.080 Andrew O’Neil: That that one makes sense, and then the other one at least. Platform. Wise seems like a priority. Is some meta tracking around their conversions. Api.
8 00:01:47.270 ⇒ 00:01:49.690 Andrew O’Neil: So I think that might be.
9 00:01:49.820 ⇒ 00:01:56.503 Andrew O’Neil: Yes, the ticket number 4 88. That seems like another priority.
10 00:01:57.670 ⇒ 00:02:11.280 Andrew O’Neil: but yeah, and and it seems like for those or Gtm the blockers are no longer there, because I do have access to the platform, which is fantastic and then for 4 88.
11 00:02:11.490 ⇒ 00:02:19.809 Andrew O’Neil: Yeah, just kind of getting into their meta events. Manager. If that’s where we want to start. I would just need access to that.
12 00:02:20.890 ⇒ 00:02:25.239 Robert Tseng: Okay, yeah, I’m gonna break out that ticket more, I mean, like.
13 00:02:27.123 ⇒ 00:02:29.959 Robert Tseng: don’t I can’t like speed up access for you and.
14 00:02:29.960 ⇒ 00:02:30.690 Andrew O’Neil: 38.
15 00:02:31.009 ⇒ 00:02:40.920 Robert Tseng: Think that’s why I mean, I have everything pulled up if you if you want to, just like Look, and I would want to just talk through that on this call.
16 00:02:41.240 ⇒ 00:02:42.090 Andrew O’Neil: That works.
17 00:02:42.240 ⇒ 00:02:51.346 Robert Tseng: Okay. So what I’m gonna do is I’m gonna update it like, the order is Meta north beam, reddit pinterest. And then Google. So those are the 5 things.
18 00:02:52.340 ⇒ 00:02:56.024 Robert Tseng: I don’t think this ticket really covers it. So I
19 00:02:57.780 ⇒ 00:03:01.080 Robert Tseng: I’m just gonna have amber. Go break it up alright.
20 00:03:04.220 ⇒ 00:03:05.710 Robert Tseng: You write a quick passage.
21 00:03:06.120 ⇒ 00:03:06.899 Andrew O’Neil: No worries.
22 00:03:16.060 ⇒ 00:03:17.930 Robert Tseng: Direct name. Reddit.
23 00:03:35.480 ⇒ 00:03:43.430 Robert Tseng: Okay? So we’ll do that. And then, so that’s yeah. So those are the. Those are the platforms that we’ll go after. And then.
24 00:03:43.430 ⇒ 00:03:43.800 Andrew O’Neil: -
25 00:03:45.290 ⇒ 00:03:59.599 Robert Tseng: yeah, pretty much like any questions you have about where the event is coming from. Like, what is it like? What is what it captures like, I mean, that’s coming from the data warehouse that we maintain. So I think. That’s kind of where we’re here and
26 00:04:00.168 ⇒ 00:04:07.119 Robert Tseng: I don’t know if you’re familiar with segment, but if you’re not like I I can. I can push things in segment to where they need to go.
27 00:04:08.000 ⇒ 00:04:13.360 Andrew O’Neil: Sounds good. Yeah, I’ve done a little bit of work with with segments. Yeah, it’s it’s a super cool tool.
28 00:04:13.780 ⇒ 00:04:14.185 Robert Tseng: Okay.
29 00:04:14.850 ⇒ 00:04:22.899 Robert Tseng: cool? Well, then, yeah, I would say, like the the Google Tag manager stuff, like, I mean, I’ll just also I’ll just my screen so you can see what I’m doing.
30 00:04:23.230 ⇒ 00:04:23.819 Andrew O’Neil: Yeah, yeah.
31 00:04:24.500 ⇒ 00:04:25.860 Robert Tseng: As far as like.
32 00:04:30.530 ⇒ 00:04:36.050 Robert Tseng: okay, whatever this ends up being, it’s gonna be split to 5. Like.
33 00:04:36.670 ⇒ 00:04:54.470 Robert Tseng: Yeah, I would say, the the platform getting conversion events. Firing to the platforms accurately is the biggest priority the Google Tag manager audit like, I think that’s more of like we could do that if I I as you’re blocked on other things like, I know that we decide some time for you to
34 00:04:55.222 ⇒ 00:05:03.077 Robert Tseng: I forgot if we said 10 HA week, or whatever. But whatever time you have, you can kind of use use there and then.
35 00:05:03.630 ⇒ 00:05:09.299 Robert Tseng: I mean I don’t. I don’t know. I think this is kind of vague, so I’ll just
36 00:05:11.690 ⇒ 00:05:13.798 Robert Tseng: you could not worry about that.
37 00:05:14.150 ⇒ 00:05:15.220 Andrew O’Neil: Okay, perfect.
38 00:05:15.220 ⇒ 00:05:17.579 Robert Tseng: I’m just trying to simplify things for you, and then.
39 00:05:17.580 ⇒ 00:05:18.659 Andrew O’Neil: Yeah, this is great.
40 00:05:21.040 ⇒ 00:05:26.790 Robert Tseng: I mean, I think this is kind of self explanatory, so I don’t really know this is necessary to say
41 00:05:28.650 ⇒ 00:05:36.019 Robert Tseng: no, I don’t. I don’t think you should remove anything here, so sorry I don’t. I haven’t been writing the tickets for this project, and.
42 00:05:36.020 ⇒ 00:05:37.139 Andrew O’Neil: No worries, no worries.
43 00:05:37.140 ⇒ 00:05:41.132 Robert Tseng: Looking at it. And I’m like, it’s it’s kind of strange to me.
44 00:05:42.060 ⇒ 00:05:44.550 Robert Tseng: yeah, I mean, I would say, those are like the 2
45 00:05:44.820 ⇒ 00:05:47.740 Robert Tseng: 2 main things for for this week. So
46 00:05:48.487 ⇒ 00:05:59.539 Robert Tseng: yeah, whatever you need to kind of help with that like, this is like, I’ll be paying attention to this most closely, so I will do whatever I can to get you unblocked faster. There.
47 00:05:59.780 ⇒ 00:06:00.940 Andrew O’Neil: Perfect.
48 00:06:02.690 ⇒ 00:06:10.349 Robert Tseng: Yeah. So maybe what would be helpful is just to kind of share briefly about like kind of segments and kind of what we have going on.
49 00:06:12.140 ⇒ 00:06:17.420 Robert Tseng: yeah. So you’ll you’ll you’ll go into Google tag manager, and you’ll kind of understand the workspace like, I think
50 00:06:18.100 ⇒ 00:06:21.419 Robert Tseng: you know this at the end of the day. It’s
51 00:06:22.740 ⇒ 00:06:28.100 Robert Tseng: like there is a lot of stuff that’s set up here. They’re like, you know, 60 tags. It’s probably yeah. And then
52 00:06:28.721 ⇒ 00:06:35.180 Robert Tseng: I think there’s like a there’s a couple parallel work streams, one on the
53 00:06:35.873 ⇒ 00:06:59.149 Robert Tseng: from the CTO’s perspective. He uses Google Tag manager to set up all this tagging and tracking so he can have like helps him with error. Logging on like seeing of the website is down like. But obviously everything is like kind of browser based tagging. And so it’s not the most reliable for conversions, which is why, we’ve been sending.
54 00:06:59.240 ⇒ 00:07:24.380 Robert Tseng: Well, we’ve been working towards sending conversion events from the server and by the server, it’s just the warehouse. So it’s there isn’t really a server. It’s not our own proprietary product. It’s a 3rd party software that sends us web hooks. That we kind of consume in the warehouse, and we do stuff to it and then push it back into
55 00:07:25.298 ⇒ 00:07:32.479 Robert Tseng: you know, platforms like like Facebook. So I set up this Facebook destination in this production.
56 00:07:33.219 ⇒ 00:07:40.489 Robert Tseng: Thing, there’s this payload that you can kind of investigate and basically.
57 00:07:41.630 ⇒ 00:07:45.589 Robert Tseng: if I were to just take a snippet of the latest data.
58 00:07:47.390 ⇒ 00:07:54.879 Robert Tseng: Yeah, it’s just like, you know, a purchase event with some hashed user data like.
59 00:07:55.390 ⇒ 00:08:02.986 Robert Tseng: And then some value of the transaction, and that gets fired into Facebook here.
60 00:08:03.960 ⇒ 00:08:12.919 Robert Tseng: before we were sending it into a custom event called order completed. Once I verify that there was actually stuff showing up here.
61 00:08:13.140 ⇒ 00:08:20.898 Robert Tseng: I kind of I I pointed it into the purchase event, because
62 00:08:21.570 ⇒ 00:08:28.920 Robert Tseng: I think, from a campaign optimization perspective. It’s better to have it in the native purchase event than it is on a
63 00:08:29.570 ⇒ 00:08:30.800 Robert Tseng: custom event
64 00:08:31.344 ⇒ 00:08:37.739 Robert Tseng: and that’s pretty much the extent of like my expertise there, like, I can like that where things kind of
65 00:08:38.260 ⇒ 00:08:41.189 Robert Tseng: like. I can see that the data went in like I think.
66 00:08:41.190 ⇒ 00:08:41.690 Andrew O’Neil: Yep.
67 00:08:41.690 ⇒ 00:08:43.319 Robert Tseng: Nuances are like.
68 00:08:43.549 ⇒ 00:08:49.039 Robert Tseng: what else we need to? Does this actually matter? Do we need to increase coverage here and like kind of
69 00:08:50.730 ⇒ 00:08:53.089 Robert Tseng: like making sure that this is.
70 00:08:54.030 ⇒ 00:08:58.450 Andrew O’Neil: Like. I don’t really know how to give the green light to this. To be honest, like, I just kind of
71 00:08:58.450 ⇒ 00:08:58.970 Andrew O’Neil: sure.
72 00:08:58.970 ⇒ 00:09:00.399 Robert Tseng: I can. I can.
73 00:09:00.720 ⇒ 00:09:07.065 Robert Tseng: I could tell you if the data is being sent or not like that’s I think that’s that’s all. I really, that’s all I really know.
74 00:09:07.310 ⇒ 00:09:07.920 Andrew O’Neil: Yeah, yeah.
75 00:09:08.310 ⇒ 00:09:11.499 Robert Tseng: Yeah, I think just hoping for you to kind of
76 00:09:12.890 ⇒ 00:09:22.500 Robert Tseng: give us like a point of view on this setup. And then, like, we’ll move on to each of the other platforms is pretty much the how we, how I thought about it.
77 00:09:22.770 ⇒ 00:09:28.790 Andrew O’Neil: Perfect. And when you refer to the date of the warehouse, is that segment? Or is that a different platform.
78 00:09:29.270 ⇒ 00:09:36.020 Robert Tseng: No, yeah, that’s in bigquery. So is our data. Warehouse. Segment is just a data connector tool.
79 00:09:36.020 ⇒ 00:09:36.730 Andrew O’Neil: Right.
80 00:09:36.730 ⇒ 00:09:38.040 Robert Tseng: Just move data around.
81 00:09:38.510 ⇒ 00:09:45.190 Andrew O’Neil: Yep, perfect. And then are you sending any data from segment into tag manager?
82 00:09:47.960 ⇒ 00:09:49.670 Robert Tseng: Hi!
83 00:09:49.990 ⇒ 00:09:50.940 Robert Tseng: Think!
84 00:09:52.800 ⇒ 00:09:53.939 Robert Tseng: No, we’re not.
85 00:09:54.310 ⇒ 00:10:15.881 Andrew O’Neil: Okay, cause. So 1st glance it looks like, so I’m I’m looking at their tag manager right now. They do have or within tag manager. They are sending events both to like, for example, the Meta pixel as well as the conversions. Api.
86 00:10:16.460 ⇒ 00:10:24.835 Andrew O’Neil: so like, if you search you think? Yeah, you’re in the the normal browser container. If you go to tags instead of triggers,
87 00:10:25.170 ⇒ 00:10:25.890 Robert Tseng: Well.
88 00:10:25.890 ⇒ 00:10:34.470 Andrew O’Neil: And then search Meta. That’s what they’ve labeled all the meta ones or just purchase either way.
89 00:10:35.260 ⇒ 00:10:37.230 Andrew O’Neil: yes. You see the Meta.
90 00:10:38.252 ⇒ 00:10:43.329 Andrew O’Neil: Subscribe. There’s so there’s a 1st time purchase, and then there’s a return purchase.
91 00:10:44.910 ⇒ 00:10:46.619 Andrew O’Neil: Yes, if you click into that.
92 00:10:47.483 ⇒ 00:10:55.759 Andrew O’Neil: you can see, it’ll open up. Yeah. So this is firing a purchase event. Okay, it’s pause. That’s that’s good.
93 00:10:56.260 ⇒ 00:11:00.169 Andrew O’Neil: Although that’s weird, because it’s not showing paused for me.
94 00:11:00.270 ⇒ 00:11:01.720 Andrew O’Neil: Let’s see.
95 00:11:01.940 ⇒ 00:11:09.190 Robert Tseng: I did ask for deposit like late last week because we I didn’t wanna have redundant stuff from like.
96 00:11:09.730 ⇒ 00:11:16.370 Andrew O’Neil: Perfect. Perfect. Yes, cause that, maybe. Okay. Maybe that’s if I go to.
97 00:11:16.700 ⇒ 00:11:20.720 Andrew O’Neil: Okay, let me see if my version is wrong.
98 00:11:21.180 ⇒ 00:11:24.820 Andrew O’Neil: Let’s see Meta.
99 00:11:25.891 ⇒ 00:11:29.239 Andrew O’Neil: Okay, yes, I’m seeing the Meta 1st time purchase
100 00:11:29.590 ⇒ 00:11:37.280 Andrew O’Neil: that is paused. Okay, good. They have to subscribe one. Now, the important thing, though.
101 00:11:37.680 ⇒ 00:11:40.089 Andrew O’Neil: if we go to Meta.
102 00:11:42.310 ⇒ 00:11:50.129 Andrew O’Neil: They did not pause it in their server container. So if you, if you go back to Gtm.
103 00:11:50.480 ⇒ 00:11:50.805 Robert Tseng: Yep.
104 00:11:52.630 ⇒ 00:12:02.750 Andrew O’Neil: So right now you’re in their their browser container. That’s the Gtm, dot. Try eden.com if you switch to. Yeah the Sgtm. And then similarly search for Meta.
105 00:12:05.370 ⇒ 00:12:09.188 Andrew O’Neil: You see, the the 1st time purchase is still there.
106 00:12:09.570 ⇒ 00:12:10.340 Robert Tseng: Yeah.
107 00:12:10.340 ⇒ 00:12:26.990 Andrew O’Neil: So we’ll need to make verify that this is not sending it, but based on the what you were showing me in Meta. If you want to flip to that? See how on the 17, th the line, the blue and the green? We’re pretty much one to one. But then
108 00:12:27.090 ⇒ 00:12:50.070 Andrew O’Neil: on the 17.th It’s there’s all of a sudden this divergence. What that looks like to me is that we’re over inflating is the is the teal. Is that the server that would make sense. Yeah. So they didn’t pause the server. So that’s why Server stayed high. Browser went down because we we pause the duplicate So
109 00:12:50.230 ⇒ 00:13:01.349 Andrew O’Neil: for for the last, and call it 10 days they’ve been sending 2 server events and only one browser event, which is why we’re seeing that divergence there.
110 00:13:01.350 ⇒ 00:13:07.190 Robert Tseng: Yeah, I mean doesn’t matter like I mean, I don’t find this reliable cause. It’ll say like.
111 00:13:07.440 ⇒ 00:13:11.370 Robert Tseng: Oh, events may be discarded, due to deduplication.
112 00:13:11.370 ⇒ 00:13:12.180 Andrew O’Neil: Oh.
113 00:13:12.320 ⇒ 00:13:27.669 Andrew O’Neil: yeah, they’re they’re okay. Yeah. So if you want to look at that to quick sense on that, they say that in practice. They’re not amazing at it. But we can see if you click view details there within Meta.
114 00:13:27.900 ⇒ 00:13:31.379 Robert Tseng: Or like a yeah, just that button. There.
115 00:13:31.700 ⇒ 00:13:34.969 Andrew O’Neil: And then go to event deduplication.
116 00:13:35.600 ⇒ 00:13:39.580 Andrew O’Neil: So this will give us a sense of how it’s going.
117 00:13:39.760 ⇒ 00:13:45.660 Andrew O’Neil: Yes. So the fact that you see that the server events have such a low overlap rate
118 00:13:45.790 ⇒ 00:13:51.859 Andrew O’Neil: that to me the so there. So this tells me that Meta is struggling to dedupe things.
119 00:13:52.200 ⇒ 00:14:21.779 Andrew O’Neil: Now, generally this is due to server events missing a key that they could match with a browser. But my hunch is that it’s it’s looking poor because we’re getting server, both from the segment data push as well as our server Google tag, manager, container. So what I’m trying to say is that I think the fix we should try 1st pausing the
120 00:14:22.030 ⇒ 00:14:28.020 Andrew O’Neil: purchase in the server container and that should resolve most things, and then
121 00:14:28.300 ⇒ 00:14:51.460 Andrew O’Neil: from there there might be additional work we can do to help improve the dedupe rate. But definitely. And I would want to look into this a little bit further to make sure. But yeah, from the glance of what you showed me and the fact that we pause the browser, but not the server. That seems like we need to. We need to do the pause in the the server as well.
122 00:14:52.170 ⇒ 00:14:58.673 Robert Tseng: Yeah, okay, so I do that, I can submit that
123 00:14:59.780 ⇒ 00:15:00.320 Andrew O’Neil: Single.
124 00:15:00.820 ⇒ 00:15:02.530 Robert Tseng: I’ll just say
125 00:15:05.980 ⇒ 00:15:09.552 Robert Tseng: Oh, dude! I don’t know what they’ve been naming their versions, but
126 00:15:11.660 ⇒ 00:15:17.509 Andrew O’Neil: Let’s see, it just looks like they’re yes, pretty standard. They’re just like telling, like, usually you just say, like what you’re doing.
127 00:15:17.750 ⇒ 00:15:18.500 Robert Tseng: Oh, okay.
128 00:15:18.500 ⇒ 00:15:22.209 Andrew O’Neil: Yeah, he’s I don’t think he’s taken. It’s not like a full release.
129 00:15:26.320 ⇒ 00:15:26.990 Robert Tseng: Bye.
130 00:15:29.710 ⇒ 00:15:35.620 Andrew O’Neil: And with Gtm, you can always roll back a version. If they they don’t like what we did, they can. They can change that.
131 00:15:35.990 ⇒ 00:15:44.090 Robert Tseng: Yeah, okay, okay, cool. So we’ll we’ll see kind of like, how that ends up going. I think.
132 00:15:46.050 ⇒ 00:15:50.130 Robert Tseng: yeah. So I mean, I find this to be a bit like redundant like.
133 00:15:50.440 ⇒ 00:15:51.430 Andrew O’Neil: Yes.
134 00:15:51.870 ⇒ 00:16:01.389 Robert Tseng: So there, like the same directive that we have to go and do our version of server side tagging like the CTO. Is doing his own thing like he’s just.
135 00:16:01.510 ⇒ 00:16:14.550 Andrew O’Neil: Right? Right? Right? Right? Yes, yeah, they’re they’re kind of running 2 implementations. So you have 2 version, yeah, 2 server Ver versions of kind of each tag. And yeah, so that kind of leads me to.
136 00:16:14.720 ⇒ 00:16:18.530 Andrew O’Neil: You know where you want me to focus of, you know, if
137 00:16:18.720 ⇒ 00:16:44.219 Andrew O’Neil: if the goal and if one of the primary goals is to use segment to send the data to the platforms rather than tag manager, that’s fine. And I think the the recommended approach would be. As we set up tracking and segment we pause the equivalent tags and tag manager just like we did with Meta, so that we can avoid, you know, over inflating their numbers.
138 00:16:46.300 ⇒ 00:16:47.160 Robert Tseng: Yeah,
139 00:16:54.760 ⇒ 00:16:56.859 Robert Tseng: okay, maybe I will.
140 00:17:03.501 ⇒ 00:17:19.680 Robert Tseng: Okay, well, so yeah, I mean, I I understand that there’s like, just that. There’s like duplicate effort going on. I mean, I can get more clarity on this as well. But my understanding is that the Google Tag manager, like Approach, has not been working, which is, he’s been doing this.
141 00:17:19.680 ⇒ 00:17:20.079 Andrew O’Neil: Sure.
142 00:17:20.089 ⇒ 00:17:29.769 Robert Tseng: Response. No one in marketing believes that the conversions are correct. Because he’s not pulling the data from our warehouse. He’s he’s pulling it.
143 00:17:29.879 ⇒ 00:17:53.569 Robert Tseng: I don’t really know. I can’t even really tell. Like, from looking through some of the stuff where he’s getting the data from. I just know it’s not coming from from us. And I think there’s a consensus that like when we report purchase events, they’re more accurate than what whatever else he seems to be pulling it from and which is why we’re being asked to like, kind of just take over that part of the scope.
144 00:17:53.729 ⇒ 00:17:57.189 Robert Tseng: And yeah, I
145 00:17:57.879 ⇒ 00:18:03.939 Robert Tseng: I think what I will do is I will make sure that he’s aware that this is what we’re doing.
146 00:18:04.219 ⇒ 00:18:10.439 Robert Tseng: Like, we’re we’re basically gonna yeah, we’re we’re trying to run our
147 00:18:11.302 ⇒ 00:18:14.679 Robert Tseng: conversion event over the same platforms.
148 00:18:14.959 ⇒ 00:18:18.279 Robert Tseng: And we’re gonna be replacing what he’s doing. But like.
149 00:18:18.280 ⇒ 00:18:19.590 Andrew O’Neil: Yes, yes.
150 00:18:19.590 ⇒ 00:18:26.660 Robert Tseng: Yeah, I don’t necessarily know how he’s. I mean, he’s not gonna take it well or whatever. But it’s just not. Whatever we’ll we’ll just have to deal with that.
151 00:18:27.260 ⇒ 00:18:27.810 Andrew O’Neil: Go ahead!
152 00:18:28.220 ⇒ 00:18:29.780 Robert Tseng: Yeah, so
153 00:18:32.970 ⇒ 00:18:33.880 Robert Tseng: I guess.
154 00:18:34.750 ⇒ 00:18:45.360 Andrew O’Neil: Do you think like cause? It sounds like you? You there’s an approach you did. Where? And, for example, with Meta, you, instead of sending it to purchase initially, you sent the event like an order created.
155 00:18:45.742 ⇒ 00:19:11.080 Andrew O’Neil: You know. Maybe like one way we could show the the CTO, you know. Hey, like our approach is better, is this, was the number of order completed events like our new event, and then over the same timeframe. This is what purchase showed. And like, look like we’re, you know, now, 95% of the true value, whereas before, maybe you were 75% that might, you know, help persuade them to.
156 00:19:12.600 ⇒ 00:19:18.639 Robert Tseng: Okay, that’s fair. Yeah, I can. Definitely. Yeah, I I should own the comms there on, like.
157 00:19:19.320 ⇒ 00:19:37.774 Robert Tseng: yeah, we’re not gonna replace everything he did as I said, like immediately, we’re just gonna create separate conversion kind of custom events show them that our conversion is more accurate by a factor of whatever. And then kind of give us the green light to go and roll it out. So okay, that
158 00:19:39.170 ⇒ 00:19:43.720 Robert Tseng: yeah, okay. And then in that case,
159 00:19:48.730 ⇒ 00:19:50.900 Robert Tseng: yeah, I guess. Where? Where do we go from here?
160 00:19:51.240 ⇒ 00:19:52.430 Robert Tseng: Like, I’ll sorry that.
161 00:19:52.430 ⇒ 00:19:58.560 Robert Tseng: And like, kind of where? How do you? How do you see your like, what? What do you? What would you like? How would you approach the rest of this? Then.
162 00:19:58.980 ⇒ 00:20:09.490 Andrew O’Neil: Yeah. So in segments have we? Set up the push to like north beam and all the other platforms? Or is is Meta the 1st kind of guinea pig for that.
163 00:20:10.120 ⇒ 00:20:14.809 Robert Tseng: No, yeah, that is the 1st one. I don’t think that segment will go to North Main.
164 00:20:15.340 ⇒ 00:20:16.890 Robert Tseng: like, if I were to. Just
165 00:20:22.040 ⇒ 00:20:26.850 Robert Tseng: I think you’re we’re we’re gonna have to do our own custom thing to get get into our team.
166 00:20:28.500 ⇒ 00:20:31.960 Andrew O’Neil: Okay, yeah. So
167 00:20:32.160 ⇒ 00:20:58.833 Andrew O’Neil: again, like, it sounds like the there’s kind of 2 work streams like one. We want to set up conversion tracking through segment. And then that’s kind of like a a parallel track to which we are gonna clean up their Gtm. By hopefully replacing their prior conversion, tracking with our better conversion, tracking with segment and that that those 2 will kind of go hand in hand.
168 00:20:59.770 ⇒ 00:21:01.509 Andrew O’Neil: I don’t know. Like if you
169 00:21:02.037 ⇒ 00:21:11.070 Andrew O’Neil: if you’re gonna be owning the segment piece of it, sending the data from segment to the platforms and you want me to like help
170 00:21:12.790 ⇒ 00:21:18.780 Andrew O’Neil: like Qa, the data when it gets accepted by the platforms. Or if you want
171 00:21:20.220 ⇒ 00:21:28.729 Andrew O’Neil: cause. Yeah, like, I can, I can be as involved or as as like leading or or supporting you. Kind of depending on your preference.
172 00:21:29.530 ⇒ 00:21:32.190 Robert Tseng: Yeah, I mean, I guess, like, kind of what?
173 00:21:33.240 ⇒ 00:21:34.910 Robert Tseng: Well, I’m just thinking.
174 00:21:37.460 ⇒ 00:21:50.140 Robert Tseng: if we push our event into Google tag manager, then the workflow stays in Google tag manager. And you would be able to kind of connect to anything from from this environment, right?
175 00:21:50.540 ⇒ 00:22:02.579 Robert Tseng: Or it would have to just create direct connections that don’t go through Google Tag manager. Some of it will go through segment. Some of it will not like that seems a bit more complicated. I don’t really know what the risks are between those 2.
176 00:22:05.320 ⇒ 00:22:24.820 Andrew O’Neil: I mean, I think, from a just a high, level architecture standpoint, the it would be probably a better setup to like of like kind of go around Google tag manager to be honest with you than to like use like push, push data from segment, or like another warehouse
177 00:22:25.050 ⇒ 00:22:34.909 Andrew O’Neil: from segment into Gtm, and then that goes to the like that. It seems like an unnecessary push. Because the only thing I guess I could us with
178 00:22:35.610 ⇒ 00:22:48.600 Andrew O’Neil: Google tag manager that I don’t know if segment can do is help transform the data. So that if it’s like whatever form it’s in, we can get it to a a payload that
179 00:22:48.890 ⇒ 00:23:01.470 Andrew O’Neil: either Facebook or North be, you know, whatever format the platform needs it in Gtm does help. We know with that if that can be done in segment, then, you know, I don’t think we really need to send the data to Gtm. Honestly.
180 00:23:01.860 ⇒ 00:23:02.500 Robert Tseng: Yeah.
181 00:23:02.640 ⇒ 00:23:13.459 Robert Tseng: So I think there will always be some events that should be just tracking cool tag manager, because they’re just it’s easier to set up there. Only, for, like the like, purchase events, or something.
182 00:23:13.460 ⇒ 00:23:13.960 Andrew O’Neil: Sure.
183 00:23:13.960 ⇒ 00:23:18.529 Robert Tseng: Super critical, though I feel like we need to go through this through segment.
184 00:23:18.530 ⇒ 00:23:19.849 Andrew O’Neil: Yeah, that makes sense.
185 00:23:20.090 ⇒ 00:23:20.920 Robert Tseng: Okay.
186 00:23:21.090 ⇒ 00:23:21.870 Andrew O’Neil: That makes sense.
187 00:23:22.350 ⇒ 00:23:22.960 Andrew O’Neil: Yep.
188 00:23:22.960 ⇒ 00:23:23.650 Robert Tseng: Okay?
189 00:23:24.580 ⇒ 00:23:28.859 Robert Tseng: And then, yeah, I mean, as far as like.
190 00:23:30.340 ⇒ 00:23:46.519 Robert Tseng: so I think, kind of the Google Tag manager audit is like, I don’t really know, like what what will be, what’s like a reasonable objective for this like, I don’t want to spend too much time on this, like whatever they have like works, I guess. But it’s kinda just
191 00:23:47.030 ⇒ 00:23:49.770 Robert Tseng: I don’t really know.
192 00:23:51.770 ⇒ 00:23:58.380 Robert Tseng: I mean, I don’t necessarily want to get more involved in the Google in the Google Tag manager stuff. But
193 00:23:58.710 ⇒ 00:24:04.090 Robert Tseng: yeah, just it’s kind of like, if I were to
194 00:24:04.260 ⇒ 00:24:10.050 Robert Tseng: like the way I would want to pitch, it is like, Hey, like leader like leadership.
195 00:24:10.180 ⇒ 00:24:17.010 Robert Tseng: I mean, you could end up being the Google Tag manager person for this moving forward, you know. And it’s like.
196 00:24:18.203 ⇒ 00:24:41.870 Robert Tseng: if we were to hand this off to Andrew. He would make these changes to our environment. It would, and that would be, lead to whatever Xyz improvements and then moving forward. He can just own that scope. So the CTO does does no longer has to be touching Google tag manager. Like, I, I want to make a proposal of like some sort of hand off here.
197 00:24:42.257 ⇒ 00:24:49.339 Robert Tseng: But I don’t know if that’s kind of I mean that’s where that’s what I think. But I don’t know anything. Yeah.
198 00:24:50.020 ⇒ 00:25:01.330 Andrew O’Neil: Yeah, I mean, from a clean, what cleanup generally looks like is you’re removing thankfully from from 1st glance of looking through their platform. You’ll find like
199 00:25:01.740 ⇒ 00:25:21.219 Andrew O’Neil: a tool that maybe they tried out 3 years ago. They’re not using. There’s still tags, or, you know, code for that lingering and tag manager, you know. So you go ahead and like, remove that of like, hey, like, are you still using app loving? Are you still using a hrefs like, you know. Generally, things like that
200 00:25:21.500 ⇒ 00:25:31.119 Andrew O’Neil: that’s like, you know, at a high level, from a tag perspective, you’re checking active platforms against what platforms they actually have tags for
201 00:25:31.630 ⇒ 00:25:40.230 Andrew O’Neil: And then, similarly, with triggers, it’s a little bit easier, like if you click over to that you can actually see it a little bit more easily.
202 00:25:40.704 ⇒ 00:26:07.775 Andrew O’Neil: If you sort do you see the second or the tags column there? Where it shows number? Yeah. So, for example. So you see that, or click again those first.st What is that? 10 those are not being used at all by a tag, but they’re they’re existing. So that that’s gonna create a little bit of bloat on their site. So you know, you go ahead and like, remove all of the triggers that are not being used by any tag.
203 00:26:08.090 ⇒ 00:26:08.420 Robert Tseng: Yeah.
204 00:26:08.420 ⇒ 00:26:18.280 Andrew O’Neil: Similarly with variables like, you know, there’s variables, because each of these like things that get created one. It’s a little bit messy, but 2. It can
205 00:26:18.390 ⇒ 00:26:31.809 Andrew O’Neil: add to the file size that your site is loading when you are loading. Gtm, so cleaning it up and auditing it helps to reduce that and make sure that you’re only running, tagging code that you need to be.
206 00:26:33.320 ⇒ 00:26:34.100 Robert Tseng: I see.
207 00:26:35.570 ⇒ 00:26:54.777 Robert Tseng: Okay? Well, yeah, I mean, if you can give me a point of view on like, if you were to own this environment what you would do differently and like just, I would like to be able to tell the team that by handing it over to Andrew to maintain like there is a marginal improvement.
208 00:26:55.570 ⇒ 00:26:59.009 Robert Tseng: because sites gonna be faster or like,
209 00:27:01.780 ⇒ 00:27:06.470 Robert Tseng: Yeah, like, I just want, I just want to be able to like measure this a bit more
210 00:27:07.150 ⇒ 00:27:18.649 Robert Tseng: cause? Yeah, yeah, I think, I think that that would help build this case. Yeah, I don’t want to go all into it because there is a bit of a dance there is like somebody there doing it right now.
211 00:27:19.500 ⇒ 00:27:19.820 Andrew O’Neil: Right.
212 00:27:19.820 ⇒ 00:27:25.349 Robert Tseng: I would like to take it off his plate so he can go do other things. But I don’t know if.
213 00:27:25.690 ⇒ 00:27:30.110 Robert Tseng: Well, yeah, I just I don’t know what by doing that. Would it actually be better so.
214 00:27:31.540 ⇒ 00:27:32.080 Andrew O’Neil: Yeah.
215 00:27:32.360 ⇒ 00:27:36.139 Robert Tseng: So, yeah, okay, so just to kind of close this out.
216 00:27:37.540 ⇒ 00:27:44.460 Robert Tseng: you know, on the meta side, remove some things here and there. And then, yeah, I’ll kind of get a.
217 00:27:45.450 ⇒ 00:27:51.599 Robert Tseng: We’ll have all the tickets for the other platforms kind of laid out and just have a clear understanding of how we’re gonna go after that.
218 00:27:52.260 ⇒ 00:28:01.620 Robert Tseng: And then I’ll have a better perspective on how to involve you there, like, in terms of
219 00:28:02.950 ⇒ 00:28:14.119 Robert Tseng: sending like getting. It’s basically wanting the marketing team to feel like North beam, reddit interest are all set up and clean and ready to go, whatever that.
220 00:28:16.030 ⇒ 00:28:20.870 Robert Tseng: Yeah. And then also, like on the for the CTO, just wanting to know, like
221 00:28:21.050 ⇒ 00:28:25.640 Robert Tseng: you’re gonna do a kind of a look over the
222 00:28:25.910 ⇒ 00:28:28.610 Robert Tseng: Gtm environment and and let me know like.
223 00:28:29.570 ⇒ 00:28:33.590 Robert Tseng: if you own this, would it be better and like, How would you measure that very much?
224 00:28:34.260 ⇒ 00:28:36.150 Andrew O’Neil: Sure, yeah, yeah, absolutely.
225 00:28:36.360 ⇒ 00:28:39.838 Robert Tseng: Okay, cool. Alright. Then that sounds like a good plan.
226 00:28:40.140 ⇒ 00:28:40.750 Andrew O’Neil: Cool.
227 00:28:41.500 ⇒ 00:28:43.219 Robert Tseng: Yeah. Okay, cool.
228 00:28:44.560 ⇒ 00:28:46.129 Robert Tseng: Alright, thanks, Andrew, helpful.
229 00:28:46.130 ⇒ 00:28:49.069 Andrew O’Neil: Yeah, absolutely, Robert. I’ll talk to you in a bit over slack.