Meeting Title: Catalyst Data Integration Planning Date: 2025-09-15 Meeting participants: Henry Zhao, Kevin Bell, Robert Tseng


WEBVTT

1 00:00:19.920 00:00:21.439 Henry Zhao: Hey, Kevin, how’s it going?

2 00:00:23.740 00:00:25.159 Kevin Bell: Hey, Henry, what’s going on?

3 00:00:25.940 00:00:30.440 Henry Zhao: Robert’s gonna be joining this call, too, so we’ll give him a minute to join. But thanks for joining.

4 00:00:31.410 00:00:32.340 Kevin Bell: Yeah, no problem.

5 00:01:01.280 00:01:02.489 Henry Zhao: Hey, Robert, how’s it going?

6 00:01:04.519 00:01:05.699 Robert Tseng: Good, how are you guys?

7 00:01:06.069 00:01:08.010 Kevin Bell: It’s pretty good, man, pretty good.

8 00:01:09.430 00:01:10.009 Henry Zhao: Alright, cool.

9 00:01:10.010 00:01:12.159 Robert Tseng: watched, Zoran’s video yet?

10 00:01:12.160 00:01:12.620 Henry Zhao: I have…

11 00:01:12.620 00:01:14.110 Robert Tseng: I should do that.

12 00:01:16.200 00:01:20.879 Kevin Bell: I got through a couple minutes, but yeah, let’s… we can watch through… or not a couple minutes, couple… first…

13 00:01:21.500 00:01:23.389 Kevin Bell: 20 seconds of it. But yeah, we can watch that.

14 00:01:24.430 00:01:28.519 Robert Tseng: If it’s relevant to what you guys are discussing, I’m just saying, like, I… yeah, so…

15 00:01:31.540 00:01:34.460 Henry Zhao: Evan, do you want to take the first few minutes just to watch that video, or do you want to…

16 00:01:34.950 00:01:36.690 Kevin Bell: Yeah, give me a minute, let me pull it up.

17 00:01:36.690 00:01:39.340 Henry Zhao: Yeah, I’ll watch it as well, and I’ll come back in 5 minutes.

18 00:05:01.310 00:05:04.569 Kevin Bell: Let me know when you guys done, I was able to watch it.

19 00:05:13.340 00:05:15.570 Robert Tseng: I’m good too, so whenever Henry’s back.

20 00:05:33.340 00:05:44.160 Henry Zhao: Cool, so I just watched the video, and then just to talk to Robert real quick, Robert, I think all we really need to do then is just to stitch it together in BigQuery, using segment, and have first touch, last touch.

21 00:05:44.440 00:05:55.880 Henry Zhao: And then for the conversation with Kevin, Kevin, I just want to know how we send that data to Catalyst, so we make sure that it’s ready to basically let them know how much we owe them, right? So…

22 00:05:56.310 00:05:58.359 Henry Zhao: That’s kind of what the reason for this call is.

23 00:05:58.690 00:06:12.759 Kevin Bell: Gotcha. So we use, so Catalyst is a tracking software, so we use their Pixel, which is integrated, or we have it set up through Google Tag Manager, and I believe we’re doing the…

24 00:06:13.340 00:06:17.590 Kevin Bell: Some of the conversions are modeled out in BigQuery as well.

25 00:06:17.590 00:06:18.630 Henry Zhao: Okay, perfect.

26 00:06:18.900 00:06:24.029 Henry Zhao: Do we… do they charge us based on first touch, last touch, any touch? Like, how does… how does that work?

27 00:06:25.670 00:06:29.190 Kevin Bell: Ew, first touch, I believe. Let me comp…

28 00:06:29.410 00:06:37.330 Kevin Bell: Give me all the questions that you need, and then I can… I have a Slack channel with them. Matter of fact, we can just… I can just add you to the Slack channel, too, if you need.

29 00:06:37.420 00:06:38.460 Henry Zhao: Okay.

30 00:06:39.520 00:06:42.899 Kevin Bell: That might be a little bit… but yeah, let me know all the questions, and I can…

31 00:06:43.550 00:06:47.329 Kevin Bell: get it over to them. Where’s… where’s the channel here?

32 00:06:47.330 00:06:51.909 Henry Zhao: Robert, correct me if I’m wrong, but if it’s First Touch, I think it’s pretty straightforward, where we would just look at

33 00:06:52.130 00:06:56.350 Henry Zhao: any anonymous ID and segment where the first touch was

34 00:06:56.750 00:07:03.510 Henry Zhao: catalysts, or, like, any affiliate, right? And then we would just pay out any conversions where that was the first touch.

35 00:07:04.340 00:07:05.529 Henry Zhao: I think that’s pretty much it.

36 00:07:07.760 00:07:17.959 Robert Tseng: You’re saying that’s how we… that’s how we would pay out for affiliate? I mean, I don’t know, I guess I… okay. I mean, I don’t know if that’s… that’s for sure how we’ve been doing it. We’ve not really had to…

37 00:07:18.120 00:07:19.380 Robert Tseng: model affiliate.

38 00:07:19.530 00:07:21.280 Robert Tseng: conversions before.

39 00:07:21.280 00:07:27.370 Henry Zhao: Yeah, but then, like, if Cutter pays out other things last touch, then there’s still gonna be some duplication, right? Like…

40 00:07:28.750 00:07:32.730 Henry Zhao: if Catalyst brought… if Catalyst brought us the… the lead.

41 00:07:32.880 00:07:35.229 Henry Zhao: And then, like, a Facebook ad closed it.

42 00:07:35.400 00:07:39.990 Henry Zhao: then, you know, first touch would be a catalyst, but the last touch would be Facebook.

43 00:07:42.280 00:07:43.560 Kevin Bell: Yeah, I think…

44 00:07:43.560 00:07:44.629 Henry Zhao: I think the other…

45 00:07:45.160 00:07:46.719 Kevin Bell: Excuse me, is…

46 00:07:47.520 00:07:55.529 Kevin Bell: We’re gonna… we’re… it’s gonna take another week or two, but we’re getting customized landing pages and intakes just for affiliates.

47 00:07:56.210 00:07:57.710 Kevin Bell: And that should help.

48 00:07:58.340 00:08:00.959 Kevin Bell: Isolate as much of the crossover as possible.

49 00:08:01.420 00:08:02.879 Kevin Bell: Or reduce it, we’ll say.

50 00:08:02.880 00:08:09.170 Henry Zhao: Yeah, that’s fine, yeah, we just need to make sure that we understand whether it’s first touch or last touch, and that we stitch it together properly.

51 00:08:09.650 00:08:11.920 Kevin Bell: Yeah, I’m sending him a message now.

52 00:08:17.720 00:08:24.510 Henry Zhao: Cool, but ultimately, we’re the ones that provide the data to Catalyst, right? Or do they just take the data coming in from their Pixel and just automatically bill us?

53 00:08:27.730 00:08:31.000 Kevin Bell: Conversions… It takes…

54 00:08:37.510 00:08:41.790 Kevin Bell: It takes it from their… it does identify the conversions based upon their pixel.

55 00:08:43.390 00:08:49.180 Kevin Bell: So we send them all conversion data, and then they model it out based upon

56 00:08:49.380 00:08:55.870 Kevin Bell: They don’t model out, but then they identify conversions based upon the, the clicking IDs, the click IDs.

57 00:08:56.430 00:08:57.310 Kevin Bell: Purchase it.

58 00:09:04.650 00:09:07.120 Henry Zhao: Okay, that, that’s good to know.

59 00:09:07.770 00:09:10.190 Henry Zhao: So we’re not sending them any additional data?

60 00:09:11.320 00:09:12.450 Henry Zhao: From BigQuery?

61 00:09:13.740 00:09:17.019 Kevin Bell: We’re… we are sending them, like, they… they are…

62 00:09:17.780 00:09:22.829 Kevin Bell: based upon the pixel that I’m looking at, they are getting data, like, all conversion data.

63 00:09:24.190 00:09:27.239 Kevin Bell: So they’ll see all conversions, and they will only…

64 00:09:27.720 00:09:31.670 Kevin Bell: If the convert… as long as the conversion shows… and let me just pull it up…

65 00:09:32.730 00:09:37.330 Henry Zhao: And also, what’s counting as a conversion? Is it just filling… finishing the tri-edin intake flow?

66 00:09:40.280 00:09:41.859 Kevin Bell: Yes.

67 00:09:42.010 00:09:42.660 Henry Zhao: Okay.

68 00:09:44.460 00:09:48.589 Henry Zhao: So we might have to pay them even if they don’t… even if somebody doesn’t convert to revenue.

69 00:09:49.680 00:09:59.549 Robert Tseng: Well, isn’t… isn’t Cutter saying that he… he wants it to be as close to revenue as possible? Like, so, I don’t know if just filling out a form should be enough for this.

70 00:10:00.680 00:10:05.170 Kevin Bell: I mean, I get that he wants to do it as close as possible, but at the same time, it’s like…

71 00:10:06.770 00:10:10.200 Kevin Bell: You know, and maybe that’s the conversation that we have to have with them.

72 00:10:13.140 00:10:20.839 Kevin Bell: you know, it’s just… there’s always gonna be some overlap. You’re always gonna have a little bit more from affiliates than you really want.

73 00:10:21.840 00:10:26.879 Kevin Bell: So we gotta figure out what is the… what is a safe… Kind of number.

74 00:10:26.990 00:10:28.929 Kevin Bell: But you can see here.

75 00:10:28.930 00:10:38.919 Robert Tseng: Yeah, if they’re letting us send data to them, then why don’t we have more say in, like, what an actual conversion is, you know, if they’re not just using their own pixel.

76 00:10:41.390 00:10:43.470 Kevin Bell: Yes,

77 00:10:45.840 00:10:53.710 Kevin Bell: I think it comes down to, like, how we’re modeling out, and then also, like, how they’re identifying it. Because, you know, I’ve been in situations where it’s like.

78 00:10:54.000 00:10:55.170 Kevin Bell: where we’ve…

79 00:10:55.410 00:10:59.759 Kevin Bell: you know, I mean, we did… we tried to do the same thing at Mochi, which is like, alright, you know, let’s… let’s…

80 00:11:00.550 00:11:09.100 Kevin Bell: Do it based upon our touch system, but then ultimately, you end up discrediting affiliates to the point where it’s like, it’s not useful.

81 00:11:09.840 00:11:14.310 Kevin Bell: So, if we’re gonna come up with some model, I, you know.

82 00:11:15.290 00:11:22.030 Kevin Bell: I think we’ll have to… I think what would be helpful is to compare and contrast, like, what is the difference in conversion numbers that we’re seeing

83 00:11:23.280 00:11:26.490 Kevin Bell: You know, currently, versus, like, what the model is going to be showing us.

84 00:11:27.340 00:11:29.389 Kevin Bell: I think that’s gonna be helpful.

85 00:11:34.290 00:11:35.619 Kevin Bell: If that makes sense.

86 00:11:36.900 00:11:43.529 Robert Tseng: Yeah, well, I mean, like, that’s kind of how we detected the, the offer situation, right? It’s like.

87 00:11:44.140 00:11:49.940 Robert Tseng: they were… But if we’re counting X number of conversions, and then…

88 00:11:50.630 00:11:56.789 Robert Tseng: I mean, basically, Cutter had me just run the analysis of, like, well, how many of these are actually conversions, and…

89 00:11:57.090 00:12:02.469 Robert Tseng: I don’t think… I mean, sure, the number is… the dollar amount is what it is. I mean.

90 00:12:03.000 00:12:04.850 Robert Tseng: All of those are valid.

91 00:12:05.090 00:12:13.710 Robert Tseng: valid conversions by our definition, it’s not necessarily, like, what they should be getting credit for, it’s probably…

92 00:12:14.330 00:12:20.909 Robert Tseng: Yeah, they’re probably getting credit for more… more than what I’m saying when I’m looking at internal data, but, like, that’s…

93 00:12:21.750 00:12:25.050 Robert Tseng: Yeah, like, I… I… how do we, like…

94 00:12:25.720 00:12:33.370 Robert Tseng: Like, I… I could… we could… we could easily model that, but it’s… it’s kind of like deciding, well, what are they on… what are they actually on the hook for?

95 00:12:37.050 00:12:41.600 Kevin Bell: Yeah, if we… If we model out the number of,

96 00:12:43.140 00:12:48.470 Kevin Bell: If we model out and say, alright, these are, you know, exactly conversions, or, you know, yeah, exact…

97 00:12:48.730 00:12:50.749 Kevin Bell: exact conversion numbers.

98 00:12:50.990 00:12:53.419 Kevin Bell: Then we would be paying based upon that number.

99 00:12:53.650 00:12:54.580 Kevin Bell: So…

100 00:13:00.010 00:13:07.040 Kevin Bell: Yeah, I guess my concern is if we’re modeling out how much

101 00:13:07.760 00:13:18.329 Kevin Bell: are we sure that these modeled conversions are going to be, like, 100% accurate? Because I think, based upon the numbers that we have now, I mean, we’re seeing, like, let’s say, for example, like, a 4% conversion rate.

102 00:13:18.480 00:13:22.299 Kevin Bell: you know, based upon this data that I was… that’s obviously not 100% accurate.

103 00:13:22.620 00:13:24.590 Kevin Bell: That would drop us…

104 00:13:24.880 00:13:31.779 Kevin Bell: You know, that could potentially drop our conversion rate quite a bit, because we’re now actually telling them exactly the number of customers.

105 00:13:32.070 00:13:34.489 Kevin Bell: Which is obviously key, we don’t want to be overpaying.

106 00:13:35.750 00:13:40.780 Kevin Bell: But then it comes down to, like, alright, we have to spend a lot of time investing in improving the conversion rate.

107 00:13:40.920 00:13:48.969 Kevin Bell: And that also drops our rankings. So, I guess my… My concern is…

108 00:13:49.150 00:13:57.190 Kevin Bell: how much of a drop-off are we going to experience from this modeled approach? Like, if it’s…

109 00:13:59.850 00:14:05.580 Kevin Bell: We want to get as accurate as possible, but if it’s, like, a 5-10% you know.

110 00:14:06.300 00:14:13.520 Kevin Bell: if the accuracy’s, like, 5-10% off, I think… I think that would be fine, but if we’re, like, 50% off, then yeah, we need to do something, you know, a much more modern approach.

111 00:14:14.760 00:14:26.049 Robert Tseng: Okay, well, I mean, I guess for now, we’ll just have to deploy this solution, and then we’ll monitor it over, like, I don’t know, like, two to four weeks, and then we’ll be able to see, like.

112 00:14:26.200 00:14:28.720 Robert Tseng: Again, we know what the actual conversions are.

113 00:14:28.720 00:14:29.620 Kevin Bell: Versus, like.

114 00:14:29.790 00:14:32.689 Robert Tseng: You know, and so we’ll be able to make that

115 00:14:32.850 00:14:36.400 Robert Tseng: I mean, I wouldn’t be able to tell you what the difference is right now.

116 00:14:36.400 00:14:37.740 Henry Zhao: Yeah, that sounds good to me as well.

117 00:14:38.830 00:14:39.450 Robert Tseng: Okay.

118 00:14:41.200 00:14:42.849 Kevin Bell: Alright, cool, cool.

119 00:14:43.560 00:14:44.460 Kevin Bell: Makes sense.

120 00:14:45.300 00:14:47.240 Henry Zhao: Alright, yeah, if you can add me to that channel, that’d be great.

121 00:14:48.400 00:14:51.499 Kevin Bell: Will do, I’ll add you to that in a minute, Henry.

122 00:14:51.500 00:14:58.550 Henry Zhao: Cool, and then we’ll get… we’ll get started, Robert, with Zoran’s solution, and then work on some stitching, and then we should be able to have some preliminary data.

123 00:14:59.080 00:15:01.369 Henry Zhao: And see the uptick also.

124 00:15:04.960 00:15:05.910 Kevin Bell: Alright.

125 00:15:05.910 00:15:07.130 Henry Zhao: Okay, thank you guys.

126 00:15:07.130 00:15:08.349 Kevin Bell: Appreciate it, fellas.

127 00:15:09.030 00:15:09.750 Henry Zhao: Take care.

128 00:15:10.610 00:15:11.140 Robert Tseng: Right.