Meeting Title: Zoran - Hannah - EMH Case Study Date: 2025-10-30 Meeting participants: Zoran Selinger, Hannah Wang
WEBVTT
1 00:00:37.910 ⇒ 00:00:38.930 Hannah Wang: A…
2 00:00:38.930 ⇒ 00:00:39.960 Zoran Selinger: Hi, Hannah.
3 00:00:41.440 ⇒ 00:00:42.490 Hannah Wang: How’s it going?
4 00:00:42.850 ⇒ 00:00:45.080 Zoran Selinger: Yeah, yeah, good, good, how are you doing?
5 00:00:45.820 ⇒ 00:00:53.710 Hannah Wang: I’m doing… okay, let me… I was in a weird position, let me get settled.
6 00:00:55.860 ⇒ 00:01:02.710 Hannah Wang: That was nice, nice meeting you, Okay, so…
7 00:01:02.900 ⇒ 00:01:06.350 Hannah Wang: Usually for this case study type of
8 00:01:07.460 ⇒ 00:01:23.569 Hannah Wang: interview. I just ask you a bunch of questions, and then I basically pull everything into… or everything’s fed into the transcript, and I take that transcript, and I make the… and I make the case study. So, if you feel like you’re being redundant.
9 00:01:23.900 ⇒ 00:01:28.870 Hannah Wang: That’s okay, because I want all the information to be recorded.
10 00:01:30.650 ⇒ 00:01:37.579 Hannah Wang: And, yeah, it’s just gonna be a list of questions that I ask you, so let me pull that up. One second…
11 00:01:37.700 ⇒ 00:01:47.319 Hannah Wang: Mmm… Okay, so I have it pulled up, and for this case study, I want to talk about
12 00:01:47.540 ⇒ 00:01:53.539 Hannah Wang: LA Mental Health, there was something specific.
13 00:01:54.020 ⇒ 00:02:01.860 Hannah Wang: That, Robert… Wanted, let me… Let me see…
14 00:02:06.830 ⇒ 00:02:11.260 Hannah Wang: Okay, yeah, specifically for attribution,
15 00:02:12.160 ⇒ 00:02:21.240 Hannah Wang: Not entirely sure what that means, but hopefully you know what work to talk about in terms of, like, focusing on attribution.
16 00:02:21.580 ⇒ 00:02:24.910 Hannah Wang: So yeah, do you have any questions for me before we start?
17 00:02:26.480 ⇒ 00:02:33.350 Zoran Selinger: Mom… No, not… not really. So that’s gonna… that’s gonna end up, being a… a blog post.
18 00:02:33.630 ⇒ 00:02:38.149 Zoran Selinger: On the website somewhere. Will you run the final draft by me?
19 00:02:38.150 ⇒ 00:02:38.610 Hannah Wang: Yes.
20 00:02:38.610 ⇒ 00:02:39.820 Zoran Selinger: Okay, good, good.
21 00:02:39.820 ⇒ 00:02:53.570 Hannah Wang: Yeah, it’ll be a case study, it’s just gonna be one PDF page. Eventually, we do want to put it on the blog page on our website, but for now, it’ll just be a PDF, and I’ll ask you for your feedback afterwards.
22 00:02:53.570 ⇒ 00:02:54.830 Zoran Selinger: Cool. Cool. Okay.
23 00:02:54.830 ⇒ 00:02:55.440 Hannah Wang: Okay.
24 00:02:55.580 ⇒ 00:02:58.850 Hannah Wang: So, just broadly speaking,
25 00:02:58.980 ⇒ 00:03:09.350 Hannah Wang: like, what type of project was this? Like, I don’t know all the specific technical terms, but yeah, what type of project was it? High level.
26 00:03:09.350 ⇒ 00:03:17.440 Zoran Selinger: Yeah, so we… we… we started, working for them, you know, just doing… trying to do,
27 00:03:17.560 ⇒ 00:03:37.159 Zoran Selinger: a typical set of services that we do. And then on Discovery, we, they kind of explain that the priority right now for them is for all of their franchisees to have some better visibility on… on the performance of their Google Ads.
28 00:03:38.320 ⇒ 00:03:48.399 Zoran Selinger: So… they were… their main analytics tool that they have is called Amplitude, so they’re on Amplitude.
29 00:03:48.640 ⇒ 00:03:53.929 Zoran Selinger: And Amplitude has a native integration with Google Ads.
30 00:03:55.090 ⇒ 00:04:00.549 Zoran Selinger: Now, they have hundreds of franchisees, right? And there’s…
31 00:04:00.690 ⇒ 00:04:05.240 Zoran Selinger: 10 to 15 key events that they… that they track.
32 00:04:05.990 ⇒ 00:04:16.870 Zoran Selinger: They kind of anonymized it so it’s compliant. You don’t really know what happened, it’s just the event names are literally the names of the famous psychologists.
33 00:04:17.230 ⇒ 00:04:17.940 Hannah Wang: Okay.
34 00:04:17.940 ⇒ 00:04:19.819 Zoran Selinger: They know on the backend what that…
35 00:04:20.050 ⇒ 00:04:25.230 Hannah Wang: event represents, right? It’s just to be compliant with HIPAA and everything else. Yeah.
36 00:04:25.230 ⇒ 00:04:33.680 Zoran Selinger: So, they wanted… they attempted to use that native integration for a few accounts that they have.
37 00:04:33.860 ⇒ 00:04:39.919 Zoran Selinger: And, just couldn’t get it, couldn’t get it working.
38 00:04:40.120 ⇒ 00:04:46.100 Zoran Selinger: So, our goal there was to… to…
39 00:04:46.230 ⇒ 00:04:52.279 Zoran Selinger: Figure out why, why this is not working. And, you know, set it up.
40 00:04:52.460 ⇒ 00:04:59.880 Zoran Selinger: Get them ready to, you know, plug in a new franchisee whenever they want to.
41 00:05:01.240 ⇒ 00:05:18.989 Zoran Selinger: We went in, we looked at those native integrations, and ended up concluding that the native integration is only for a single sub-account, basically for one franchisee, and one event only.
42 00:05:19.580 ⇒ 00:05:20.050 Hannah Wang: Okay.
43 00:05:20.050 ⇒ 00:05:25.169 Zoran Selinger: So you can imagine having hundreds of franchisees, and…
44 00:05:25.460 ⇒ 00:05:27.920 Zoran Selinger: For each of them to at least have
45 00:05:28.420 ⇒ 00:05:43.790 Zoran Selinger: be interested in 3, 4, 5, 10 key events that I want to have in their Google Ads accounts is conversion actions. So, that could… we could end up creating, you know, hundreds and thousands of connections.
46 00:05:44.330 ⇒ 00:05:56.039 Zoran Selinger: in the interface, which is… obviously, you can… that’s unmanageable completely, right? We hope… we hope that, just pulling
47 00:05:56.200 ⇒ 00:06:10.479 Zoran Selinger: pulling the subaccounts into their MCC, which is… MCC is an overarching account in Google Ads. You can have sub-accounts, and you’re automatically the, kind of, the manager of
48 00:06:10.630 ⇒ 00:06:13.840 Zoran Selinger: Of all the… all… of all the accounts.
49 00:06:14.360 ⇒ 00:06:15.630 Zoran Selinger: Oh, so…
50 00:06:16.410 ⇒ 00:06:27.210 Zoran Selinger: we hope that we could use that native connection to just do it, like, out of the box. We set it once for everyone, and whoever comes into the MCC account.
51 00:06:27.210 ⇒ 00:06:36.789 Zoran Selinger: has this available, but it’s not the case. So it’s really, like, the permutations of number of accounts times the number of key events.
52 00:06:38.470 ⇒ 00:06:44.460 Zoran Selinger: We decided, let’s… let’s build a custom solution.
53 00:06:44.460 ⇒ 00:06:45.110 Hannah Wang: Okay.
54 00:06:45.110 ⇒ 00:06:51.660 Zoran Selinger: Let’s build a custom solution for this, because it can still be a pretty simple solution.
55 00:06:51.680 ⇒ 00:07:09.880 Zoran Selinger: Like, when you look at, when you look at all the, all the moving pieces in the, in the cloud solution, it’s very few pieces that… that’s there. But the interface itself is a simple Google form and a Google Sheet.
56 00:07:09.900 ⇒ 00:07:12.759 Zoran Selinger: And everything can be managed… managed there.
57 00:07:12.970 ⇒ 00:07:22.829 Zoran Selinger: The setup itself, the form that they use to set up one account times one account.
58 00:07:23.000 ⇒ 00:07:29.129 Zoran Selinger: One conversion action, which is one of the key events, is identical to amplitude.
59 00:07:29.130 ⇒ 00:07:30.449 Hannah Wang: Right? But…
60 00:07:30.490 ⇒ 00:07:37.940 Zoran Selinger: That’s the… the only… they manage it in one place, it’s just one spreadsheet.
61 00:07:38.270 ⇒ 00:07:44.629 Zoran Selinger: They just need to submit a form submission with literally 3 fields.
62 00:07:45.010 ⇒ 00:07:51.750 Zoran Selinger: And that’s it for each one of those cases. What happens is,
63 00:07:51.870 ⇒ 00:07:56.630 Zoran Selinger: So the solution, basically, is built on Google Cloud Platform.
64 00:07:56.760 ⇒ 00:07:57.750 Zoran Selinger: we have…
65 00:07:58.070 ⇒ 00:08:07.110 Zoran Selinger: two, two Cloud Functions, we have a BigQuery table, and we have one PubSub topic between those two Cloud Functions.
66 00:08:09.230 ⇒ 00:08:16.180 Zoran Selinger: We have a… we have a piece that sits in Amplitude, which is a webhook connection that sends events
67 00:08:16.300 ⇒ 00:08:22.350 Zoran Selinger: to the first Cloud Functions, which is… which has an endpoint, and it’s a controller that can receive those requests.
68 00:08:22.860 ⇒ 00:08:24.760 Zoran Selinger: We get the click ID,
69 00:08:24.840 ⇒ 00:08:49.140 Zoran Selinger: we… we… we look up the settings. Okay, so this is the… one of the key events. Do we have settings for those… that key events… those key events in our BigQuery table, which is essentially a mirror of the Google Sheet? Google Sheet is just, you know, the interface. When you make changes to the Google Sheet, those same changes happen in BigQuery as well.
70 00:08:49.140 ⇒ 00:08:50.890 Zoran Selinger: and we read the BigQuery.
71 00:08:50.890 ⇒ 00:08:51.730 Zoran Selinger: Stable.
72 00:08:52.540 ⇒ 00:08:58.939 Zoran Selinger: So, if that exists, simply a message is sent to a PubSub topic.
73 00:08:59.410 ⇒ 00:09:02.419 Zoran Selinger: The worker that is going to upload
74 00:09:02.530 ⇒ 00:09:08.610 Zoran Selinger: the actual conversion into the Google Ads service using the API picks that message up.
75 00:09:08.750 ⇒ 00:09:11.010 Zoran Selinger: Processes it, and that’s it.
76 00:09:11.810 ⇒ 00:09:14.609 Zoran Selinger: It’s… it’s a very, very simple system.
77 00:09:14.970 ⇒ 00:09:22.980 Zoran Selinger: So, basically, what, what is, what’s happening,
78 00:09:23.830 ⇒ 00:09:43.490 Zoran Selinger: inside the amplitude, instead of having, you know, thousands of small connections in the screen, you… you… your interface in the amplitude is one connection, which is the webhook that sends data to… to our… our system, and you have that
79 00:09:43.580 ⇒ 00:09:50.909 Zoran Selinger: you have that really simple Google form, where, okay, we plug in a new, new franchisee.
80 00:09:51.680 ⇒ 00:10:05.169 Zoran Selinger: we invite the account to be under our MCC, we go in, we create the conversion actions inside the account, get the conversion action IDs, and we submit those into the Google Form, and it’s…
81 00:10:05.510 ⇒ 00:10:11.820 Zoran Selinger: Immediately, in place, and working from From that moment onwards.
82 00:10:12.420 ⇒ 00:10:12.910 Hannah Wang: Okay.
83 00:10:12.910 ⇒ 00:10:14.140 Zoran Selinger: Yeah, so…
84 00:10:14.560 ⇒ 00:10:18.270 Hannah Wang: I have a couple clarifying questions.
85 00:10:19.950 ⇒ 00:10:20.850 Hannah Wang: I guess…
86 00:10:21.310 ⇒ 00:10:28.659 Hannah Wang: I should have looked into the client before this, but LA Mental Health, like, what do they do exactly? I’m assuming Nate with Mental Health.
87 00:10:28.660 ⇒ 00:10:45.340 Zoran Selinger: Yeah, they, they gen… generally connect, mental health professionals with, with, with their patients. I see. So they have, they have clinics, they have online, they, it’s essentially, a huge, huge,
88 00:10:45.710 ⇒ 00:10:53.320 Zoran Selinger: company that… that, you can… any health… mental health professionals can sign up.
89 00:10:53.520 ⇒ 00:10:59.619 Zoran Selinger: As far as I understand, they can sign up and get certified, and then they get matched with patients.
90 00:10:59.620 ⇒ 00:11:00.230 Hannah Wang: Okay.
91 00:11:00.230 ⇒ 00:11:00.670 Zoran Selinger: True.
92 00:11:00.670 ⇒ 00:11:03.869 Hannah Wang: And then, what is a franchisee?
93 00:11:03.870 ⇒ 00:11:08.720 Zoran Selinger: A franchisee is, so basically those, those clinics…
94 00:11:09.270 ⇒ 00:11:17.410 Zoran Selinger: And those mental health professionals can run their own campaigns to generate, generate patients.
95 00:11:17.650 ⇒ 00:11:19.629 Hannah Wang: So that, okay, I see.
96 00:11:19.630 ⇒ 00:11:23.990 Zoran Selinger: So they, they are all, they, they basically… those specific…
97 00:11:24.320 ⇒ 00:11:33.599 Zoran Selinger: clinics that are under… under their franchise. They can use their platform, but they… they… they run the,
98 00:11:33.770 ⇒ 00:11:36.759 Zoran Selinger: The marketing… marketing themselves.
99 00:11:36.870 ⇒ 00:11:39.300 Zoran Selinger: So they have… they have their own…
100 00:11:40.150 ⇒ 00:11:44.569 Zoran Selinger: Google Ads accounts, and they run ads for their location, right?
101 00:11:45.040 ⇒ 00:11:49.120 Hannah Wang: I see, so Ellie Mental Health keeps… track…
102 00:11:49.120 ⇒ 00:12:06.749 Zoran Selinger: It’s a platform, yeah. It’s a platform that all of those franchisees use to, you know, for their patients to set appointments with them on their particular location or online, and they generate
103 00:12:06.880 ⇒ 00:12:08.640 Zoran Selinger: the traffic themselves.
104 00:12:08.830 ⇒ 00:12:27.499 Hannah Wang: I see. So it’s like, yeah, Elements Health is just a platform to, I guess, market. Or it’s just to host, like, marketing and… Yeah. Okay, got it. Okay, so I know you, like, gave a good overview and a deep dive into the process, so I’m just gonna kind of break it down and ask more specific questions, so…
105 00:12:28.070 ⇒ 00:12:47.069 Hannah Wang: I guess I want to understand, like, the context of everything that was happening before we came in to help Ellie Mental Health. So, what… specifically… specifically for the attribution part of it all, like, what was the working environment like before the project started? I know you mentioned…
106 00:12:47.070 ⇒ 00:12:52.340 Zoran Selinger: So, I wasn’t there at the start of the project. So that’s a question for Robert.
107 00:12:52.520 ⇒ 00:12:52.930 Hannah Wang: Okay.
108 00:12:53.140 ⇒ 00:13:02.980 Zoran Selinger: So I just… I just don’t know. I particularly came in, my first, my first task was to, to…
109 00:13:03.700 ⇒ 00:13:14.479 Zoran Selinger: audit their amplitude and Google Analytics setup. That was my first ask. Then we decided that GA4 is a no-go.
110 00:13:14.790 ⇒ 00:13:17.070 Zoran Selinger: Because compliance and…
111 00:13:17.070 ⇒ 00:13:17.500 Hannah Wang: boost.
112 00:13:17.500 ⇒ 00:13:28.279 Zoran Selinger: has problems with it, but mainly because of compliance. GA4 is a no-go, and… which left their Google Ads
113 00:13:29.030 ⇒ 00:13:29.960 Zoran Selinger: Kinda…
114 00:13:31.690 ⇒ 00:13:41.279 Zoran Selinger: because G… I didn’t explain this, G… Google Analytics 4 was an alternative for tracking conversions for every franchisee.
115 00:13:41.280 ⇒ 00:13:41.670 Hannah Wang: Okay.
116 00:13:41.670 ⇒ 00:13:58.040 Zoran Selinger: But just, we couldn’t… we couldn’t use it. This is why we have to… had to rely directly, connect Amplitude to Google Ads. It had to be a direct connection. And then we decided, okay, that’s an absolute priority right now, and then for the last, you know, month.
117 00:13:58.040 ⇒ 00:14:06.240 Zoran Selinger: That was the only work that we’ve done for them, but now, we’re gonna… I think we’re gonna continue.
118 00:14:07.300 ⇒ 00:14:08.540 Zoran Selinger: adding some…
119 00:14:08.620 ⇒ 00:14:10.030 Hannah Wang: Nope, services.
120 00:14:10.330 ⇒ 00:14:24.349 Hannah Wang: Okay. Yeah, so… but the whole… the beginning and onboarding, I wasn’t there for it, so… Okay, so you wouldn’t happen to know, like, the challenges Ellie was facing in terms of, like, why we…
121 00:14:25.180 ⇒ 00:14:32.260 Hannah Wang: like, what issue were they trying to solve, I guess? And, like, why were they needing this help of, like.
122 00:14:32.260 ⇒ 00:14:41.389 Zoran Selinger: Yeah, so what I know is they use, they use, they use Amplitude, they use Google Analytics,
123 00:14:41.400 ⇒ 00:14:56.530 Zoran Selinger: they see those two systems and being… it being completely, like, they are not reliable, their… their… their settings are… their setup implementation is incomplete, especially Google Analytics, so they just…
124 00:14:56.760 ⇒ 00:15:01.250 Zoran Selinger: They just know they’re kind of… They’re running blind.
125 00:15:01.460 ⇒ 00:15:05.369 Zoran Selinger: They don’t really have reliable, reliable analytics.
126 00:15:05.550 ⇒ 00:15:10.220 Hannah Wang: So that’s probably the main purpose, from what I could gather.
127 00:15:10.220 ⇒ 00:15:16.070 Zoran Selinger: This is why they got in touch, to kind of figure out, okay, what do we need to finally do?
128 00:15:16.330 ⇒ 00:15:22.890 Zoran Selinger: To have a proper Proper analytics in place, so we can make informed decisions.
129 00:15:22.890 ⇒ 00:15:23.950 Hannah Wang: Hmm.
130 00:15:23.950 ⇒ 00:15:30.830 Zoran Selinger: So I don’t know if what I could see from Amplitude is something that we’ve done before I joined.
131 00:15:31.210 ⇒ 00:15:32.720 Hannah Wang: I see.
132 00:15:32.950 ⇒ 00:15:40.730 Zoran Selinger: Or… or… Or it was… or we never touched it, I don’t know. I’m not sure. Okay.
133 00:15:40.730 ⇒ 00:15:45.760 Hannah Wang: No worries. But the GA4 is basically completely unusable for them.
134 00:15:45.890 ⇒ 00:16:00.240 Zoran Selinger: And they were looking to us for that final confirmation that that is the case. They weren’t sure that that’s really the case, and I went in and I…
135 00:16:00.420 ⇒ 00:16:09.079 Zoran Selinger: they were basically… they were sending… there’s a way to send data without your website being tagged. There’s an API where you could send the data in.
136 00:16:09.410 ⇒ 00:16:15.220 Zoran Selinger: But the sessions teaching and users teaching, and everything else really depends on the pixel.
137 00:16:16.140 ⇒ 00:16:19.830 Zoran Selinger: And they did not wanna… they don’t wanna have the big saline.
138 00:16:20.530 ⇒ 00:16:35.569 Zoran Selinger: So, the data that they were sending in, completely unusable, so… and they were looking to us to explain that, and finally give them the confirmation that that is the case, and we have to look for alternative solution.
139 00:16:36.030 ⇒ 00:16:36.570 Hannah Wang: Okay.
140 00:16:36.570 ⇒ 00:16:44.420 Zoran Selinger: And I guess there’s… they are decided on amplitude, and looks to me like that can work really well for them.
141 00:16:44.600 ⇒ 00:16:45.000 Hannah Wang: just…
142 00:16:45.000 ⇒ 00:16:52.229 Zoran Selinger: need to configure the optimization, the integrations that they need, that’s it. And this being one of them.
143 00:16:53.390 ⇒ 00:16:59.759 Hannah Wang: So is this project… did we… or how long did this project take? I’m assuming we… we’re wrapping it up right now?
144 00:16:59.760 ⇒ 00:17:11.499 Zoran Selinger: We are wrapping it up, yes. We… I mean, the development is done, but just for the last week, we are just simply monitoring how it’s working. There is a next phase of this.
145 00:17:11.760 ⇒ 00:17:21.289 Zoran Selinger: Which is being talked about right now. They do eventually, so the reason why we did it on GCP is because they want it right now.
146 00:17:22.569 ⇒ 00:17:28.709 Zoran Selinger: But we have to rewrite this, we have to rebuild this in Azure.
147 00:17:29.080 ⇒ 00:17:29.710 Hannah Wang: Okay.
148 00:17:29.710 ⇒ 00:17:34.709 Zoran Selinger: Because they… because of compliance and everything else, they want to have everything in Azure.
149 00:17:34.970 ⇒ 00:17:35.540 Hannah Wang: Okay.
150 00:17:35.540 ⇒ 00:17:41.470 Zoran Selinger: So, right now, we are… we have the conversations about whether we are going…
151 00:17:41.640 ⇒ 00:17:48.919 Zoran Selinger: excuse me, whether we are going to do it, or they have… internally have the resources to do it. Looks like they don’t.
152 00:17:49.580 ⇒ 00:17:53.460 Zoran Selinger: I think the main Azure person left.
153 00:17:53.730 ⇒ 00:18:05.890 Zoran Selinger: And the replacement is, just swamped at the moment, so I’m not sure, how that’s gonna go. Sounds to me like we’re going to do it, but let’s see.
154 00:18:06.230 ⇒ 00:18:06.630 Hannah Wang: Okay.
155 00:18:06.630 ⇒ 00:18:07.960 Zoran Selinger: No conclusion yet.
156 00:18:08.220 ⇒ 00:18:15.569 Hannah Wang: Okay, and how… when did you start this process, or this phase of the project, I guess?
157 00:18:15.570 ⇒ 00:18:19.680 Zoran Selinger: It’s… I think start of the month was… yeah, yeah, yeah.
158 00:18:19.680 ⇒ 00:18:20.300 Hannah Wang: Okay.
159 00:18:20.850 ⇒ 00:18:23.839 Hannah Wang: I’m gonna ask you to go…
160 00:18:23.970 ⇒ 00:18:37.000 Hannah Wang: or I’m just gonna ask more questions about the solution itself. So I know you gave, like, an overview. Would it be possible to share your screen with me and kind of walk through it? It doesn’t matter, like, I’m not gonna screenshot any of.
161 00:18:37.000 ⇒ 00:18:37.730 Zoran Selinger: That’s true.
162 00:18:37.730 ⇒ 00:18:43.549 Hannah Wang: If it’s… if there’s compliance stuff, but I also just want a visual, so I understand everything.
163 00:18:43.570 ⇒ 00:18:46.469 Zoran Selinger: Of course, let me do this, share.
164 00:18:54.010 ⇒ 00:18:56.680 Zoran Selinger: Should I do… yeah, Windows should be fine.
165 00:18:58.370 ⇒ 00:19:04.210 Zoran Selinger: So this is a… This is a, the dashboard?
166 00:19:04.290 ⇒ 00:19:23.980 Zoran Selinger: Yep. So sometimes we do get… so, just two things, unattributable clicks, which are… which are basically, sometimes we get events with clicks that do not belong to the accounts we set up already in the system, so those are unattributable clicks, and we have successful runs of our two cloud functions, that…
167 00:19:23.980 ⇒ 00:19:26.219 Zoran Selinger: That, we run.
168 00:19:26.460 ⇒ 00:19:41.400 Zoran Selinger: So that’s a very simple dashboard, because the whole system is very, very simple. Let me, so the settings, and, settings is… are very simple, you see? We just…
169 00:19:42.020 ⇒ 00:19:47.140 Zoran Selinger: account ID, conversion action ID, and the code name of the key event.
170 00:19:47.140 ⇒ 00:19:47.570 Hannah Wang: Hmm.
171 00:19:47.570 ⇒ 00:19:52.759 Zoran Selinger: That they have, and so when you pull that in, that is,
172 00:19:53.040 ⇒ 00:19:56.909 Zoran Selinger: That’s gonna work. Let me just show you the form.
173 00:20:10.830 ⇒ 00:20:14.149 Zoran Selinger: So this is a simple… this is a very simple form.
174 00:20:14.350 ⇒ 00:20:15.369 Zoran Selinger: You pull the…
175 00:20:15.530 ⇒ 00:20:22.080 Zoran Selinger: customer ID, which is the Google Ads account ID. You have the ID of your conversion action.
176 00:20:22.180 ⇒ 00:20:26.390 Zoran Selinger: And there’s a… there’s a checkbox for…
177 00:20:26.610 ⇒ 00:20:36.690 Zoran Selinger: For the different events you want to enable, and there’s also an other. These are the events that are inside the amplitude, so they have to match.
178 00:20:37.180 ⇒ 00:20:46.079 Zoran Selinger: Completely. So when the event called Marshall Linehunt comes in with a Google Click ID, Then we propagate it.
179 00:20:46.810 ⇒ 00:20:52.499 Zoran Selinger: If, obviously, if we find a setting for Marshalline in the table already.
180 00:20:52.830 ⇒ 00:20:54.449 Zoran Selinger: Right? So…
181 00:20:55.080 ⇒ 00:21:09.419 Zoran Selinger: if the event called Marshalline comes in, I… we… we look up all the rows that need it, we… and then we, you know, look up the… the Google Click ID and propagate to the…
182 00:21:10.350 ⇒ 00:21:12.440 Zoran Selinger: To the platform.
183 00:21:13.600 ⇒ 00:21:15.230 Hannah Wang: So this is tracking…
184 00:21:15.230 ⇒ 00:21:16.219 Zoran Selinger: Yes, go ahead.
185 00:21:16.400 ⇒ 00:21:18.750 Hannah Wang: So this is tracking, like, wit…
186 00:21:19.110 ⇒ 00:21:23.370 Hannah Wang: Which, like, when users click on a specific franchise.
187 00:21:23.370 ⇒ 00:21:33.199 Zoran Selinger: Either click or submit or, you know, different actions. Let me just give you… give you this view as well. So…
188 00:21:33.610 ⇒ 00:21:36.900 Zoran Selinger: Where that ends up, is here.
189 00:21:37.120 ⇒ 00:21:39.620 Zoran Selinger: And we just chose Orlando.
190 00:21:41.890 ⇒ 00:21:43.570 Hannah Wang: Oh, cool, okay.
191 00:21:43.570 ⇒ 00:21:45.559 Zoran Selinger: Yeah, so this, we are in Google Ads.
192 00:21:46.100 ⇒ 00:21:52.650 Zoran Selinger: And in Google Ads, there’s a con… there are conversions, conversion actions.
193 00:21:52.780 ⇒ 00:22:01.930 Zoran Selinger: Which is basically, you know, that can be a purchase, email submission, whatever. It’s a conversion event. The way it works in Google, we set it up here.
194 00:22:02.410 ⇒ 00:22:03.959 Zoran Selinger: Let me set it up here.
195 00:22:05.150 ⇒ 00:22:10.769 Zoran Selinger: And then… when there’s Virginia satir… here.
196 00:22:10.940 ⇒ 00:22:18.149 Zoran Selinger: And if… if the event, if the Google Click ID comes in… that was generated by this account.
197 00:22:18.570 ⇒ 00:22:28.370 Zoran Selinger: and this event happens, then we send a conversion for this conversion action. And then, obviously, in reports.
198 00:22:28.640 ⇒ 00:22:46.070 Zoran Selinger: when they have reports, when they have conversions here, one of the columns, they can look at how many Virginia events they had for… from which campaign and all that. That’s the whole point of this. This is the… this is how they know whether the… if…
199 00:22:46.340 ⇒ 00:22:49.610 Zoran Selinger: The campaigns are successful or not.
200 00:22:50.900 ⇒ 00:22:51.270 Hannah Wang: Got it.
201 00:22:51.270 ⇒ 00:22:55.000 Zoran Selinger: That’s the conversion signal, that’s the goal of the campaign.
202 00:22:56.480 ⇒ 00:23:04.410 Hannah Wang: And so we tell each franchise… like, I’m assuming we tell the franchisee, like, oh, this campaign is working, or because Ellie…
203 00:23:04.410 ⇒ 00:23:07.380 Zoran Selinger: No, no, no, so they have… this is their account.
204 00:23:07.630 ⇒ 00:23:11.469 Zoran Selinger: So this is… Account is run by a franchisee.
205 00:23:12.490 ⇒ 00:23:13.360 Hannah Wang: Got it.
206 00:23:13.360 ⇒ 00:23:19.559 Zoran Selinger: what… only thing we do is we set up those conversions that I just showed you. So we go here…
207 00:23:19.870 ⇒ 00:23:25.470 Zoran Selinger: And we set up new roles. David Grant, Virginia, and then we have Archa.
208 00:23:25.620 ⇒ 00:23:35.409 Zoran Selinger: In particular for this account. These are the three actions that they are interested in tracking, and that’s what we set up, yeah. So…
209 00:23:35.410 ⇒ 00:23:36.010 Hannah Wang: Got it.
210 00:23:36.470 ⇒ 00:23:51.390 Zoran Selinger: So those 3 is literally, you see, 372699 is the account ID, and… So this is it, this is it. These are… in particular, this is the row for David Grant in that account.
211 00:23:52.240 ⇒ 00:24:00.670 Zoran Selinger: in Virginia, and we have Marsha as well, yeah. So these three rows are in particular settings for that particular account.
212 00:24:01.320 ⇒ 00:24:01.870 Zoran Selinger: And those.
213 00:24:01.870 ⇒ 00:24:02.330 Hannah Wang: Got it.
214 00:24:02.330 ⇒ 00:24:04.079 Zoran Selinger: Internet conversion actions.
215 00:24:04.630 ⇒ 00:24:08.199 Hannah Wang: I see, and we have each of those for every… every clinic.
216 00:24:08.200 ⇒ 00:24:18.210 Zoran Selinger: Every. They… so, when they agree with the franchisee, that franchisee wants this, they go in and can now plug in that franchisee
217 00:24:19.960 ⇒ 00:24:29.480 Zoran Selinger: themselves. They know how to do it, they create the conversion actions, they go… they go here into the form, they submit
218 00:24:30.690 ⇒ 00:24:41.229 Zoran Selinger: they submit a value for every single event they are interested in, and that’s it. They have… that’s their user interface. They don’t have to code anything, nothing.
219 00:24:42.600 ⇒ 00:24:46.420 Hannah Wang: So, sorry, who’s the one filling out this Google Form again?
220 00:24:46.420 ⇒ 00:24:51.050 Zoran Selinger: So this is… this is, the marketing managers from… from…
221 00:24:51.050 ⇒ 00:24:51.520 Hannah Wang: Okay.
222 00:24:51.520 ⇒ 00:24:52.610 Zoran Selinger: Grammarly, yeah.
223 00:24:52.850 ⇒ 00:24:55.660 Zoran Selinger: So, Alison and Adam.
224 00:24:55.830 ⇒ 00:24:57.990 Zoran Selinger: Are the ones that are doing that.
225 00:24:57.990 ⇒ 00:25:00.730 Hannah Wang: Allison. I’ve seen that name before. Okay.
226 00:25:01.380 ⇒ 00:25:09.019 Hannah Wang: Cool. Okay, and then, in terms of… obviously, like, in a case study, we kind of want, like.
227 00:25:09.390 ⇒ 00:25:14.720 Hannah Wang: metrics, if we have any. Are there any, like, numbers that you can give me inter…
228 00:25:14.720 ⇒ 00:25:28.319 Zoran Selinger: That would be great, but the volume of clicks moving through this system at the moment is very low, because these are very few accounts that are, at the moment, plugged in. They are basically adding
229 00:25:28.550 ⇒ 00:25:36.940 Zoran Selinger: Adding a few daily, but, it’s still… the volume is very low, so when you see this,
230 00:25:37.950 ⇒ 00:25:42.280 Zoran Selinger: this dashboard, the volume here is not huge. Yeah.
231 00:25:42.470 ⇒ 00:25:48.159 Zoran Selinger: Of, kind of, the volume of clicks going through, through the system is not, is not huge.
232 00:25:48.420 ⇒ 00:25:49.120 Hannah Wang: Got it.
233 00:25:49.850 ⇒ 00:26:05.860 Hannah Wang: Is there… it’s okay if we don’t have any, like, numbers or volume, but, any general feedback that we got from Ellie? Do you happen to know? Like, oh, like, was this project, like, did they say it was helpful, or, like…
234 00:26:06.330 ⇒ 00:26:07.210 Hannah Wang: Yeah.
235 00:26:07.210 ⇒ 00:26:18.010 Zoran Selinger: So I think, also, in the latest conversations, Robert was part of them, so you can, you can see, you can hear, and maybe even…
236 00:26:19.170 ⇒ 00:26:23.010 Zoran Selinger: Utam is talking to… to the Azure guy.
237 00:26:23.410 ⇒ 00:26:29.099 Zoran Selinger: So that’s… that’s where the… the latest comments are… are in those conversations. I wasn’t in them.
238 00:26:29.130 ⇒ 00:26:43.420 Zoran Selinger: So maybe you can… you can hear, from them. I did not receive any, any feedback in terms of, this is not working, or… or they, they are using it, they are using it. There was,
239 00:26:43.550 ⇒ 00:26:51.329 Zoran Selinger: I had a portion of the guide, so we created, obviously, a user guide for this. I had old screenshots.
240 00:26:52.040 ⇒ 00:26:58.690 Zoran Selinger: created that a few weeks ago, and then Google changed the user interface.
241 00:26:58.690 ⇒ 00:27:00.530 Hannah Wang: Oh my god, so…
242 00:27:00.530 ⇒ 00:27:08.129 Zoran Selinger: So, basically, a 3-week, 3-4 week, old, guide was out of date.
243 00:27:09.130 ⇒ 00:27:11.450 Hannah Wang: That’s a bummer, oh my god.
244 00:27:11.450 ⇒ 00:27:21.990 Zoran Selinger: Yeah, but fine, we figured it out, we figured out the new interface, and it’s, as per usual, it’s worse than before.
245 00:27:23.650 ⇒ 00:27:34.420 Zoran Selinger: as it gets with Google, so that’s normal for them, so nothing actually new. They make changes often, and usually for the worse.
246 00:27:34.530 ⇒ 00:27:36.950 Zoran Selinger: It is what it is.
247 00:27:37.790 ⇒ 00:27:45.319 Hannah Wang: Okay, cool. So yeah, I know you mentioned, as we wrap up here, like, we’re trying to migrate off of
248 00:27:45.750 ⇒ 00:27:48.450 Hannah Wang: GCP to Azure, is that correct? Or…
249 00:27:48.450 ⇒ 00:28:08.940 Zoran Selinger: Yes, yes, that’s their platform of choice. When we needed to start working on this because they wanted a solution immediately, but their… their internal processes, they were just doing a little bit of switching to… so their switch to Azure is recent.
250 00:28:08.940 ⇒ 00:28:16.530 Zoran Selinger: Okay, so it’s still ongoing, they just transitioned the website to Azure and platform, so…
251 00:28:16.530 ⇒ 00:28:26.659 Zoran Selinger: there’s… this is not gonna be a priority for a while. That’s why they needed, something early, and now it’s a question of, okay.
252 00:28:26.870 ⇒ 00:28:38.960 Zoran Selinger: will they give us enough access so we can do it? Because that’s… that also needs to be very tight, because they’re in the… in this health space, so the compliance is a big, big deal here.
253 00:28:38.960 ⇒ 00:28:39.640 Hannah Wang: Totally.
254 00:28:39.640 ⇒ 00:28:40.250 Zoran Selinger: Yeah.
255 00:28:40.740 ⇒ 00:28:41.440 Hannah Wang: Okay.
256 00:28:41.440 ⇒ 00:28:42.020 Zoran Selinger: Yeah.
257 00:28:42.300 ⇒ 00:28:48.839 Hannah Wang: Cool. Awesome. Well, I appreciate your time, and like I said,
258 00:28:49.070 ⇒ 00:28:55.930 Hannah Wang: we’ll kind of draft up the case study. It’ll probably be in Figma, so if you don’t have access, I’ll give it to you.
259 00:28:55.930 ⇒ 00:28:56.330 Zoran Selinger: Okay.
260 00:28:56.330 ⇒ 00:29:02.629 Hannah Wang: And you can just look at it and add comments. Just confirm that everything is accurate, because
261 00:29:02.840 ⇒ 00:29:07.669 Hannah Wang: I mean, it’s my first time hearing this, so I’ll try to make sure it’s accurate, but I’m not… yeah, I wasn’t.
262 00:29:07.670 ⇒ 00:29:10.200 Zoran Selinger: Or select, we can go back and forth.
263 00:29:10.540 ⇒ 00:29:10.890 Hannah Wang: Yeah.
264 00:29:10.890 ⇒ 00:29:14.499 Zoran Selinger: And there’s… of course, yeah, no problem.
265 00:29:14.900 ⇒ 00:29:21.739 Hannah Wang: Awesome. Okay, we’ll keep in touch over Slack, and it was nice meeting you.
266 00:29:21.740 ⇒ 00:29:22.760 Zoran Selinger: Auto.
267 00:29:22.980 ⇒ 00:29:24.639 Zoran Selinger: Yeah. Excellent. Thanks.
268 00:29:24.820 ⇒ 00:29:25.710 Zoran Selinger: Bye-bye.