Octus - Rosa's Meeting Point

GTM & GA4 Audit Meeting Agenda

Opening

• Thank you for taking the call on Monday.
• This will be a brief meeting to discuss audit progress and key findings
• Goal: Share hypotheses discovered during the GTM and GA4 investigation

Key Findings Summary

1. Implementation Quality Issues - Deduplication: Multiple tags and variables serving similar purposes - Navigation: No folder structure - tags, triggers, and variables scattered without organization - Naming Conventions: Unclear naming makes it difficult to understand purpose at first glance - Documentation: Missing tracking plan documentation and implementation notes 2. Strategic Tracking Gaps - No Data Taxonomy: Lack of structured framework to understand user flow and behavior - Unclear Tracking Objectives: Difficult to determine which metrics align with business goals - Event Hierarchy Confusion: No clear distinction between standard events and key conversion events 3. Technical Implementation Challenges - User Segmentation: Logged-in vs. prospect differentiation incomplete - Lead Qualification: MQL/SQL tracking not properly implemented - Event Classification: Confusion between events and key events (conversions) - Attribution Issues: Medium and source attribution inconsistencies - Redirect Data Loss: Reorg.com → Octus.com redirect stripping referrer data - Cross-Domain Tracking: Multiple staging domains creating tag coverage problems 4. Marketing Attribution Concerns - UTM Parameters: Inconsistent UTM implementation affecting campaign tracking - Cookie Consent Impact: Potential data loss from cookie consent affecting paid traffic attribution - Custom Channel Gaps: AI referrals (ChatGPT, Claude, Perplexity) not categorized - Paid Traffic Underreporting: Limited CPC data from Google Ads and LinkedIn despite active campaigns

Progress Update

We have two types of audit, surface and details, and for the detail audit it will focusing on business goal. So hence, the meeting today is to discuss that a little bit if that is required for this audit.

Completed

• Surface-level configuration audit (Pillar 1)
• Traffic source analysis (Pillar 2)
• Custom channel report recommendations for AI referrals

In Progress

• Deep-dive event tracking analysis (Pillar 3)

Audit Approach: Four Pillar Framework

Pillar 1: Foundation (Technical Configuration)

• Data maturity assessment
• GTM and GA4 configuration review
• Identifying setup discrepancies

Pillar 2: Who (Traffic & Users)

• Understanding website users and their origins
• Traffic source analysis
• Channel mix evaluation

Key Finding: Identified traffic from ChatGPT, Claude, and Perplexity that should be assigned to custom channel reports. Also seeing limited CPC traffic from Google Ads and LinkedIn Paid Social.

Pillar 3: What (User Journey & Events)

• Key journey tracking
• Event tracking analysis
• Funnel leak identification

Pillar 4: Implementation (Technical Execution & Documentation)

Current Issues Identified:

• Naming conventions unclear (not user-friendly for first-time review)
• No folder organization (tags, triggers, variables not grouped)
• Lack of documentation (not following software development standards)
• Duplicated variables and tags due to poor organization

Key Hypotheses & Questions

Hypothesis 1: Reorg.com → Octus.com Redirect Tracking

Observation: Found inactive implementation attempting to track page referral URLs from reorg.com to octus.com

Questions:

• Is tracking traffic from reorg.com to octus.com still a priority?
• Are you open to alternative redirect methods that may add minimal loading time but improve data accuracy?

Hypothesis 2: User Journey Funnel Analysis

Observation: LinkedIn and Google Ads reports created in GA4 suggest interest in user journey tracking

Questions:

• Do you want to track the complete user journey (landing → trial → demo)?
• Are you trying to identify drop-off points in the funnel?
• Has this funnel visualization been achieved, or is this a desired goal?

If desired: I can develop a tracking plan focusing on critical key events and recommend which current events to consolidate or remove.

Hypothesis 3: Dual GA4 Properties

Observation: Two GA4 properties detected:

1. Reorg Marketing GTM (previous website)
2. Octus Marketing (current, but receiving some data from Reorg Marketing)

Questions:

• Is the plan to migrate all tracking to Octus Marketing over 12-18 months?
• Can we consolidate to one GA4 property using two Google Tags in one GTM container?

Hypothesis 4: Internal & Cross-Domain Tracking Issues

Observation:

• Multiple staging domains (staging.octus.com, app.reorg.com, etc.)
• Tag coverage problems across domains
• Current IP-based filtering not foolproof

Recommendation:

• Create custom channel grouping for internal traffic
• Implement UTM parameter policy (utm_internal=true) for employee bookmarks
• Filter internal traffic at the channel level rather than relying on IP filtering

Hypothesis 5: Custom Dimensions & Parameters Strategy

User Type Tracking: Prospect vs. Customer

Current Setup: Custom dimension user_type with values:

• prospect (default)
• customer (triggered by login click)

Questions:

• Is differentiating logged-in users from prospects critical to your analytics strategy?

Value Proposition - How This Helps Analytics:

• Behavioral Segmentation: Compare how customers vs. prospects navigate the site (e.g., customers may go directly to product features, prospects browse marketing pages)
• Content Effectiveness: Measure which content converts prospects vs. retains customers
• Funnel Analysis: Separate conversion funnels - prospect-to-trial vs. customer-to-upgrade
• Personalization Insights: Identify what logged-in customers are searching for or struggling with
• Churn Indicators: Track when customers stop logging in or change behavior patterns
• ROI Calculation: Differentiate acquisition costs (prospects) from retention costs (customers)

Underutilized Custom Dimensions Identified

Observation: Several valuable custom dimensions exist but aren’t properly implemented:

• referrer_page_url - Could mitigate redirect data loss issues
• login_status - Reinforces prospect vs. customer differentiation
• form_name - Critical for form submission tracking and conversion attribution

Questions:

• Were these custom dimensions intentionally built for specific use cases?
• Is the underutilization due to lack of documentation or incomplete implementation?
• Should we prioritize activating these dimensions properly?

Hypothesis 6: Redirect Technical Challenges

Reorg.com → Autis.com Redirect Impact

Critical Technical Question: Is the 301 redirect stripping referrer data?

Investigation Needed:

• Current redirect behavior analysis
• Referrer header preservation check
• Server-side handling capabilities

Questions:

• Do we have server-side access to implement redirect modifications?
• Is the redirect a hard 301, or can we implement alternative redirect methods?
• Are there development resources available to test redirect scenarios?

Potential Solutions (if server access available):

• Meta refresh with delay to preserve referrer
• JavaScript redirect with manual referrer passing
• Server-side redirect with referrer header forwarding
• Cookie-based session bridging

Trade-off: May add 50-200ms loading time but significantly improve data accuracy.

Next Steps & Access Requirements

Current Deliverable

• Audit report with quick fixes available now
• Can be shared with developers and marketers immediately

For Deeper Analysis

To conduct comprehensive deep-dive auditing (Pillar 3), I’ll need access to:

1. Website access - To test forms and user flows
2. Google Ads account - To verify conversion pixels
3. LinkedIn Ads account - To verify conversion pixels
4. Testing email - For demo and contact form testing (SQL/Non-SQL validation)

Question: Do you prefer a general audit or in-depth testing including form submissions and conversion tracking?

Discussion

• Review each hypothesis
• Address questions and priorities
• Determine scope for next phase
• Identify quick wins from current audit findings

Reference

o, alright, so in this particular video, I’m just specifically want to collect as much points so we can pretty much have it, we can pretty much use these points in our conversation with Rosa next. So what exactly we want to talk about? So number one, the first point that we want… When we have a meeting with her after this, the first thing we want to mention is we want to thank her for taking the call because it’s Monday and it’s morning. So I believe she probably have much better call. So thank you so much for taking the call. So this meeting is going to be a short one. I just want to talk about the process that we currently ongoing and the progress that we have made and there are few hypotheses or There’s there are few hypotheses that I’ve made during the investigation process of auditing the GTM and GA4 so So here are some of the least that I think we probably can have it. So number one So before we go for the list of expectation, I just want to show first our progress and maybe along the way I’m going to you’re going to see where I’m where where exactly I figure out where exactly I have the points for my hypothesis like where where my hypothesis is coming from and then at the end we I’m gonna list out all of the hypothesis and we’re probably going to address it one by one and And we can make this as casual as possible because I love going through all of this stuff and I think this is also one of the points which is I think authors have been spending a lot of time in GTM and GF4 but I think there’s a sweet spot that haven’t yet been achieved. and I just want to ensure that this audit can help them bridge the gap to achieve it. So the first one that I’ve realized is that So when we trying to write, so that was like the whole introductory and everything. So the first one that we want to talk about is the approach that we’re doing when we are the thing. So the idea is that we use this approach, we call it four pillar approach. And this four pillar approach is simply every time we want to implement GTM or JL4, there’s four pillars that we have to figure out. The first pillar is the foundation, like the technical foundation. This is the data maturity for the website, the company. Basically here, what we’re doing here in this report is that we just go through the configuration for GTM and JL4 and see if there’s any discrepancy that we actually have to address. So That’s pillar number one, just the foundation setup, the general setup for everything. The second part will be the who, which is basically now we’re talking about traffic. So the whole idea of we’re building GTM and J4 is to understand who actually our user, who exactly is our website user and as well what they do. So the second place is who who are they? So where are they coming from? What’s the percentage of them? What’s the channel makes of all of them? And why is it that they do it at the website and on the website? So that will be the number three, which is what they do and what they do will be what user, what key journey that we’re tracking. what event that we are tracking to understand their journey properly or to see if there’s any leaking funnel over here and there that they were doing something else that we’re not expecting that they should do. And lastly will be the implementation. So this is the technical implementation, especially in GTM, which is it’s a half technical marketing. It’s almost like development rules, but it’s not so development. It’s also marketing rules. Usually from what I’m seeing as of now, a lot of people in technical marketing have been implementing GTM and J4 without documentation in mind because they’re not approaching it as a developer, a software developer. So a software developer, they will have documentations, they have PR reviews, they have to submit everything. Everything is documented, but not for technical marketing, even though it’s almost the same. Like you developing a a tracking plan. So you have to kind of document everything from scratch. So we are seeing some problem with the documentation, with the technical implementation as of now. The first one would be The naming convention, it’s not so friendly. The first time I went through the GTM, I didn’t really understand what was going on. I have some ideas, this is for GA4 and everything, but why is it doing this way? Why don’t you just general all events, for example, like GA4, this is all events, track everything. Why you have to do it very specific? Like, is there any request for that? Is there any reason we’re doing that? Because of maybe we want to track a specific clicks from this page and then to another page. So that will be the problem that I’m seeing. The second part would be problem. See, the second part I’m seeing is problem as well is the organization like there’s no foldering. Everything is. put up either the same branch so it’s really hard to also navigate around that as well and to understand like oh this whole bunch of stuff because it’s pretty much everything isolated like the tags the trigger and the variables all of them is isolated and So when we don’t have a proper folder link, everything seems like for everything. So we have no idea like, oh, this tag actually only use these variables because these variables is built for this tag. So we’re going to have duplicated variables. We’re going to have duplicated tags because we’re not organizing it properly. So the technical, the pillar four is all about that. Just making it pretty, making it user friendly, making it developer friendly, developer experience, technical experience much better so we can work on it. So those are four pillars that I’m building here in this full audit report. So the configuration are basically done because it’s pretty much this is the surface level audit that I’m talking about. The deeper level audit will be onto the pillar three, which is the event tracking. So I think I kind of solved some of the problem that I can especially for pillar 2 we are seeing some source and referrals coming from chatgpt cloud and perplexity we should be able to assign them as a custom channel report I also seeing that we have some coming from CPC, coming from Google Ads, coming from linking paid social as well. But it’s not a lot. And I’m not sure why is that. Maybe because of the cookie consent or maybe because of there’s problem with the UTM from every age. from every platform itself. So I’m not really sure what’s happening. So my suggestion, if you want me to dive deep into that, I will be glad to do so if you give me access to both, not both, actually three things like the website for me to be able to see if there’s any part. For example, we have two J4. so i noticed that we have two j4 so uh the first j4 i believe um the reorg marketing gtm that was the previous website j4 and we don’t retired that j4 immediately so instead we slowly mitigate to the new j4 which is autos marketing but then the autos marketing also have data streams and also some other parts coming from reorg marketing So I believe the idea is to migrate one by one until a year or a year and a half and we only use one JFO. And if that was the plan, well, I think in the Audi, I can be more thorough. I can actually say that, oh, we should just remove this JFO and just use the same Google Tag because we can use two Google Tags for one GTM container. There’s no problem with that. And we can definitely send two data to two different GA4 and there’s no problem with that. It’s just I want to understand what’s happening because there are some events send data to GA4 reorg but we’re now using GA4 Autos Marketing. So I want to kind of organize it. So this is more on pillar four. All right. So If you want, I can give you this as of now. Maybe we can go through it together at other time. Maybe where we can just have a call just where I go through the audit in overall and I can I’m going to work on this. If you want, I can already send you this slide so you can already send it to your developers or to the marketers so they can already fix some of the quick fix. because all of the specific events, the form submissions and all of that will be very specific to the tracking plan, which we’re going to go to the second part of the meetings, which is some hypotheses that I have built throughout auditing the GTM and the investigation. So I notice, I believe there’s four points. The first point is that I noticed that It has been a problem of the redirection, redirecting from reorg.com to autist.com. So it is a problem. So I noticed that some of the implementation in the past is trying to figure out that can we track people coming from reorg.com at GA4 for autist.com. So I’ve seen this implementation of page referral URL, which is not active. wondering if that’s still something that you guys want to track something that you want to try still to understand whether people is coming from real calm to that so there’s a way we can do that but it probably requires some different type of redirect so there’s more of a developer side it probably add just a few millisecond in loading time but I will have a lot in the accuracy but it will help a lot in the data accuracy. So that’s number one when it comes to page referrer. Number two, I’ve noticed that you come up with a report at JFO, which is a LinkedIn report and also Google Ads report. So I believe what you want to see is user journey. So we now doing all of this campaign and we’re sending people to our website. So what they do in that website and what steps that they take. And what exactly that we’re dropping? So are we dropping at the trial? Are we dropping at demo? Are we dropping at the first landing page itself? they actually went to another pages right so so I believe you want to see it in the funnel form is it something that you already achieved or is it something that you want to achieve here as you focus if you want to achieve here I can somewhat will you a tracking plan that they have a really important key events so So that in the pillar three, I can start to remove an advice to, OK, maybe we should remove these events and just focus at these events. OK, so that’s that’s hypothesis number two. Hypothesis number three is that I want to understand like, OK, so hypothesis number three is about internal and also cross domain referral, I’ve noticed that there’s a lot of domains. So there’s staging domains, there’s appreorg.com staging domains, there’s wroteinreorg.com staging domains. But we have problem with tag coverage, which is basically the GFO or GTM tracking all of these domains, but it’s not properly tracked. So the best way is to actually just remove this and consider that any of traffic coming from this will be like internal traffic. So we can actually do that. What we can do is that we can develop a custom channel group that ensure that if they’re coming from app dot reorg.com if they’re coming from all of this subdomain if they’re coming from internal utm parameter if we do some kind of any parameter and then just put internal we then filter it out into very specific uh channel grouping so whenever that we’re building report we just exclude this channel uh we can actually do that and i advise you to actually do that instead of um trying to build up, you know such massive filtering function that, like you say, it’s not foolproof, like tracking from IP address, which is not foolproof. So what we can do instead is to create some kind of policy that probably help that probably all of the employees are bookmarking the same URL, which is using the UTM/internals/through internal underscore equal to true and whenever they went through the website or they went through any of the domains that they’re working on and they still have the same domains and they don’t have a new tab uh then and they don’t open up a new tab um we still can track that they are they actually is internal traffic and we can exclude them from all data that we have so um yeah so I think I think those are the three hypotheses that I want to highlight after going through all of the audits. This hypothesis is important because I want to understand. Yeah. So the last one I want to talk about is that if you want us to somewhat test, is it more depth or you want it to be like general auditing? Because if you want it to be more depth, I would love to test much further to a point that I even test the demo form. I even test the contact us form, whether it’s non SQL or SQL. And that will need us, that will kind of need some more engagement, which is I kind of need to have the testing email, probably some access for Google ads and LinkedIn ads as well, if that’s possible. So I can check on the conversion pixels So that would be the last part. Okay, so that’s all.