Meeting Title: Brainforge Data x Carlos: Discovery Date: Dec 4 Meeting participants: Shivani, Awaish Kumar, Carlos Hernandez
Transcript:
Them: Hey, guys.
Me: Hello.
Them: How you doing?
Me: Good. How are you?
Them: Dude, my cold ended up turning into a sinus infection.
Me: Brutal.
Them: So I just started antibiotics, but I’ll probably go cover off at, like, once I do the intro, you guys, because it’s a little.
Me: No problem. S will be effective. I feel like every time I get on, it’ll crush it, but, yeah, that sucks.
Them: Let’s let Carlos in the room. Hey, Carlos. How are you? Tell us. In the last 24 hours, I’ve gotten a full on sinus infection. I had a cold earlier during Thanksgiving week. And then like yesterday when we were talking, I was like, my head’s really hurting on these. On these? Metric scar? No. And then I got on the flight and I. Was like, oh, I’m in a lot of pain. No. Oh, my God. Getting on a plane with the sinus infection.
Me: Yeah, your head’s going to explode.
Them: I was. I was actually crying. Like, not like, you know, just, like, pain crying. I was like. Oh, my God. My. My eyes were just, like, tearing up and. Yeah. My boyfriend was like, we’re gonna get through this. I’m sorry. I literally actually was on a flight with someone who had a sinus injection, like, a couple months ago, and they were actually crying as well. So I. So that’s just where I’m at right now. And so that’s why I probably go camera off it a little bit. Like what you guys saw because it’s like. It’s just really like went like real quick, 0 to 100 situation over here. But Carlos, I wanted to introduce you to the Brainforge folks. Uttam, who’s a founder CEO of Brainforge and awaits who’s on our team, who’s working with Elemnt. And Utam is based in Austin. And I wish you’re based in Pakistan, right? Yeah. Cool. So we got. We got some spread out over area codes right now, so it’s nice for them to join us today. And if I just, like, share the context, like, you know, as I’ve connected a little bit with source medium, my understanding that I want to play back and like, check with you on Carlos is. That source medium has kind of like a templatized way of helping businesses. And that’s kind of what we’ve been doing with them, specifically, like, mostly focusing on acquisition.
Me: No.
Them: Of customers. For E commerce specifically, not retail and things like that. And so in this world of really having more channels and eventually want to. Wanting to link like, like, let’s say the first three month project is like kind of this like omnichannel thing with Brainforge to think about how do we get all of the revenue sources of data, kind of like in one warehouse. And what does that look like? What insights can we glean from that? A longer term vision that I hold is like eventually putting that supply inventory data into that same warehouse so that you can actually do better supply demand forecasting, because you have a lot of the demand metrics in the same place as the supply metrics. So there’s, like, a lot of benefits that will unlock from bringing the data into one place and being able to, like, do different cuts and views and dashboards and things. And so if, if I play it back correctly, Source Medium is kind of like, specifically mostly used by you, N. Blake. Maybe other people are using it, but you, I’m imagining you two are like the top stakeholders, like, looking at dashboards. Yeah. So it would be myself, Blake from our partnerships team, and Jason, who’s our head of tech. And then we also have Landon, who is on our CX team. So for context and by the way. Very nice to meet you both. We also have Landon, who’s from our customer experience team. So our. We’re a little different in the sense of most businesses. We’re not really E Comm. Anymore. Most businesses will have a retention team. We don’t have a retention team. So we have, like, retention functions that are split throughout the business. And our CX team actually manages our email workflows as well as our email campaigns, and that data comes into Source Medium. So we also have CX that occasionally goes into Source Medium, but primarily it is myself and Blake and then Jason, and then we do have CX that goes in there a little bit as well. Thank you. That’s perfect. And so that’s the, the gorgeous data that would be in that struct or in that. In that looker table.
Me: Great.
Them: And so, so I think that I just wanted to share the, like, high level vision, right, of, like, where we want to get to where all the data is, like, in one warehouse. And rather than using, like, BigQuery over here, either we’re using bigquery for everything, or we’re using Snowflake. For everything. And then we pick our own bi tool. And it might mean eventually recreating some of the, like, rate dashboards that are in the looker dashboard so that we all have it in one place. Right. So, like, that’s not what we’re starting off with. That’s not the highest value. ADD thing, but I think that’s, like, eventually we might want to, like, consolidate how our data stuff is done, I guess. Right. So with that, I think I wanted to create space. You. You and I, Carlos, have done a little bit of, like, you’ve oriented me to your E Commerce dashboard and some of the metrics. That you think about? I think Uttam and Wish will probably have like some high level questions and then some like really detailed questions to go through. How do you define this metric and, and, and things like that. So, like, today is a kickoff of a discovery, but it might mean that, like, we’re following up in a week to do like, a bit of a deeper dive. Does that run right? Utam in a Wish. Did I. Did I throw that? Okay.
Me: Question. Yeah, that’s perfect. We did go through Carlos, like your ecommerce growth dashboard, and our job is just to get a sense of how you’re thinking about metrics.
Them: Okay?
Me: And kind of the priority metrics. Of course, I’m also interested in what value you’re getting out of source medium today and what’s working, what’s not working. And then, yeah, just kind of, like, hear a little bit about how we arrived. At the current structure for reporting on the ECOM business. Of course, our primary goal initially is to support the commercial reporting side. So we’re talking about ecommerce. We’re also talking about retail wholesale data. And so just understanding each channel. We did already speak to source medium and we also have spoken to Jason. So do have a good understanding of those two pieces. But yeah, probably first conversation of many, but also just put a face to a name.
Them: Yeah, absolutely. And it’s utam and Awaish Reset. Yeah, I wish. I wish. Okay. Perfect. Yeah. So. Absolutely. I think that I can walk you through ecom growth dashboard. And talk you through the different data sources and where we get all that data from.
Me: Right.
Them: Well, I think the best way to describe what the goal is, like the whether it’s source, medium or any of these other data sources. There’s two things that I’m looking for is the data to be consistent and then for the data to be accurate. Obviously, like, those are two very obvious things, but in reality, that’s literally the only thing that I care about. I’m using multiple sources right now. And I believe at this time that those are the most accurate sources that I have for this data. So I think it’s like you said, if the best format is for me to walk you through, through the growth dashboard and then give you an idea of why, you know, why I have each individual, well, maybe not each individual metric, but, like, why we’re looking at certain metrics. I think that might be the best way to format this call. Is that what you’re thinking, too?
Me: That’s. Why? We defense way to. Night. Yeah. That’s perfect. And throughout this, I think feel free to explain a little bit about, like, hey, we’re doing this, but it’s kind of jank right now. Or this is new, or this is something historical. Love, especially to hear the context they give you. Maybe some brief background. We do quite a bit of work on the data side in ecom. So a lot of 50100 million dollar brands, all the data around combining Shopify, Amazon, wholesale retail and so very, very familiar with. You don’t have to. We’re very familiar with the ecom world in terms of data. Also very familiar with probably a lot of the data challenges that you’re dealing with. Pulling reporting out of source systems modeling here at Spreadsheet. So ultimately, our goal is to achieve both of those things for you and also scale your reporting to other folks. And then ideally, along the way, we’re also replacing a lot of manual data preparation. So even as you go through as much as you can talk about, like, hey, this is something we do on a daily, on a weekly, on a monthly basis, it takes, like, 10 hours to do.
Them: Oh, okay, now you’re talking. Now you’re talking.
Me: That type of stuff helps me quantify.
Them: Because if you’re talking automation, I’m all for it.
Me: Yeah. That type of stuff helps me quantify how we can impact your day to day. And to tell you how we’re working with Shivani and Phil. Is right now, we’re in a mode of just looking at everything. Kind of the theme of our engagement is like, start slow to then move fast. So, please, anything where you’re like, I wish we had this data or this taking so longer because it’s taking so long. I’ve never been able to address this metric. Those are the kinds of things that’s very, very helpful from the context side.
Them: Yeah. Absolutely. All right, you can see my screen now, right?
Me: Yes.
Them: Let me actually do this. So I think the best way to. To. To start is and. And I’ll get through as I walk through this. You know, what we were previously doing and how we got here, but I think that it’s easier to start at the top and show you the. The report that we provide to the entire company. And then we’ll drill down into how we specific how we get that data ourselves to be able to provide the report. So I’m sure. I’m sure you’re familiar with, like, what are we have okr. Deliverable Tracker, which is where we report all of our key metrics to the entire company. So I specifically operate within the ECOM business, with an ECOM side of the business.
Me: So.
Them: I do help a little bit with retail, but primarily with ecom. So what I myself and my team report on are total ecom cells. So that is going to be at this time it is Amazon, it is Shopify and it is Walmart 3P. So Walmart third party. We don’t do Walmart. 1P we also have a report on all spend. So we have performance spend, which is we have performance. In Branson, Performance spent is anything that is attributable, meaning that there’s some type of view or click attribution and then brand spend are things that are not potentially we can attribute, like out of home, brand partnerships, things like that. Tv. I know there is some attribution for tv, but for, you know, we’re just labeling it brand.
Me: At this. Time.
Them: And then we have metrics like ecom mer, which is just marketing efficiency ratio and that’s revenue divided by all spend.
Me: Which. Is just.
Them: Attitude roas. And then we do have one retention metric that we look at in my group for econ, which is just our total subscriber count for across all of everything that I’ve mentioned is across all of those three those three different marketplaces or ecom sites or DTC site. And then as I mentioned Brian spent. I won’t go into these individually because this is going to go into the other in the growth dashboard. But when I report to the rest of the company every single month, those are the metrics that I’m reporting on. So that’s very high level. That’s going to go to the entire company. That’s going to go to the executive team. Below that. You have, which sounds like you’ve looked at which is the growth, the growth dashboard. In here. Actually, before I continue, any questions on that side of things?
Me: And then for all of this. Are you pulling directly from Shopify?
Them: So most of it is coming from source, medium in ad platforms. So there is some. There’s some ga. Now, I’ll go through. Through the metrics once I go into the individual ones.
Me: Okay?
Them: But. Yeah. So source medium is a primary. Primary source of the data. Which all of it obviously comes from Platform and then Amazon as well. We get some data from source medium for Amazon and we get some data from Platform itself. Because there’s certain limitations that I see that I’ll get to in a second. So in here right now. Ideally I this growth dashboard does have owned ecom in retail. I I don’t actually I put this as a, you know something I want to look at down the road right now we don’t have a way to get automate pos sell so what I have to do for for retail. I have to actually ask our finance team to look at POS sales and sort of build, you know, because it’s. It’s not perfect because they work off. I believe that they work off of a weekly calendar, so sometimes there’s overlap across two months, so we have to, like, manipulate the data. To make it look like it’s one month, but so that I don’t have that included in here yet. So in here.
Me: And do you know the pause on that? Do you know if that’s what’s coming from spins like, that’s what they’re getting?
Them: Oh, I don’t know where it’s. I’m getting it from Dan. I don’t know if you guys have met Dan yet. He’s in our finance team. He’s on Shivani team.
Me: I think we met Dan, like, maybe. Yeah, maybe a month ago, but, yeah, that’s helpful.
Them: So I get it from Dan, and I’m assuming he get. He’s able to access POS sales. I’m not 100% sure how he gets them. So in here, this is where you start to get into. So we’ll look at. We’ll look at November, right? So this is where I. And I’ll go over quickly. I’ll go over this quickly because then I’m going to get into how I get these metrics. In this growth dashboard, you have everything that rolls up into E comm, and you’ll see a lot of those same exact metrics. That. That you just saw, right? So you’ll see all of the. The sales spent, and we are all of that. Below that. I have Amazon rolled up in here so you can see all those individual metrics. Shopify. And walmart.
Me: Great.
Them: So where I’m getting all of this, this is where we get into the. We get a little more granular. Is I’m getting all of these are rolling up in from these two tabs. Well actually these three because of brand spend, but they’re rolling up from this tab where include that includes all of our drink mix revenue across those three channels. So it’s again Amazon, Shopify and Walmart. And then we have sparkling. In here. You’ll have the way that this is broken up is into these individual like revenue channels within those individual platforms. And you know, everything is broken up into non paid and non paid. So you’ll have. For Amazon specifically, I’ll keep using November Target as the example. You have Total Drink Mix, Amazon sells here, but this is all rolling up from all of these different sections here, which right now other sells and all of that. And I promise I’ll get to how I’m getting the metrics. I think just want to give you the. The high level first. So all of that rolls up into there and in here, the way that I’ve built this is. Everything in here that’s blue is a manual input. So any I know. Typically in finance, that’s more blue is used for assumptions. But, you know, in this scenario, blue is used for inputs. So anything that’s in here that’s blue would be an input. Anything that’s, that’s in black font is a, is a calculation.
Me: Question. Happy. There’s some color coding, by the way. I mean, I used to work a lot of finance data, and, yeah, we do a lot of Excel work, but usually you don’t. We don’t get nearly this much. Like the format is actually really, really helpful to see it this way.
Them: Okay. I appreciate that. It took me a while to build this. This is. This is. I wouldn’t say night and day from what we did in the past, but it’s definitely a step up.
Me: But.
Them: But. Yeah. So everything in here is blue. That is. That’s an input. And we have, you know, what our target is and what our projection is. And I’ll go into the metrics and where I get those in a sec. So below that, you have the same exact thing for. For Shopify, the difference between. Shopify and Amazon is because we use source medium in in Source Media muses GA4 we’re actually able to break up our total revenue into a lot more I I I can keep calling them revenue buckets or channels.
Me: The difference? Between. Your. Bucket.
Them: So in Amazon. We only have. We don’t have the ability to break up organic direct or returning customer sales that are non paid. So the only non paid channels that we have for Amazon.
Me: Training. Customer. That we. Have to.
Them: Are these two, which is other? Amazon sells and subscribe and Save sells.
Me: Subscribe. To success.
Them: Below that. The, the remaining is performance spend. We turn in customer sells, which is basically any sort of revenue from returning customers that is attributed to ads. And then we track new customer sales separately. And the reason we do that is we want to be able to track customer acquisition costs versus our, you know, our total cost of purchases. So we separate attributed revenue from NU in returning customers to be able to get that number.
Me: Separately. In the reason why. We do that? Is we. Custom. Ize. Advert. Ising. Custom. Ers to be able to get and isn’t all. First, your guys are fulfilling all Amazon orders. You’re not doing any fba.
Them: It’s NBA. It’s all fba.
Me: Okay?
Them: And then for Shopify, this is what I mean. For non paid we’re able to break it up into a lot more calling it channels. So we have subscriptions, email, direct organic SEO online.
Me: You’re relying on the shopify attribution.
Them: We are using so source media, we’re relying on source medium. Source medium is the one that is able to allocate, obviously using the source medium in identifying where, what the attribution channel is.
Me: Obviously. Using a nfor. Ce medium. If I can find. You’re relying on their attribution models to power it. Sure.
Them: Correct.
Me: So maybe if I can pause on that.
Them: Yep.
Me: The history of that. How involved were you in determining that attribution model? Like, how confident are you in that?
Them: That’s a good question. I would say 90, 95%, me being involved in that initial process, zero. It was actually brought on before it was brought on years ago, before I actually came on. But the way that I understand it, as long as that. As long as our UTM structure isn’t. Doesn’t break or we don’t. Change it, the data would have consistently been reliable. However, throughout the the years I’ve noticed some inconsistencies, and sometimes the inconsistencies happen due to ga because they’re using GA data.
Me: Init. Ial process. Ut. M structure. Isn’t. Our respons. Ibility.
Them: G8 changing something like, for example, where when we launch, when Google launched GA4, I think now it’s been a year and a half, two years maybe. There was a huge, huge shift in attribution for direct versus Google Ads. And they went through this period of, I guess, Being more accurate with the attribution. So there was about three months of data that was what I consider completely unreliable. And, I mean, that was like the, I think, the biggest concern that I ran into. And then now this. Since I shifted to this, you know, this type of format for reporting. I’ve gone back and I’ve noticed a couple of issues with the way that things are being reported in there.
Me: Report. Ing on.
Them: I think that what I’ve identified, for the most part, All of the data is there. However, sometimes the way that reports are being created in there are not completely accurate. But like I said, I think it’s 90, 95% accurate. And, and I’m actually working on that this, this month. That’s part of my, one of my, we call them sprints, but One of my primary goals for this month is to go through all the source medium reports and try to poke holes in them individually to try to figure out where data is. But if I had a, you know, throw a number out there, I think 90, 95% accurate.
Me: Reports. Have been created. That’s. Called. Just. Because. You. Try to. Okay?
Them: Awesome. And then the last thing. So for then, for Shopify, we have all the different paid channels, which is, you know, I don’t think I have to walk through them. You can just see all the different paid channels that we have. And then we have Walmart at the, at this time, we only have. We actually only have Walmart organic. So we’re just running through organic. There’s no paid, but there will be paid down the road. So. But right now it’s just all coming from non paid cells. And then for sparkling. It’s all the same things that you saw for drink mix. And then not. I don’t think there’s really anything to go into for BRN spend. So I think the next thing that I would do is now I can walk you through the, through the blue, through the inputs and then show you where I get all of that data. But before I do that, does that. All make sense. Any questions there?
Me: Yeah. Before. You. Aw. Ait.
Them: Yeah. So all this data, like the ad spend you do is into it on these individual platforms, or you’re using any, like, platform, like naming to basically. Yeah. So we do. Most of the time we buy direct from the platforms. The only DSP we use, like, why am I drawing a blank on, like, the biggest TSP that there is?
Me: The big. Gest tsp that there. Is.
Them: It’s not one that we use, but anyway, so we, we do use some DSPs. We’ve used stack Adapt in the past. Demand side platform where we, you know, they, they have access to inventory that is not necessarily direct, but most of our ad buying is direct. So we buy directly from Meta. We buy directly from Google. The only time that we don’t buy direct is when we’re buying on the Amazon dsp. So Amazon’s DSP has access to Amazon owned inventory and they also have access to other inventory like you can buy on Netflix and Hulu directly through Amazon’s dsp. But most of it. All of it lives in the DSP console, even though we’re not buying direct. And then we also buy linear tv, which, you know, we’re using Tatari as our agency of record there.
Me: Some. News. Necessary. Direct. Ly. Time. That we don’t want. Inventory. And they also have. Access. To. Netflix. Directly. To NFL’s d. Sp but. Use. The console, even though. We’re not coming direct. Ly. With your team. Mate, which. Brings. To tony. Being a. Super.
Them: And I, I’m not 100% sure, but, you know, but they’re obviously not buying direct, right? They’re. They’re working with, with the third party to buy inventory across different networks, but. Yes. Does that answer your question?
Me: Question.
Them: Yeah, that’s for the. You are running ads for different products. But what about brand span? Like. That how. How that is done. Yeah. So for brand spen aki nwaki, through the different brand spin that we have,
Me: With different. Brands that we have.
Them: So we have TV brand spend right now. We don’t have anything planned for the rest of the year. But what we’re doing, this is Tatari. So we’re using Tatari as our agency and we’re buying Linear with them. So that’s how the brand spend. This next thing is our. This is Amazon DSP brand spend. This is what I was just mentioning. Where we buy, we’re buy. This is the. The Amazon DSP has performance spend that we can get attribution for and that’s broken up into these other two tabs because it’s attituated. But Amazon DSP also has brand spend, which is those, like, streaming TV that I was just mentioning. And then that’s how we, we keep this here. But it all reports back to the, to the Amazon ad console. So we have access to that data through there. And I, as I go through the inputs, I can pull up all those. Different consoles to show you where those come up. And then YouTube is the only other at this time, the only other brand spin that we have, which is just bought directly through Google’s ad console.
Me: Agen. Cy and revised. Training. Where we. Can’t. Get out of. These. Cons. Oles. So we have just two out of. 800. Different. Consoles to show you what? ’s going on? At. This point. Which. Is just directly. Out.
Them: And we do. We did put on the budget to potentially buy some audio in 2026, but obviously none in 2025.
Me: To potentially. By. 2000 and 20. Six. But. Obviously. In2025.
Them: And then we also have. This is the brand partnership. This is more of a traditional brand partnership with a MMA organization. So we. We sponsor a MMA organization that runs. Their format is changing this here, but next year, in the previous years,
Me: This is. The. Soviet. Union.
Them: It was a format that their season started in April and it ran all the way through through August. So that’s the only brand other brand partnership that we have and we do have not this year, next year. So I’m out of Homespan as well. That will be it will be leaning into. But this is all of our brand spend and most of it will usually we’ll be able to get from the individual consoles, except for the PFL because we just get invoice and we’d run ads on their events. They do a couple of AD REITs throughout the the broadcast and we we get a couple of TV placements because they run on espn.
Me: Most. Of it. Is. Throughout. This. Year. So the other thing we saw is we have access to a copy of the Partnerships team dashboard. So what about that sort of spend? Is that accounted for? Okay?
Them: Yep. That that is accounted.
Me: That’s all related to a product.
Them: Yep, that is it all. It’s all going to be performance spend and that is right here.
Me: Okay?
Them: That lives right here.
Me: Great.
Them: Yep. Okay, so I’ll walk you through the inputs now. I think that’s the part you guys are waiting for. And where we get the data from. I’ll actually start high level because we have so that this is one of the things that’s a little annoying. And you’ll see as we’ll go through some of these metrics. There are certain metrics that I have to pull specifically from, from source medium, some that I have to post specifically from, from Amazon, and so on. So I’ll start. I’ll actually start with the inputs here, and I’ll actually start with Walmart, because Walmart is the perfect example of it. So for for one more aov. So the three inputs here, we have aov, you have traffic, and then you have. I’m calling it right now, I’m calling it Source Medium because the goal is to be able to bring in data into Source medium. But it’s really Walmart, new customers, and I get all of this data from. From Walmart itself.
Me: Specifically. For. Those. Specific. Input. S. Customers.
Them: Do you guys want me to show you exactly where I get the data from? Is that helpful? Do you want me to just tell you? Okay.
Me: And we’ve been in seller central before, but, yeah, it’s just helpful to see the process.
Them: Okay? So I’ll start with with Aov in traffic. Because those are the easy ones. So what I’ll do for aov, I come into their analytics. Console or report. And then I get. I look at this. This number up here, this is their AOV number. So I get the A number from here.
Me: Here.
Them: And then the. The new customer, this number right here. I actually get from this customer? Insights. And. This number here is the new customer sells, the percentage of new customer sells. And the reason I use that number is because that’s how I identify our total new customers. And that’s the number that I use for our customer acquisition costs, of course.
Me: Is. Custom. Er service. Because. You’re doing this, like, day after close or.
Them: This. This I’m doing. So that’s. Yeah, I’m doing this day after close. However, I do actually have a daily tracker. Where I will right now. I don’t do this myself. We have. We have a coordinator that, like, has to do this. Has to do it manually. The way we were doing it before. Keep in mind, this just changed this month. In the past, I was able to automate our daily report where all those metrics that I just walk you through, we also track on a daily basis because we obviously have to, you know, node whether to move, spin up and down. So all of those are tracked on a daily basis. In the past, with our previous reporting template that wasn’t as detailed as this, we were using two ways to automate it. We were using supermetrics. There’s one way to automate it. And then we were also using amc. We were using a sheet that was connected through AMC that had queries that were bringing in Amazon data.
Me: Having. Those business. So.
Them: And then there was still like, it was probably at that time, I would say it was like 60, 80% automated. Now, right now, it’s zero percent. So we. We have to pull this every single day. But so for the new, this number, this input right here, the source, medium, new customer sells percentage. I have to go to Customer Insights. Walmart is the only marketplace right now that doesn’t actually give us a true new customer number, a new customer order number. So what I have to do is I have to go into the customer insights and then I have to look at the, the repeat order percentage. And they don’t even have a. I mean, it would be helpful if they. They had a. An actual selection for dates.
Me: That.
Them: But I have to look at what I think is the most accurate window. So, for example, At the end of the close. What I have to do is I have to look at the last 30 days.
Me: And. Then you use a percentage of the last month.
Them: Yep. Yep. And exactly. And then for the one that we’re doing, if we’re doing this on a daily, for the daily numbers, what I have to do is look at the last seven days, and I just have to continue to use that rolling seven days.
Me: Damn.
Them: Yeah. And I mean, if there’s a way, this is one of those where, like, if we can get that accurate data. This is the most accurate data that I have at this time. Where, like, if I’m looking at the end of the month, I have to look at the last 30 days. If I’m doing it at the last. I’m looking at the daily numbers. I have to use a much smaller window.
Me: Exactly.
Them: And just for my own, like, verification. Sorry guys, but between YouTube, Utam and Awaish, are you taking detailed notes about these, like, clunky definitions and like, methodologies?
Me: Yeah. I mean, we. Yeah, I have Granola going on my side, and this is. We actually just did this for someone else. Like, we’re in seller central, like, last week.
Them: Okay? Okay, gotcha.
Me: We’re really familiar with like Walmart data, basically like part of like any solution that we develop. It’s like you want to avoid going for aggregates because what’s going to happen is on things like refunds, discounts, where there is like a time bound decision.
Them: Cool.
Me: Shopify, Amazon and Walmart are all calculating it very differently. Like the attribution differently. And of course, if you just literally, if you just select the wrong filter or something, like, you could get the wrong input, right? So it’s fragile in that sense. But, yeah, familiar with.
Them: Yeah. I will say the, the plus side is that I think like you can almost call Walmart revenue non material. It’s just so small at this point that it doesn’t actually impact the overall number. So. But it’s still like, I still want to try to get some type of accuracy with the data. This is, this is at this time, after even confirming with Walmart, this is the most accurate data that, that they have. If, obviously, if there’s a different solution that is more accurate, I’m all for it.
Me: The. Type. Yeah. This is. The one. Okay?
Them: Awesome. All right, so the next thing here is traffic. Actually, give me a second. I wrote a note here of where what the report is that I use. Actually, this report. So this is. No, I don’t want the. Our. My traffic. I want. To do this. Amazon’s traffic. Wait a second. So what I use here for traffic numbers, I use the item sales report. And then I go to item, item, page views. This is. Yeah, this seems to be the most accurate. My understanding is that it could potentially include some reseller traffic. However, we don’t have any resellers on Walmart. But even if it for me, obviously I would prefer to use, you know, traffic numbers that don’t have resellers. That number, to my knowledge, is not available. But we’re using item page views because I want to look at the overall traffic to the, to the website or, excuse me, to the pdp. We don’t have a brand store right now. When we do have a brand store for for Walmart, then you know, I’ll probably want to look at that traffic as well. But right now we just have PDP images. Cool. All right, so, yeah, that’s. Those are the. On all the inputs for. For Walmart. Going into. This one’s a little bit. Actually, a little bit more complex. So for Shopify. What I do. I’ll start with traffic. I just use g. A numbers and I use sessions, and I use all sessions. So what I’ll do is I’ll just go to g.
Me: So none of the traffic data from Shopify is getting leveraged anywhere.
Them: No, because I want to. Well. The reason I’m using. GA data is because we’re using Last Click.
Me: Cat.
Them: Revenue, so I want to be consistent with that. That’s the reason I’m using G8 data.
Me: Okay?
Them: So then I just go to the reports and then look at the. The acquisition reports and get the sessions from there.
Me: Great.
Them: And then for aov, what I’m using.
Me: Great.
Them: Is. I’m actually using source medium.
Me: Great.
Them: In source medium under the orders deep dive report. There’s this acquisition channel report. And I obviously make sure to include. So what’s included in here? Is all online D to C, which is shopify, but it does exclude a wholesale customers.
Me: Great. Okay?
Them: And then I’m using this AOV number here, which is all our total A number.
Me: So Shivani, like questions I’ll have for source medium are going to be like, how are they calculating some of these things? Basically. Like, what is the logic under the hood?
Them: Yeah. Like, does that air table definitions tell you or. We need to go deeper with them.
Me: Yeah, I probably need to call Chase again.
Them: He has, like a calendly sort of thing.
Me: Okay. Yeah. It’s basically because, like, Carlos is taking it from source medium. Source medium is either. It’s either going to be straight by Shopify Data. Or combining and doing something. So I just want to understand, like, what the logic is. But at least it’s good. I think at least the good thing I’m seeing here is it’s very. We’re very consistent, at least in the decisions from where we’re sourcing things. You know?
Them: Yeah, I appreciate that. That’s a big thing that I was trying to do. As we’re putting together.
Me: That’s the thing. Whether it’s right or wrong. The tough part is we go to places where they’re like, oh, in this situation, I go here in the switch. Then it’s like, really tough. At least here. We’ll. We will. We. We’re. We were like, okay, we’re doing it. We’re do or die by this source. And that’s. That’s, like, at least a bet. That’s one thing to go, understand?
Them: It’s apple sabbles. Like the number fluctuations mean something.
Me: Something. Yeah, exactly.
Them: All right. Yep. So that’s a in that right here. So the next thing is going to be new versus returning customers, which also comes from source medium. So in here, they have multiple filters, but basically what I do is they have an order sequence filter. And then I filter by first and repeat.
Me: And then. Of course, everybody.
Them: And then that’s how I get. I get these two numbers here.
Me: Get. These. Two numbers. Here. So I guess one question I had is why not ask source medium to just build. Some of this there. I don’t, I, I, I kind of understand the, the reason. I kind of understand that there’s a lot of manual inputs, but, like, was that ever on your mind to be like, hey,
Them: You build what you’re looking at here within source medium.
Me: Yeah.
Them: Yes, definitely. I think that that’s the. That is the goal. To be able to do that. If I can just keep it all in source medium. Great. However, I still like a spreadsheet. Because you can manipulate the data and, and, you know, if we ever have to do an analysis, I mean, have you ever done an analysis and not exported data? Like you’re going to export the data regardless, so. But visually, yes, I would love to have this in source medium.
Me: You have to backend up.
Them: But we’re just, we’re getting to the point, which is. It’s good, but we’re getting to the point of just having to build the complete custom plan for us, basically. But if we get alignment, I think if, if Shivani and Dan are aligned, if the ECOM team is aligned on, on. What What I have here, visually, in terms of metrics and reports.
Me: To the. Point of just having. Custom. Basically. I. Think it’s. Important.
Them: Then I think that would, then at that point it would make sense to move forward with Source medium to have them build the sort of this custom report for us. So, yeah, I think we. That’s definitely on the table. All right, so moving on to Amazon. So the, the buyback buy box win rate, I just get that from Amazon directly. I can show you. Where I get that. Unless you guys are familiar with it, but actually, I’ll just pull it out. It’s just pulling it from the Amazon business reports. Because this is to me, to my knowledge, this is the most accurate way to get the buy box. And you just get it from this. It’s this. This column here.
Me: Can you? Just get it.
Them: The feature offer percentage and then just set the the specific date.
Me: Feature. Specific. Date.
Them: There is a little bit of a delay. Usually you won’t have data, you know, a couple of days.
Me: Using.
Them: Until after. But this is the the buy box percentage the reason why. I keep track of this is that’s one of the most important things on Amazon, is winning the Buy Box. And from issues in the past that we had with Buy Box. There seems to be a one for one ratio between buy box percentage and total revenue gain or loss depending on what you have. So it’s not. Obviously it’s not accurate, but from what when we’ve seen an impact to the buy box, it seems to be that 1% equals about 1% of total revenue 1 on on all ASINs.
Me: Tor. Okay?
Them: So that’s where I get that traffic I also get from. Because we don’t have this in source medium. I mean, if we. Again, we can. I’m assuming there’s a way for us to pull this into source medium, but source medium would be getting it from here, and this is where I would. Want it from. And that’s just from the sales and traffic report.
Me: That’s just. From the traffic. Report. Great. Yeah, we’re from. We’re really familiar with this as well. So yeah.
Them: Perfect. And then Source Medium. New customers versus returning. It’s the same exact thing I just showed you guys in Source Medium, but I just changed the the channel to Amazon. Yep. And then this is. Sorry, go ahead. This is actually one that we’re having. We’re running into a pretty big issue right now. So in the past. So I build this template. Because I wanted to know what our blended customer acquisition costs.
Me: Because. I wanted to know. Custom. Er.
Them: Was because at attribution is isn’t always reliable. And I think we’re seeing a lot of cross channel sales where we are advertising to people on on Meta and then they’re going to go buy on Amazon. And we we’ve actually started working with an mmm that has showed showed us that that is actually happening with some of the channels.
Me: In. I think we’re seeing a lot. Of training. Sites. Where? Advert. Ising. Different. Actually. Happening.
Them: But another way to do that too. I thought it would be first party data.
Me: With. A.
Them: Would be to get just look at all total new customers and then look at what our blended customer acquisition costs. That didn’t account for any of the issues with the attribution, however, with with Shopify.
Me: New. Customers. And then. Coming. From u. M.
Them: Having ran into any issues with this.
Me: How. To bring into.
Them: With Amazon. I did actually notice a pretty significant issue in the past where I pulled all of these numbers. Historicals were pretty consistent.
Me: You know? There is a pretty. Significant. Distribution. Consist.
Them: However, when we switched to this reporting template in November, what I’ve actually noticed is that it looks like there’s actually some type of delay of the categorizing of new customers.
Me: Ent. However. We switch to this. Reporting table. Not. Ices. Are. Tagging domain. Of. New customers.
Them: It sounds like you’re already familiar with that. I wish I would.
Me: It’s. Something you’re refusing. I asked like. How? How? How. Long are you waiting to do? The yeah.
Them: I Awaish. I wish you were. I was talking to you a month ago because. Because I was looking at all the historical data. I didn’t actually have that context because it all looked consistent. I mean, we’re not waiting, right? We’re just putting the. And we’re just. I guess. Let me ask. You.
Me: Context. Because it’ll look consistent. We’re just. Putting that. I can. Finally ask you.
Them: How long should we wait? Is it. Does it take 15 days? Does it take 30 days?
Me: How much should be the. Weight. Is it? 15. That’s the tough part, because, I mean, What? What we will end up doing. Is. I mean, we will build. We’ll basically build backup from, like, the most raw data we get from Amazon. And usually.
Them: But it continues to update, right? It seems like for me, what I’ve noticed was when I was looking, I looked at this in the first week of the month, right?
Me: For me what I notice. Was when I was. Looking at? This in the first week of the month. Right.
Them: In our new. Our total new customers were up by about 30, 50%. I know it’s a big range, but I don’t remember exactly.
Me: In our new. Our total new customers. Were up. On 30%. I know it’s a great measure. Ment. Number.
Them: But our spend was actually down, and this is our down season, so that made no sense to me.
Me: So.
Them: Now that we are, you know, four days removed from November, that gap is much closer, and it looks like.
Me: What we’ll basically do in, like, these types of situations is we’ll. We’ll do. We’ll both take a snapshot. On close. And then you’ll publish revised. You know, on whatever floats. So there will be some stuff that floats.
Them: Yeah.
Me: And for like. A good example of this is like, returns, and you may need to revise returns for the past month.
Them: Yep.
Me: But like you already may have reported out like we didn’t like mid month returns. Or, like, refunds or whatever.
Them: Yeah.
Me: So often, the way we. We would handle this is you. Do you still publish a snapshot? And then you have something that’s more live.
Them: Yeah, that’s what we’re having to do this month, essentially. And I just. So, Shivani, for your context, in the end, a month report, I just published what we have, and then I put a note that there’s a latency issues with the data. And I think, I think that’s probably where we’re at. We’re. Well, land is like. We’ll publish it and then we’ll have to. We’ll have to revise the data and we’ll try to figure. Well, I think we’ll need to identify if that’s clear.
Me: And then. You put a note? To. I. Want to be. You. Can also understand, like, hey, are we. Is it often off by a certain amount? And you can input some, you know, assumptions, but again, it depends, like, if it’s really being used, like, closer to accounting, then it’s tough. Right? And so some of this, it will float for a little bit. And you will have to go look at, like, actually have, like, a couple. Some stuff got refunded or some stuff got adjusted, and here’s a floating version of it.
Them: Yeah, and by the way, this is. It’s not AMC New Customers. It is new customers using the hashed emails.
Me: Personal. Ization is using the. Okay? Yeah. Yeah, yeah. Yeah.
Them: Okay, cool. Perfect. Yeah. No, I mean. Great job, Shivani. Sounds like they. They definitely, you know, the team is great. So this is some of these things I brought up, like, I’ve. I brought up in the past, and, like, sometimes, you know, there’s just.
Me: Great. Sometimes. You know, it’s like a very. It’s like a. Yeah, it’s tough. And you may be the only one that they’re like, thanks, but, like, what should we do? You’re like you introduced it. I have no idea. What should we do? So there’s some ways of doing. There’s some ways.
Them: Yeah.
Me: Of guarding against this type of stuff, so awesome.
Them: Awesome. Well, yeah. So those are the inputs.
Me: Can I ask a question about, like, how you’re doing the forecasting? Is that, like, a loaded question?
Them: That’s a great question. Two there’s two ways that I’ll tell you how I wanted I we were how I was doing forecasts for paid.
Me: How you’re doing the car? Yeah.
Them: Was using historical, in, year over year trends because there’s a, a very clear seasonality curve to our business.
Me: Because. Clear. Ance seasonality. Closed toward business.
Them: Where our conversion rate dramatically increases during the summer.
Me: With. Decreases. During the summ. Er.
Them: Our AOB increases because one people, you know, more people are going to buy electrolytes in the summer. In people that are customers are probably going to buy more. Right. Because they’ll go through it a little bit more quicker.
Me: Dmp increase. S. Because people. Like. Temporary.
Them: More quickly. So in the past, historically, paid ads have been forecasted using historicals and then also, like, using, you know, things like, like. Like using ideas of, like, diminishing returns. Right. So if we wanted to hit a certain target, We were to increase spend. But do we get to a point where by increasing spend, we’re going to take a big hit in conversion rate, where we get diminishing returns? So that’s how paid ads were forecasted.
Me: Increase. In. Spend. We’re going to take. A big. Risk. So you’re forecasting the budget, or, like, tell me, kind of like, how it works? Are you saying? Okay, great. Like, we need to spend this much, and that’s the budget. Or like, yeah. Like, what is the operational stuff? Going on?
Them: Yeah. No, that’s a good question. So in the past, the annual budget was. Was given to. Given to me by our executive team. So then I have to, you know, I was the one setting up the customer acquisition target and then distributing it based off of the the customer acquisition target. So that’s how paid has worked. Shivani and I and a couple other team members actually just had this conversation about non paid revenue.
Me: We just had. This conversation. About?
Them: The way that non paid revenue was forecasted is we were given an overall revenue target.
Me: Forecast. Ing is we were given. An open. Market.
Them: So we accurately forecasted paid ads, but then we still had a revenue target that we had to forecast to. So essentially what we had to do was, you know, I had to be as accurate as I can where I needed to be. But when you go back to some of these channels, like, you know, for, for, for Shopify if we had, we still had a gap between what my forecast was and what the target was, and that gap was a million dollars. I honestly just had to put in that million dollars in there to because in the past, I think executive team asked us to build. I keep and I’m not using finance terms, so. If there’s a better term, please use it.
Me: We still. Had a resident target. And we had a full. Price, too. So essentially what we. Had to do. Between. A5 forecast. Was. How. Many. Times. Use. It.
Them: Instead of building a forecast, it was really building a roadmap to the target.
Me: It’s been. A forecast. It was really building. A roadmap to the. Target. Yeah.
Them: And so the, like, the TLDR on that in my mind is that it’s not really forecasting. Exactly. It’s just target setting. That’s what I’m saying. Yeah, that’s what I’m saying. Building a roadmap to the target. It’s not really a forecast.
Me: I feel like you said I like.
Them: Yeah. I’m like. It’s like.
Me: Yeah. So this is awesome. Yeah, you’re exactly right. You’re setting a budget, and you basically are backing into the numbers.
Them: Yep.
Me: And so one thing that’s, that’s certainly like we, we do also is just, like, trying to understand, like, how accurate we are. We are, are we to forecast. You know, after setting it, and so, yeah, you’re totally right.
Them: And I think Dan is going to be the stakeholder who’s going to be, like, trying to figure out, like, okay, if I were to go from late of, like, how do I think a revenue forecast should be built?
Me: Pacing.
Them: Versus just let me set an OKR the back. It’s like, well, I think. And I mean, we’ve had some more recent conversations, so. By the way, I need to start by saying I know that was not the correct way to do it, and I’ve known that’s not the correct way. I know.
Me: My intention at all to say.
Them: I know, I know, I know.
Me: Understand? Yeah.
Them: That’s not at all feedback. That’s just like, oh, bad. Okay, we have we to be just more highlights. Like, ooh, there’s room to, like, make this better and, like, that.
Me: For example, Carlos, if you’re like, damn, this channel has great cac, we need to ratchet it. Okay, like, what happens? Like, how do things get adjusted? Like.
Them: I’ll walk you through the. The. Perfect. So we just actually had this happen this month, and. Yeah.
Me: This. Happen this. Month or the other way. Right. But like, how do you roll with the punches on a month to month? Basis.
Them: So I. My recommendation, actually, is we. We have to use one true North Star, right? Because we have. We have. I think there’s two different metrics we could. We can look at. Two metrics I think we need to look at. There needs to be an overall marketing efficiency, which there is already a metric, which is marketing efficiency ratio, right?
Me: There’s two. Different metrics. We can. Two metrics. I think we should look. At. The effic. Iency. Which then. We measured. Which is one of the inefficienc. Y ratio. Right.
Them: I think that is like the true guiding light for more or less spent, right?
Me: I think that. It’s like the. Truly. Guideline for what? We spend. Right.
Them: However, there’s also customer acquisition costs. And I think this is a a good example of why I built this, because it is going to actually solve that problem that you just mentioned where, you know, in the this last month, our total Ecom sales were actually down 5% month over month.
Me: So. Customer. This is. A good example. Of I. Built this because it. Is going to actually. Just. Metric. You know? Five. Percent on number. One.
Them: However, our performance targets were slightly below where we want it to be.
Me: How. We’re. Slightly doing.
Them: However, mer so performance targets are all based on attribution, right? And we know attribution is not perfect.
Me: This. We. Don’t have to be. Perfect.
Them: So when we look at that, we can take a look at Mer. And mer was actually up 6% month over month.
Me: So. When you look at that. We can take a look at. Any other.
Them: And this is where the, the ECOM team has conversations with the finance team and say, hey, our performance targets are, are below where we wanted, where we want them to be.
Me: Team. Has conversation with. The finance team and say, Hey.
Them: We know that. You know, attribution is not perfect.
Me: We. Haven’t seen our actuation. It’s not perfect.
Them: However, MER is actually quite a, it’s, it’s up quite a bit above target and we’re pacing to be below. Do we, do we feel comfortable continuing to push spend so we can hit what the revenue target is? And that’s just one example of the way that that works. And, and those are like the recent conversations that I’ve had with with Dan and Shivani is that that’s where I think we’re going to get to. And then, you know, for next year,
Me: How. Is. Actually. Target. And our patients. In england. Doing? Feel comfortable. Continuing to push. Back. So we can. Hit.
Them: This conversation just happened yesterday, so it’s very recent. What I’m doing is updating my 2026.
Me: The. 20. 26.
Them: Okrs or projections to actually update to, to be updated to what I think are actual projections are. And then at that point, what we do is we have a conversation. Okay, this is where we are. We’re up or down below what the target is. If we want to get there, we were going to need to potentially spend an additional 5 10%.
Me: To. Actually ident. Ity. N. Ifty. Exactly.
Them: And that’s. That’s the. That’s the right way to do it. It’s just. What? Just to give you some more context.
Me: Just. To give you some.
Them: This has been our business for the, since I’ve been here and I, I’ve been full time for close to, to four years and I’ve been working with the brand for like over five and a half years. Like, this has been the business. There has been no time or a need for anybody to sit down and input this together because, like, our business is growing. We need to continue to do what we’re doing. We have less than 20 team members. Like, nobody’s going to have the time to do this whenever a business is so successful. And I think now our business is, isn’t necessarily slowing down in terms of growth. Retail is up quite a bit and overall revenue is up.
Me: Time. For. Close to 40. Minutes. So. Successful. Our. Business isn’t necessarily.
Them: But things are getting more complex. We’re seeing a slowdown on E comm, and I think we’re. We’re potentially, like, taking our foot off the gas a little bit on spend because we’re not seeing attitude be where it’s been in the past. So I’m building this to be able to to solve for that problem of like, our business is getting more complex. Let’s try to get all the right reports and the right data for that.
Me: Because. You are launching new channels, right? That’s the thing. There’s, like, wholesale. There’s other new channels that are being launched, but similarly, across your ad channels, like, where do you start to reallocate where you see that, that efficiency. And then similarly, like on the forecast, on the forecasting side, like, how can you start to not only do pacing. But also start to add inputs, like in the middle of the year. And say, like, okay, yes. If I just take it and divide it by six. I can get there. By, like, okay, if we were to ratchet things up, like, what can we expect? You know, so there’s a lot. Of room.
Them: There’s still, like, I still applied in the previous year, still applied that seasonality curve. And again, for. For the paid side, it was accurately, it was a lot more accurate than the other side, than the other revenue. However, like, yes, I applied that seasonality curve where I, you know, I, I, I forecasted that we were going to have more revenue during May, August, but it was still all being forecasted with the goal of making sure that the year end matches what our target is.
Me: Yeah, exactly.
Them: Which it shouldn’t be, right? It should be projected. It should be your projection. And then you come in and look at your target compared to your target.
Me: Another. Yeah. You do projection versus target, and then you pace to see, like, what you’re gonna hit for sure.
Them: Yep. Right. Exactly.
Me: This is great. Like, this is, like. I mean, you’re pretty far down. Like, that’s usually where you want to be, where you can start to look at true projection, and then you can layer on all types of models to do better projections. You still can actually just do based on actuals pacing. And, like, it’s a really great way for you to tell a story and then to start tweaking things to see what happens.
Them: Yep. Yeah. Yeah. And this. Actually, by the way, I think our call is set for an hour. But I. I see on the in the doc that we have this for an hour and a half. I do have time to go over, but I know if you guys do.
Me: Separate. No. I just put an I. We could. We can leave an hour. And I think this is fine. I only have one more question about, like, like Geo. Like, I don’t see any dimensionality around, like. Like location.
Them: Yeah. So the reason I don’t have that is because I primarily. I mean, just for. For paid ads. It is all dynamic when it comes to jio, and it’s based off performance.
Me: When. It comes to. Performance.
Them: I mean, I do actually feel pretty strong about this, so I’m open. But I’m open to obviously, hearing your opinion.
Me: I’m all. Good.
Them: You can’t outsmart the the AD Consoles algorithm.
Me: You. Can.
Them: So for what I mean by that is I would never suggest to our media buyers that they look at like what states are performing better and only run campaigns in those states.
Me: So. What? States. Totally.
Them: The. The algorithm will identify that on its own. In Target where it’s seeing the best performance, so that’s why I don’t have any of that.
Me: Best. Performance. For me, it’s.
Them: Where?
Me: On the. Kind of on the sales side. And.
Them: Yeah, that’s what I was going to say. Yeah.
Me: Yeah.
Them: That. Like, now we’re. We’re getting into distribution, and we’re getting into retail.
Me: That’s actually. More. Why? It’s like you want the ecommerce. Sales to give you that sense of like, where is there geo concentration? You know, and. And that will indicate those channels where you do have to be a lot more specific about, like, where you’re placing things.
Them: Yeah, I think that, like what that is for me, it’s in addition to our ecommerce reports.
Me: Yeah.
Them: But yes, I, I definitely think you’re right that it is going to be more relevant now that we, we have a physical presence, specifically with distribution. We’re going to be, we’re going to, we’re being strategic with location so we’re going to have to run geo, you know, geo based campaigns. And in that is in the, it’s in the plan, but obviously it’s not, it hasn’t been solidified because we, we need to know what those locations are going to be. But yeah, absolutely. But to me, it’s separating our E Comm. Business and reporting from just overall business.
Me: Specifically. Because. To me. It’s separate. From.
Them: I think, I think that’s exactly right because eventually there will be geo questions we want to ask, like, and, and I think that will tie mostly to somebody you’ll eventually meet with them, which is Paul, and he’s thinking a lot. We actually have somebody else joining on the distribution side soon. But like, about like, which cities do we want to. And how is business there? And so you want to understand, like, how N A is doing versus N as an example. Right. But like, right now, we don’t, we don’t have that view. It. I think that’s okay. But that will, like, as we get the retail data. The spins data, all of that stuff in one flow. Then we’ll want to make sure, like where we don’t have a geo tag, that we’re able to develop one.
Me: Your. Country. That. I want to.
Them: And then we’ll figure out how are we dividing up the country to ways that makes sense to us. Yeah.
Me: Make. Sense to us. Yeah.
Them: And I, I do want to share one more thing. I know we’re maybe we can do it a different time, but there’s obviously a lot more inputs in, in these two sections. But I do actually want to get into a big issue that I ran into when I was actually doing end of month report versus the way that I originally set the forecast. So the way that I. I want to look at each individual channel is okay.
Me: I. Do want to share? These. Two sessions.
Them: And let’s try to always limit when we put a when we’re forecasting solely based on the revenue number. What I want to forecast. I want to forecast the inputs because revenue is the output. So I want to be able to build the forecast using the inputs as what we forecast. However, because we divided drink banks and revenue, the a lot of the times the inputs are going to be a V conversion rate and traffic.
Me: Inputs because. I. Think. Puts. Are going to.
Them: I don’t actually have visibility into just traffic in conversion rate.
Me: Just.
Them: I do have a V, but I don’t have visibility into conversion rate and traffic. I brought this up to Source medium, and. And they’re. They’re saying that’s part of, like, this custom thing that I. This custom plan that I was telling you about, but I would kind of quickly walk you through. Just know that that was like one more forecasting thing that I wanted to share with you is I try to always limit from us using revenue as the, as the input. It should always be the output. And like, I’ll walk you through a couple of examples here. So looking at the revenue, like the, the November target right. So essentially this is subscriptions. What I want to be the inputs here is the monthly. The forecasted estimated monthly ship rate.
Me: Be. Discuss. Ed. But. You do? Fore. Casting different. I trust. You. Have. The. Input. Into. A couple of. Examples. Here. Forecast. It estimates. To completely.
Them: Total subscription count and then a ov and obviously for that we can look at historicals in year, year over year trends. So those would be the inputs. There an area where the inputs are different are going to be for like for November again, some of these look, it’s going to be a V and then going. To be new customer revenue. And then right now I don’t have it because of the issues that we had. Traffic would be another input here and then the output would be, would be revenue. So some of these look different, but the way that I, I really want to get to is where for some of these channels, like, email, direct SEO. I want a conversion rate.
Me: Subs. Cription. Which. Are different. To. Be inform. Ing a free. Custom. Er. Email. N. Co. Email you do on statins or, like, opens.
Them: What’s that?
Me: Like, email you would do in, like, sends or, like, opens.
Them: I would. I would. We. I mean, yeah, either. Either one of those would be. I’m open to any of those being the inputs. These are the ones that I came up with.
Me: Would. Be. The general specific traffic equivalent is. Yeah.
Them: Yeah, exactly. But the problem is that what I do want to do is I want it to be consistent so I can’t use, like,
Me: Yeah. Consist. Ent. So I can’t. Yeah. Yeah. You need to either use the top most top funnel. Probably everywhere, so whatever.
Them: Exactly.
Me: Sends or impressions.
Them: Because. Yeah, because the problem I was running into is I’ll use meta as the example, right? So for meta, one of the inputs that I was using is, well, I can trust meta, you know, outbound click data and use that as the traffic.
Me: I. Can trust.
Them: However, we are using last click revenue as the as the, you know, the metric used to report. So that means I have to use sessions with from GA4. But within GA4 at this time, I. I don’t. I’m not able to separate meta traffic for drink mix in meta traffic for for sparkling. So what I ended up having to do is you use. I’ll use meta as the example. I had to go back to to using.
Me: But. This. Time. I’m not. Able to separate. Chris. Parking. So. The ex. Ample. I. Go back to. New.
Them: This. These are all the ones that are marked out right here. These were all the inputs that I previously had, and these are inputs. So, like, CPM is an input that our buyer controls. Right? Traffic, conversion rate, click through rate, Those were the inputs for the revenue to be the output. However, I don’t have a way to separate those right now.
Me: Site. Previously has. Sort of. Cpm. However. I don’t. Believe separate. Things. Right. Now.
Them: So I had to go back to using these as the inputs instead. But at least right now I’m not having to use revenue as the as the input, and I could still use some other inputs. But what I ideally would like to do here is like as a meta buyer, when you’re building a forecast, what are those individual metrics that you actually have influence over. And you don’t necessarily have an influence over revenue, you have an influence over click through rate and ncpms. And that’s what I want to be the inputs. So this kind of gives you an idea of how I look at forecasting and ideally My goal would be to. To make sure that we have the ability to have the inputs be things that we can control for those specific channels.
Me: They. Put somebody. When. You go to? See if you. Can go. Actually have the. Input. You don’t necess. Arily have an aut. O. So. This kind of gives you an idea. For speed. Things. That we can control. Specific. Checklists. Great. Great. Okay.
Them: Cool. I know we’re at time, so happy to set up another time to go through and, and go through some of the, the, the inputs, but I wanted to, you know, to see if this was, this was all helpful for you guys. If you have any, any sort of a high level questions.
Me: So. This was. This was. I feel like we went through.
Them: Yeah, it’s sort of information and, yeah, it’s really helpful. We would need access to all these platforms to, like, get more deeper into those and also to the new sheets. So we can get the metrics and the.
Me: The growth sheet. We don’t have access to Shivani, but we do have the a copy of the Ecom.
Them: Yeah.
Me: One. I guess we have the copy of the Ecom bro Dashboard. There was the other one that you shared.
Them: Yeah. You have the 2026 version? However, because that some things we just talked about, how we need to revisit that. And then also for the 2025, it’s, it’s more accurate because it actually has one month of reporting and it’s surface. Some of these things were like, okay, we’re not going to be able to get that data, so we have to. We have to get it a little bit differently, so I’m happy to share that.
Me: Because. They actually have. Record. Ed. It. Served us some of these things. We’re like, okay. We’re not going to be able to get. Back to him. So we have. To. We have to. Get some different. Ways.
Them: That sounds great. And then. And, like, if it’s. If it’s better, Carlos, like, I. I’ll send you a folder where I have stuff that I’m sharing with redforge. So if you wanted to share a copy of certain things and plop them into that folder, that’s like an. Easy way to do that, I would say. Or like. Or if you feel comfortable sharing the file with them directly, like, whatever. Whatever path makes sense.
Me: To feel comfortable. Sharing. Yeah. I really just mostly want to see the structure.
Them: Yeah. So I think, like, do a copy and put it into that.
Me: Yeah. Since we’re not, like, reconciling around nearly close to doing anything like that. Mainly just getting understanding the reporting structure.
Them: One. So, like, await. Let’s just follow up with Carlos with clear next steps over slack or something, and then we can make sure we get you everything that you need. Does that sound good?
Me: Yeah.
Them: And then I’ll just create a channel for the four of us just so that we could. Or a chat for the four of us. And then one real quick question, Carlos, that we’ll jump is just like. Do you feel with your row sets, like the rose sets, meaning all the rows that you have listed do? You feel like they’re organized in a way that you’re like, yes, this makes total sense to me. Or when you’re like, these are the inputs to revenue, and then revenue is at the top. And would you. Is that something that you’re like, I would love another set of eyes on this. I think the. The. The format I’m pretty happy with. Okay, cool. Are good there? Yeah. Like, when. When Utah was just talking about, like, email sends email. This, like, my brain just works in funnels, which is like, what are the. Like, what are the steps to get us to the number? And so, like, that’s. Why? I was just asking if that’s like. Yeah. Feel like.
Me: What are the. Steps to get us to the. Level. It’s. You can, in this view, it’s really channel based on the revenue side. But then you may start to look at like where you’re spending money and then the output revenue associated with marketing channels versus the sales channel. So that’s the kind of flexibility to where like whoever needs to look at it a certain way. You can do both, right? Yeah.
Them: Well, well, we’ll plant that seed. Utam Awaish will get those files that you need, and then why don’t we think about if it makes sense to do another touchpoint with Carlos cubby? Is that so good?
Me: And then. I’ll compile like our questions for source medium based on what we found today.
Them: Okay?
Me: You know, because this was helpful to understand. Carlos. Like what you’ve asked them.
Them: Perfect.
Me: In the past and sort of like how you worked with them. And then, yeah, we’re kind of. Gonna. We’re. We’re working with Shivani to drive towards, like. A source medium question versus outside of that and like, yeah.
Them: Awesome. Yeah.
Me: Yeah.
Them: It sounds like you guys have a joint Slack channel, so you probably have access to me on Slack as well. Happy to be in that channel if. But if it’s just more people, too many people, then just send me.
Me: No. It’s fine. It’s not noise for you. I don’t mind.
Them: I mean, at this point, I have so many to chat for the four of us, because I think clubs were just talking about getting access for, like, systems that are going to be beyond this. So let’s just create a chat for the four of us and if it makes sense to add more people right now, that chat Carlos is like, the two of the Brainforge guys, me and Jason, and it’s mostly about, like, prepping. Okay, so we’ll do it a separate one right now. I was just making sure that he had access to me. If you guys have a joint channel, you probably have access to everyone on an element.
Me: So. Let’s just expect. Some. More people. Have. Access to everyone. And are you doing the forecasting, Carlos? In. In another sheet.
Them: Doing so. What I’m going to be doing for 2026 is we have the. The one that you. You’ve seen.
Me: Is. Ufc.
Them: I’m going to duplicate that and update. So in the past, what I’ve always done is I have an actuals one, and then I create a workbook.
Me: Out. Put. Yeah.
Them: And then the workbook updates. As you know, I get some of the actual. So I’ll be having a separate workbook for that.
Me: Upd. Ates. I was given. I guess some of the actual. So. Rry.
Them: Cool. Thank you so much, Carlos. Appreciate.
Me: Thank you so. Much. Thank you. This is great. Sorry, I know this is really, really dense, but this is great.
Them: You’re good. No, I mean, this is. This is great for me too. Like, I. I mean, I honestly just went through all of this, and you see, there’s a lot of stuff in there, and it doesn’t. They were not her to have a, you know, a separate pair of eyes in there, because I. I. I had. I basically built this from scratch. And as you can imagine, it took a while. And I know there’s going to be a couple of things that pop up and questions that pop up that are going to be helpful.
Me: This is. Great for you. Too. They. Just went through. This. From scratch. You know? There’s going to be a couple of. Things that. You want to see that? I think it’s impressive that you’re managing it like it’s a lot. There’s a lot.
Them: Yeah. Awesome. Thanks, guys. Thank you. Okay, I’ll see you on the next one.
Me: Thank you. Perfect.
Them: Okay, bye.
Me: Thank you.