Meeting Title: Uttam <> Jie Date: 2024-02-19 Meeting participants: Unknown
WEBVTT
1 00:00:00.000 ⇒ 00:00:03.149 Hey, John, how’s it going? How are you.
2 00:00:04.580 ⇒ 00:00:20.700 hey? Sorry? 1 Si think my speakers go in the wrong place to fly back to Lima to be at sea level. Yeah. So I just checked into this hotel. Not. I’m not feeling perfectly well yet, but not been fun.
3 00:00:20.810 ⇒ 00:00:50.100 Well, hopefully. This meeting is interesting and hopefully takes keeps your mind occupied for a little bit. So yeah, I’m excited. I’m excited. I thank you for putting together a big jam as well. Hopefully, it’ll help clarify my thoughts. I feel like I always flip flop between optimism and pessimism. And like, one moment I’m like, yeah, this could work in the next minute, I’m like, Oh, my God, I don’t know. Hopefully, like my job here is like as a sounding board. But also.
4 00:00:50.190 ⇒ 00:01:12.839 again, I’ve been through the process of not only like understanding the goals. This product see kind of getting a sense of what levers you have, not only from the product side, but like 50 million dollars of annual revenue usually don’t have that. So that’s kind of the the segment that we’re going to go after. And so for the Cfo. We want there to be a dashboard for them to
5 00:01:12.920 ⇒ 00:01:32.739 you know we we show them a few of the things that we have built so far right? And I think there’s 2 things that resonates with them. One is, they want something that they can show to external parties, not internal parties for external parties, just a couple of questions. So
6 00:01:32.830 ⇒ 00:01:55.860 on the marketing side. So I guess one would be helpful to understand. Data is difficult. On the spend data. We have quick books, right and and the quick books. And that suite data gets joined together and unified. So there is, you know, balance sheet. And Pml, let’s focus on the Pml.
7 00:01:55.880 ⇒ 00:02:07.409 When we look at the Pml. There is de different levels of tagging right? And so be like first tagged, as you know, above the line below the line income versus expense. And then there’s
8 00:02:07.440 ⇒ 00:02:26.030 operating income operating expense within expenses. There is cost of good sold and blah blah blah right? And so the there is some kind of Ml there that’s trying to look at the line item, description, and then trying to tag that against what they, what sort of the taxonomy of
9 00:02:26.440 ⇒ 00:02:44.170 if you want to see what the role shape of that data is and all that kind of stuff. There, I can. I can. I can show you. Yeah, I’m familiar with with all those sources, and like, kind of what you get out of it. That’s a loss that’s a loss.
10 00:02:44.180 ⇒ 00:02:58.009 You’re not gonna be able to do much there. The nice thing on shopify is shopify has a like a table where you can see interactions. And then that’s linked to the customer. So you can get the refer
11 00:02:58.030 ⇒ 00:03:11.899 but at minimum you can at least see how many customers came in and what they bought right. And the revenue resulting from there average order, value all that stuff. And it’s not where we’re gonna be when we finish, obviously. But
12 00:03:12.690 ⇒ 00:03:24.659 so if you click on insights on all and our platform right now, you’ll see this. It’s like a kind of Sh hmm.
13 00:03:24.790 ⇒ 00:03:51.519 random stuff. So if you click on industry insides, you’ll get this. Industry insights Page. There is people in the industry that calls the aim or marketing efficiency, ratio or efficiency ratio, right? And so based on that you can sort of have this like kind of like an Excel model where you just every single possible point along this
14 00:03:51.540 ⇒ 00:04:02.200 event horizon, or what’s possible. And then we can map out what’s what you expected. New customer revenue is at each point, and then you can
15 00:04:02.510 ⇒ 00:04:24.900 given a cost of delivery and con you’ll get that contribution margin. yeah, this grand. It’s gonna show. Not a but it will show something. Once I we we link stuff and get it in properly. We haven’t. We haven’t, because there’s some relationship between your
16 00:04:24.900 ⇒ 00:04:46.340 already spend and your new revenue. So that’s gonna drive the other piece of that. Now, between those 2 things, you have the beginnings of a model that will tell you your top line right? And then what we’re gonna do is we’re going to build a Pml right? And what I call this is going to. It’s going to be a
17 00:04:46.340 ⇒ 00:05:06.500 color coded. Right? So, for example, your if your non operating expenses are 50 of your top line revenue. Then you maybe you’re probably overspending on non operating expenses. You may want to cut it down right? And so we’ll rank that up against everyone else in your industry as well. Right?
18 00:05:06.650 ⇒ 00:05:18.919 That’s one thing that’s like that’s furthering the benchmark layer that that marking a layer of the pnl that normally would not see when you look at your own. Pnl,
19 00:05:19.160 ⇒ 00:05:43.499 yeah, those 6, those 6 like kind of like products or features. So do you think this like, how do you think this page evolves? Does this become like a consolidation? Does this become like a couple more boxes like, what what do you? How do you think this current, like the way currently exists compared to where it’s gonna be?
20 00:05:43.670 ⇒ 00:06:11.729 Come in. You have no idea what’s going on. You don’t know what this product is first thing we’re gonna show you is the benchmarks. The benchmarks is gonna give you a sense that of the power of our platform, because, you know, before, you would just see numbers. And now you see numbers that’s ranked against like other numbers, right? And so the you’re in the 9 is the percentile. And then you can also choose right? Like, you know, you wanna look at the past 6 months.
21 00:06:11.830 ⇒ 00:06:33.540 or you want to look at the past 12 month average. And you would like average all out, you know. So and then you can look at different industries and stuff. So that’s the first thing you look at right? So now you have a sense of where you are and where the rest of industry is. But that’s like a backward looking thing right on when the product needed to get out.
22 00:06:33.600 ⇒ 00:06:49.280 And so one reason I went with looker initially is because I could have data people be building the product, which means I don’t have to wait for front end and design. If we want to push out a new dashboard.
23 00:06:49.370 ⇒ 00:07:08.159 I have someone who’s an analyst pretty much build a dashboard. So the key concern there was, I was like, Hey, we have 3 months to get something out. We price those differently. And but the actual cost of goods to manufacture those 2 are quite comparable. So it kind of depends on like, are you gonna just have one
24 00:07:08.180 ⇒ 00:07:20.619 product suite offering for all customers, and that could change over time. The second thing is is timeline, and how often you want to make. And and you may find that like, okay, we want some really specific visualizations.
25 00:07:20.640 ⇒ 00:07:39.719 That’s when I would suggest doing it. But the the development time gains. You’re gonna have a thought spot. I expect will be very, very high compared to. you know, building on your own. But you’re right. You won’t be able to do that specific pnl, benchmarking design that’s like, again, looks awesome. That’s something that probably
26 00:07:39.740 ⇒ 00:07:55.729 can probably build a little bit, Jakey and thought Spot, but ultimately should probably built to engage our existing customers further, like, What’s the roadmap for this product like? Is this a way for us? Or, again, as you mentioned the beginning of the call, we just want features that
27 00:07:55.900 ⇒ 00:08:12.630 are more like what you would see in a typical software product. So having a data product like this. cause, I wanted, how do I? How do we measure that this thing is is working right? Do we want to see, like, okay, cool out of, like, our out of the 500 customers that have access to this. 50% are logging in once a week.
28 00:08:12.750 ⇒ 00:08:25.409 Okay? But they’re not only logging in. They’re actually uploading new data. right? Or they’re they’re getting delivered data out right? And so thinking of not only how are they using the data in the platform, but I want to see people
29 00:08:25.590 ⇒ 00:08:43.569 like getting scheduled deliveries. exporting stuff. I wanna see people screenshotting and and asking support questions about about the data. What things mean? Those are all things that show that. Okay, this data is really compelling and allow you to get that engagement. And the last thing is like.
30 00:08:43.740 ⇒ 00:09:12.230 I’m after at at working at my other company like my biggest thing was, how fast can you make iterations? You’re gonna get a whole host of feedback when you put this in front of people. They’re not gonna be able to wait a month. The product, though, to give you to give you my honest opinion, I think the things that you’re able to show and showing the Pnl view is something that I’m I haven’t seen anywhere. I think that sort of stuff you would only expect to get in a quick book or netsuite, but because you have.
31 00:09:12.290 ⇒ 00:09:23.260 like both the spend items, any other revenue side, I think that’s an awesome product. I think I want to understand what kind of like the goal is.
32 00:09:23.310 ⇒ 00:09:38.340 And then understand? Like. yeah, II it would be, I think if you’ve gotten good customer feedback, I mean, I like what I really love what the product looks like. I gotta be. I’m gonna let you in on some secret. II kind of cheated a little bit here.
33 00:09:38.410 ⇒ 00:09:55.840 Reason why I’m cheated is because we we have a customer that we fed through a process by Ctc Ctc common. Threed collective is a company that does mostly like. You know, it’s it’s like an agency, right? And so they’ll have like
34 00:09:56.120 ⇒ 00:10:09.010 big accounts. They just I feel like I’ve heard of them. Yeah, I’ve heard of common thread. Yeah. So this is a lot of the stuff that I found through their content online. Anyway, we paid
35 00:10:09.020 ⇒ 00:10:22.800 for one of our customers to go through this process and in your hand and go talk to the Cdc. They’ll charge you a lot, but you know they’re good. What they do. You know, we’ll go talk to them. We’ll probably do a partnership with them.
36 00:10:22.870 ⇒ 00:10:35.950 But you know the baseline, at least we can confirm that we are providing some level of value here. Right? Cause. I think there is a lot of value in delivering this kind of use that is.
37 00:10:35.980 ⇒ 00:10:56.169 focus on the Cfo not focus on the marketing guy or some other guy. The Cfo. Who’s trying to get a hold of their business and trying to figure out, am I spending too much? Am I spending too little right like. That’s the key thing that they’re like trying to figure out. And they just don’t know what where everything is.
38 00:10:56.300 ⇒ 00:11:13.400 I think you guys can pull this off and hearing that makes me really happy, because I think you can actually charge a lot for this product. depending on how you want to do it. I think you can charge us even a stand alone as an option, and you’ll get people paying for this because
39 00:11:13.530 ⇒ 00:11:25.080 people always, especially on they want to prove that their efforts are working out for them. To do. This level of analysis costs a lot of money, and you can take the common thread collective, as maybe like the minimum it would cost.
40 00:11:25.170 ⇒ 00:11:41.699 and the maximum is like, if you were to hire someone internally to come. Do it. So I think if you can think of like, what’s the smallest portion of this that you could get out the fastest and get it out there. I think, like, I really like what the product looks like. I like that you didn’t go and do
41 00:11:41.930 ⇒ 00:12:00.349 something that’s totally left field. I’d like that. It’s stuck to actually the Cfo. There’s a really clear persona, and it’s a persona that you already have in contact with. And you already know that there’s a market for a product similar. That’s like, you know, 30 K. For 3 months. If you’re able to do this sitting on the same data and you’re able to.
42 00:12:00.440 ⇒ 00:12:24.329 I know again the one thing I’m concerned with is, how long does it take to build and how fast you can iterate after you put it out there? But that would that II like what it looks like. And I think the big things to understand is, I wanna understand the shareability? Are they able to export stuff? But I’m not worried about that stuff like I feel like once they start asking for that. Then I then I know I have it. You know III
43 00:12:24.370 ⇒ 00:12:50.040 is. The deliverability stuff is is is more standard and and clear cut for me. Yes, for me the difficult part is like, you know, all these companies. They don’t do their quick books properly, or everything’s labeled right. And then they they combine different thing line items into a single line item. And then, you know, they don’t good enough.
44 00:12:50.220 ⇒ 00:13:05.440 If it’s not. it’s not enough that that’s the problem and the quick book stuff. At least, it’s like your you’re. You have access to what they have as like the source of truth. Pnl, you can say, Hey, this is using your pnl
45 00:13:05.710 ⇒ 00:13:32.320 if you see data in accuracies. You have a little bit of a caveat of like, hey? We’re we’re only as good as the data you give us. Oh, okay. So now it’s a game of us having to classify those segments like, create these buckets. Nice thing is like you. Probably all. You probably have some sort of customer segmentation stuff already done that they’re gonna eat the benchmarks up, and especially if you make it again, as you said, nice way of screenshotting or really quick way to export.
46 00:13:32.360 ⇒ 00:13:51.599 To Pdf. the benchmark product is like a very compelling product across a lot of different, like a lot of different analytics, customer facing analytics, the benchmarking tool, especially because you guys are the hub for all this data.
47 00:13:51.800 ⇒ 00:14:05.579 yeah, I appreciate it. Thank thank you. affirmations. Good. Okay, so I have some questions for you, and, like, you know, in the time that we do have well, firstly, I’m not sure if you
48 00:14:05.580 ⇒ 00:14:30.050 add some kind of thing you wanted to go through with the fig jam and stuff, or is it not worth going through? Or II mean, I’ve just been taking notes. I’m gonna make some updates to that based on like our conversation. So let’s it does big jam as well in terms of the data sources that we do have. So okay, I see it now. Yeah, I made mainly is like, I wanted to see whether this was gonna be. So. So let me give you. Let me give you a sense of how I think about this problem.
49 00:14:30.050 ⇒ 00:15:01.839 As I just face this problem with another client that I’m working for. So I’ve done the entire stack of data from Etl all the way to like analysis and dashboarding like. But what happens is everybody’s good, something dashboarding. You never need someone who’s like a full time database engineer or dashboarder, right? Typically, it’s a data person like me who’s like that’s the last thing they do is like, go and build a dashboard. However, that’s the main consumable. How to translate.
50 00:15:01.980 ⇒ 00:15:29.260 you know, line charts, bar charts that you know that whole art of that work. The problem is the the reason what I found is like, that’s not a full time job like you need that person to come in, make a wire from a dashboard, get it signed off, go and build it. And I don’t think you need that person. Decisions based on like database best practices, and come up with a dashboard that flows nicely and is designed well, and so what I, what I have him doing at my company is.
51 00:15:29.400 ⇒ 00:15:43.120 I have a couple of clients that require dashboarding. I’m like you handle all the dashboarding last mile delivery. Because if you ask me, I’m just gonna put line charts and bar charts up, but those are the actual deliverables for those clients. And so what we’re going through is an exercise is like.
52 00:15:43.460 ⇒ 00:16:07.259 for example, what are the key dashboards? We wanna make? What are their goals? What are the sections? Wireframe it out, and then he’s building them. That’s that I agree is is a lot. It’s like cost way less, and you don’t require that many hours. But there is an art to that and so that’s kind of like, what I went to the market and tried to find is that he can start from scratch and be like.
53 00:16:07.580 ⇒ 00:16:18.659 Yeah, here’s the goal, this dashboard. Here’s the kind of options I have. Here’s the flow of this dashboard product. It’s not as expensive as my work. It’s being these types of dashboards.
54 00:16:18.730 ⇒ 00:16:33.190 You don’t have the. You don’t really have the amount of. I don’t know if you’re writing this yourself in Javascript. You can do everything he wants. But maybe you’re not. Maybe you’re just doing it in a bi tool. And so you’re very much limited
55 00:16:33.300 ⇒ 00:16:52.870 to the bi tools whims. Right? but there’s still something about like the amount of white space. How you do comparison thoughts has a lot of flexibility. They have a lot of different types of charts. The main thing is like you would contract through him to him through brain forage. He has availability. Why don’t we
56 00:16:53.750 ⇒ 00:17:05.299 get on a call like for 15 min or like. And then what do you think about just throwing him at like one thoughts about dashboard for like an amount of time? And I can. I can.
57 00:17:05.599 ⇒ 00:17:28.439 running through a little bit of what we know. We could put some constraints on it in that call, and then just be like, is that? Yeah, II pulled it up. I was just literally thinking about that. I’m like, I have some feedback, but we should just take that and say, make a copy of this engineer. He’s not a visual guy, so it’s got we should do is like, make a copy of this.
58 00:17:28.500 ⇒ 00:17:54.559 and then go at it like you know, and II would give him access to it because I what I what I’m more, what I’m more interested in is not for him to give me a a mock up that I would then have to implement. I just want to implement meaning like, make a copy like, literally, just make a copy and edit it for this. Spend some
59 00:17:56.050 ⇒ 00:18:09.770 I’ll be like, yeah, just do edit. And then maybe we just I’ll say like, Hey, take this right now. We’re he’s working on another client for 120
60 00:18:10.090 ⇒ 00:18:24.829 but I think we can try to bring him on. I’m just gonna call and make sure he has availability. Maybe I send you a note and then let’s just let’s just throw him in on one of these. I think he’s gonna crush it. You’re either you. It’s a couple of ways, one, you can create an input.
61 00:18:24.870 ⇒ 00:18:47.570 feel that allows people to enter in those fixed budgets, if you’re like, Hey, just send us a Csv file format. In this way. You have some sort of etl that grabs that via gives that via 5 train. You could have it. That just be a direct upload that then does the same flow goes through Dbt and gets uploaded. Those are the easiest ways, the complication, if it’s like
62 00:18:48.000 ⇒ 00:19:02.959 I’m just, I’m just what I’m seeing is this like a month and a budget column, if it’s like more complicated. If it’s more rows, if it’s more columns, then it gets a little bit complicated. But you could just ask them to input, that data, you can set up a simple input form or
63 00:19:02.990 ⇒ 00:19:11.220 have them upload a Csv. The the only complication is there may be some time it takes the Etl, that bring that into the Dvt model.
64 00:19:12.950 ⇒ 00:19:29.120 I don’t. Okay. So with thought Spot, there is a way for thoughtspot to directly pull. Read data from a Google sheet. Right? Okay? Great. Oh, yeah, that’s the part I was like.
65 00:19:29.310 ⇒ 00:19:49.640 And you can have a formula that ties those 2 things together. And then you they can pull it again. so I don’t know if you’ve seen anything like it seems kind of like hacky. It seems like we’re just hacking it, but if it works, it works. But I don’t know if there is a more elegant
66 00:19:49.720 ⇒ 00:20:01.750 way to do this kind of stuff in what you’ve seen before. The the most elegant way is to have a very clear like file ingestion, or like a data input form
67 00:20:01.810 ⇒ 00:20:17.920 that then becomes product data that then goes and incorporate itself into the model. So you would have like a page that’s just like and put in all the stuff or edit previously inputted. Yeah, I look through a bunch of the docs. And you know, Hub Bot has a bunch of different like embed things like, Are you guys, gonna
68 00:20:18.030 ⇒ 00:20:52.249 you are, you guys thinking of also pushing out the ability to do the search like the I think you think. Think of. Think of the one pushing out the one thing that’s just like requires lease, interaction. It’s just pure consumption. And then, as people start to use that, and that gets scheduled, then you could start to layer on. So that’s what we did. We built just one basic read, only dashboard. People then were like, Oh, I love this like we would schedule meetings to do onboarding and things like that. And then they said, Oh, I wish I had this feature. Oh, I wish I could do exports. Okay, cool, then build a feature list. Then we can start saying, Hey, you give us this data.
69 00:20:52.400 ⇒ 00:21:13.509 Would you expect this out? Would you be able to interest in that cool right? So start with just a giving that value away. Bear start to use it from the thought spot. Analytics, you can tell who’s scheduling? Who’s accessing. Who’s viewing that page? I’m gonna let me know about your friend. Well, we can schedule time for me to give him download, and then I can. Also.
70 00:21:13.680 ⇒ 00:21:25.959 I think, once we’re in a good spot, I can also try and give you access to some of the data. So you can have Po, take a poke around just to see if you have any other thoughts and stuff.
71 00:21:25.970 ⇒ 00:21:34.959 But yeah, II think I have a pretty clear mind of like what I want to do. It’s just a matter of am I doing it right? And.