Meeting Title: Brainforge Signal Combinations Sync Date: 2025-08-13 Meeting participants: Uttam Kumaran, Giselle Agot, Mustafa Raja, levkatreczko, Henry Zhao, Ryan DeForest
WEBVTT
1 00:00:55.200 ⇒ 00:00:57.000 Uttam Kumaran: Hey guys.
2 00:01:04.930 ⇒ 00:01:05.920 Uttam Kumaran: Hey, love.
3 00:01:06.200 ⇒ 00:01:07.820 Uttam Kumaran: Hey guys, what’s going on?
4 00:01:08.310 ⇒ 00:01:08.950 Uttam Kumaran: Hmm.
5 00:01:20.160 ⇒ 00:01:21.790 levkatreczko: Sam and Ryan a shout.
6 00:01:21.790 ⇒ 00:01:22.300 Uttam Kumaran: Yes.
7 00:01:22.400 ⇒ 00:01:24.070 levkatreczko: Lee’s gonna join us today.
8 00:01:38.360 ⇒ 00:01:39.400 levkatreczko: How you guys doing?
9 00:01:40.600 ⇒ 00:01:41.670 Uttam Kumaran: Good.
10 00:01:42.050 ⇒ 00:01:43.300 Uttam Kumaran: How about you?
11 00:01:43.900 ⇒ 00:01:44.739 levkatreczko: Pretty good.
12 00:01:45.770 ⇒ 00:01:47.879 levkatreczko: Pretty good, pretty busy over here.
13 00:01:48.160 ⇒ 00:01:50.750 levkatreczko: In the process of hiring a couple BDRs.
14 00:01:51.530 ⇒ 00:01:53.319 Uttam Kumaran: Nice. Chasing kids around.
15 00:01:53.320 ⇒ 00:01:54.910 levkatreczko: Having fun there.
16 00:01:54.910 ⇒ 00:02:04.059 Uttam Kumaran: pretty… it’s pretty busy here, too. We’re… we’re growing, our team is growing, but of course, like, we’re… we have to keep our client,
17 00:02:04.150 ⇒ 00:02:17.420 Uttam Kumaran: like, our client experiences to be stable while we’re building the company, too, which is always an interesting challenge, you know? We’re not a SaaS tool, right? So, it’s, like, it’s a lot more of people, process, like, the way we communicate.
18 00:02:17.760 ⇒ 00:02:21.410 Uttam Kumaran: … So, interesting challenge, but…
19 00:02:21.650 ⇒ 00:02:27.229 Uttam Kumaran: It’s good, it’s still, like, the work is great, like, you know, we’re able to do this, like, we… it’s sort of like…
20 00:02:27.560 ⇒ 00:02:45.669 Uttam Kumaran: running a team on, like, hard mode, because we’re managing, like, several different clients, but if we were just, like, doing this for one person, we would never get to use all these, like, interesting tools, like, we’re using Clay for your stuff, for some other folks, we’re using NAN, we’re building, like, RAG systems for other people, so we’re… as a…
21 00:02:45.670 ⇒ 00:02:49.170 Uttam Kumaran: And that’s just on the AI side, so we’re able to do a lot of really cool things.
22 00:02:50.070 ⇒ 00:02:50.849 levkatreczko: That’s awesome.
23 00:02:51.600 ⇒ 00:02:52.550 Uttam Kumaran: Yeah.
24 00:02:53.250 ⇒ 00:02:57.879 levkatreczko: Yeah, I gotta get my hands on NAN at some point. It’s a future project.
25 00:02:58.680 ⇒ 00:03:00.010 Uttam Kumaran: Yeah.
26 00:03:07.760 ⇒ 00:03:11.019 Uttam Kumaran: So… Okay, maybe we’ll just, …
27 00:03:12.110 ⇒ 00:03:18.389 Uttam Kumaran: I can just get started, we’ll just share stuff and start getting your feedback on things, so… let me…
28 00:03:19.110 ⇒ 00:03:21.519 Uttam Kumaran: Share my screen…
29 00:03:27.280 ⇒ 00:03:28.270 Uttam Kumaran: Maybe not see this.
30 00:03:33.050 ⇒ 00:03:34.520 Ryan DeForest: Yo.
31 00:03:34.520 ⇒ 00:03:35.650 Uttam Kumaran: Yo.
32 00:03:38.200 ⇒ 00:03:39.600 Uttam Kumaran: Okay.
33 00:03:40.220 ⇒ 00:03:46.790 Uttam Kumaran: So, just wanted to kind of give, some updates on what we talked about the last week.
34 00:03:46.910 ⇒ 00:03:54.020 Uttam Kumaran: So… I’m gonna kind of jump a little bit between Notion and Play, but …
35 00:03:54.220 ⇒ 00:03:59.129 Uttam Kumaran: Couple of things. So, last time we talked about, as kind of a, again, a little bit of a…
36 00:03:59.510 ⇒ 00:04:16.660 Uttam Kumaran: reminder, we’re kind of working our way through these signal combinations. We’re kind of now in progress on these top three. There is some overlap, right? Because some of the combinations use the same signals, but I just want to walk through, like, how far we are on these, and…
37 00:04:16.760 ⇒ 00:04:25.200 Uttam Kumaran: So where we expect to get this week. So if we just talk about the first one, so last week we talked about a couple things. One is,
38 00:04:25.550 ⇒ 00:04:40.660 Uttam Kumaran: we were looking at, hey, what tools are on their site, and we think we had more than just the go-to-market tools. We had a certain… like, we had a list of all sorts of tools, so we just refined, that, so we’re just getting the critical go-to-market related tools.
39 00:04:40.660 ⇒ 00:04:50.490 Uttam Kumaran: So it’s a lot cleaner. For example, if we look at this example, we have, both Salesforce, HubSpot, Outreach, Pipedrive, we have a couple things on enrichment.
40 00:04:50.510 ⇒ 00:04:59.140 Uttam Kumaran: a couple of them things on orchestration, so it’s a lot more narrow now. We can continue to narrow this down, but I think this is a better spot. …
41 00:04:59.510 ⇒ 00:05:19.050 Uttam Kumaran: And so, the other thing is we can also get this as a list. So, this is the list that we’ll end up pushing into Salesforce. It’s just the list of the tools that are relevant. So, when you’re interacting with a lead, you can say, hey, I notice you have all these. It’s a lot shorter list than what was previously there. We also sort of…
42 00:05:19.140 ⇒ 00:05:34.499 Uttam Kumaran: kind of, like, wrapped up, I think. I sort of wanted us to move on from the job stuff, but we actually got quite a bit further. We are now using a little bit of a waterfall, but also, we are using Klagent to do a couple things. So, kind of the flow right now is
43 00:05:34.770 ⇒ 00:05:48.029 Uttam Kumaran: One, we have an agent that actually goes in, looks at their jobs or careers, directly on their site, and scrapes all of them. It’s usually the source of truth. Second, we use their stack, and we have one other enrichment source
44 00:05:48.160 ⇒ 00:06:07.730 Uttam Kumaran: That we’re using to pull jobs, so at the end of the day, I think we’re gonna… roughly, like, 80% accurate on open jobs that they’re gonna have, which I’m pretty comfortable with. We can continue to spin our wheels on this, but I think I’m comfortable just to identify, do they have open job opening… do they have job openings across
45 00:06:07.870 ⇒ 00:06:10.520 Uttam Kumaran: Specific, you know, sectors that we’re interested in.
46 00:06:10.860 ⇒ 00:06:13.939 Uttam Kumaran: We also have, like, the job URL and everything. Yeah, go ahead.
47 00:06:14.210 ⇒ 00:06:19.939 levkatreczko: Yeah, I just wanted to jump in with a quick question. I remember on the last call, you guys had an agent, I think, hooked up
48 00:06:20.510 ⇒ 00:06:36.489 levkatreczko: to their stack via the HTTP API, and I was curious, like, I would assume that that’s because you guys have a key there. Is their stack, like, natively integrated in Clay, or is that, like, would that be a play that we would need
49 00:06:37.350 ⇒ 00:06:38.890 levkatreczko: the key to run.
50 00:06:39.380 ⇒ 00:06:42.280 Uttam Kumaran: Yeah, Mustafa, do you wanna explain? Yeah.
51 00:06:43.110 ⇒ 00:06:51.000 Mustafa Raja: Yeah, we have a key. The stack doesn’t, natively integrate with Play, so it’s going to be a HTTP call.
52 00:06:51.560 ⇒ 00:06:55.490 levkatreczko: Okay, got it. So we’ll have to… Sad about that, Ryan.
53 00:06:56.570 ⇒ 00:06:57.190 Uttam Kumaran: Yeah.
54 00:06:57.380 ⇒ 00:07:01.509 Uttam Kumaran: So I think what we’re… what we’re also gonna do is, and Mustafa actually
55 00:07:01.660 ⇒ 00:07:05.839 Uttam Kumaran: sent this to me, I just… probably… maybe you can put it at the end here, is the cost…
56 00:07:05.980 ⇒ 00:07:10.690 Uttam Kumaran: So one thing that we’re gonna start to map is, like, what does it cost to run this entire workbook?
57 00:07:11.000 ⇒ 00:07:22.579 Uttam Kumaran: Like, basically per lead. And then, at the moment, my goal is, like, get it accurate, and then there’s gonna be several different ways we can get the cost down. We’re not talking, like, a dollar.
58 00:07:22.690 ⇒ 00:07:28.980 Uttam Kumaran: per run, right? It’s a lot less, but I want that to be something we know so that we can make those trade-offs, because
59 00:07:29.290 ⇒ 00:07:41.410 Uttam Kumaran: Nothing in life is free, especially some of these API calls, so I want to make sure that… my directive was get it right with as many sources as we can. We’ll share what the cost is, and then we can say, okay, actually, we don’t need
60 00:07:41.580 ⇒ 00:07:49.200 Uttam Kumaran: we can, like, deal with, like, some inaccuracy in jobs, but we also… we always want the form stuff to work, and so that’s a conversation we can have.
61 00:07:49.420 ⇒ 00:07:50.320 Uttam Kumaran: …
62 00:07:50.770 ⇒ 00:07:55.480 Uttam Kumaran: But yeah, we can… we’re using their stack, and then what was the other one, Mustafa, that we’re using?
63 00:07:56.650 ⇒ 00:07:59.770 Mustafa Raja: We did… we also did Core Signal, ….
64 00:07:59.770 ⇒ 00:08:00.310 Uttam Kumaran: Yeah.
65 00:08:00.310 ⇒ 00:08:04.740 Mustafa Raja: And then we have a waterfall that comes, integrated within Clay.
66 00:08:05.000 ⇒ 00:08:22.920 Mustafa Raja: So this waterfall only runs if our clay agent is unable to find a good, good number of jobs. If it’s pretty confident in, in what it has found, then our water flow for that particular lead isn’t going to run.
67 00:08:23.810 ⇒ 00:08:31.400 Uttam Kumaran: Yeah, like, for example, for Cortex, it was actually able to go in, find the jobs tab, click into it, scrape all of it.
68 00:08:31.520 ⇒ 00:08:42.510 Uttam Kumaran: and it’s, like, great, I have all the job openings. That is, like, a much better source of truth than having to stitch together from 5, but for example, for Tab’s platform,
69 00:08:42.820 ⇒ 00:08:53.499 Uttam Kumaran: it wasn’t as confident, it found a couple, but then we may… if it’s not as confident, we can run the conditional waterfall. So we’re already doing a little bit of optimization, but long story short, I think we’re, like.
70 00:08:53.860 ⇒ 00:08:56.530 Uttam Kumaran: getting jobs where I got a good place.
71 00:08:56.860 ⇒ 00:08:57.930 Uttam Kumaran: …
72 00:08:58.500 ⇒ 00:09:04.170 Uttam Kumaran: In terms of the other thing, so, we also started to narrow down, like, how we’re doing scoring.
73 00:09:04.270 ⇒ 00:09:11.760 Uttam Kumaran: … like, again, I wanted to make it very simple, like, All lead scoring, and …
74 00:09:11.930 ⇒ 00:09:23.380 Uttam Kumaran: category scoring, you can get, like, very complex for basically, like, no gains on. So I… I was like, let’s just do 1 through 5, and we can basically say, what is the overlap score? Like.
75 00:09:23.820 ⇒ 00:09:41.639 Uttam Kumaran: And it’s based on a couple of categories, like, if there are two for a given tool category. I think this is something that we’ll continue to refine, but I think 1 through 5 is fair. I would even go if you want to do 1, 2, 3, or if, like, Ryan, you could tell me if we want this to be
76 00:09:41.760 ⇒ 00:09:45.639 Uttam Kumaran: anything different, but I felt like it’s a good place to start that isn’t, like.
77 00:09:45.950 ⇒ 00:09:49.140 Uttam Kumaran: Previously, we were at, like, 1 through 100. It was kind of arbitrary.
78 00:09:53.080 ⇒ 00:09:53.760 Uttam Kumaran: Nope.
79 00:09:54.020 ⇒ 00:09:55.000 Ryan DeForest: Nope, I like it.
80 00:09:55.340 ⇒ 00:09:57.670 Uttam Kumaran: Okay, cool. …
81 00:09:58.040 ⇒ 00:10:06.100 Uttam Kumaran: Great, and then the other thing that’s kind of work in progress, I think we’re gonna do a little bit of a technical spike on the form fills.
82 00:10:06.230 ⇒ 00:10:25.290 Uttam Kumaran: Form filling and understanding the HTML at every single stage of the form filling process is a little bit more nuanced. I would say we’re, like, 50% there, but we want to do a couple things, Lev, that, like, we walked through last time, which was make sure we get the HTML exports at each stage. …
83 00:10:25.340 ⇒ 00:10:32.840 Uttam Kumaran: We also probably just need to do, like, a little bit deeper dive into, like, what are the nuances of form filling, …
84 00:10:33.060 ⇒ 00:10:39.179 Uttam Kumaran: So it’s just a little bit… it’s gonna be a little bit tougher than just telling a Klagen to go grab this stuff, is kind of what I’m saying.
85 00:10:39.500 ⇒ 00:10:45.680 Uttam Kumaran: But I don’t think this is… Mustafa and I talked about this today, I think we probably just need another day or two on this.
86 00:10:45.850 ⇒ 00:10:46.660 Uttam Kumaran: …
87 00:10:46.800 ⇒ 00:10:55.440 Uttam Kumaran: kind of the goal here is… is also for our couple of other combination signals. There’s, like, different paths, right? Like, …
88 00:10:55.700 ⇒ 00:11:13.670 Uttam Kumaran: trial versus demo, different nuances about the form filling that are existing in several single signal combinations. So this is, like, an extremely crucial enrichment step for us to get right. So we’re just gonna probably do a little bit of a spike on, like, making sure we can nail this. …
89 00:11:14.430 ⇒ 00:11:21.260 Uttam Kumaran: But that’s probably… at that point, we kind of wrapped up everything we need for this first signal combination, I would say.
90 00:11:30.120 ⇒ 00:11:31.900 Uttam Kumaran: I think this would be….
91 00:11:33.560 ⇒ 00:11:38.360 Ryan DeForest: I think that part would be either, like, also the most important, but also the most difficult to get right.
92 00:11:38.670 ⇒ 00:11:44.909 Uttam Kumaran: Yeah, this was, like, last time we talked about, like, oh, they have Marketo here, and then they’re using maybe Cal.com in one place.
93 00:11:45.030 ⇒ 00:11:47.269 Uttam Kumaran: I don’t, I don’t, I don’t think, …
94 00:11:47.780 ⇒ 00:11:52.009 Uttam Kumaran: Yeah, nothing so far is, like, impossible. I think we just need to do… it’s a little bit deeper.
95 00:11:52.270 ⇒ 00:11:54.110 Uttam Kumaran: … Yeah.
96 00:11:54.360 ⇒ 00:11:59.749 levkatreczko: On the topic of the cost of running this type of play, I’m curious…
97 00:12:00.210 ⇒ 00:12:04.279 levkatreczko: Do you anticipate it being, like, one of the more expensive routes?
98 00:12:06.270 ⇒ 00:12:15.179 Uttam Kumaran: Maybe I’ll let Mustafi, you can give your feedback as well, but for me, not really. Like, I think all we’re gonna need to do is have some sort of browser automation.
99 00:12:15.260 ⇒ 00:12:26.900 Uttam Kumaran: where we can render… like, typically, we’ve used browser-based, but there’s now a couple of other options, where we would render the page, and then extract the HTML, and then pass that to an agent.
100 00:12:27.070 ⇒ 00:12:34.739 Uttam Kumaran: I’m not 100% sure yet, but I do think that some of the job APIs are actually probably more expensive.
101 00:12:34.910 ⇒ 00:12:39.969 Uttam Kumaran: The other thing is this signal is core to many different,
102 00:12:40.300 ⇒ 00:12:51.700 Uttam Kumaran: combinations. Like, if we look at the list of combinations that we’re going after, we have, like, form friction, DIY forms, so…
103 00:12:51.900 ⇒ 00:12:53.280 Uttam Kumaran: I think, like
104 00:12:53.380 ⇒ 00:13:00.350 Uttam Kumaran: if I were to say, we’re gonna look to get it right, and then we’ll see what we can do on the cost side, but I don’t think it’s gonna be, like.
105 00:13:00.570 ⇒ 00:13:03.330 Uttam Kumaran: much more expensive. I think it’s actually just gonna be, like.
106 00:13:03.540 ⇒ 00:13:05.939 Uttam Kumaran: We have to find the right orchestration tools.
107 00:13:07.950 ⇒ 00:13:09.589 Uttam Kumaran: I don’t know, Mustafa, you disagree.
108 00:13:10.590 ⇒ 00:13:35.510 Mustafa Raja: Yeah, I pretty much agree with it. We see that, in the third sign… third, combination, we have scheduler present. Now, this type of data doesn’t come with, their stack, so, with form friction, we are… we would be able to fetch, the schedulers that the particular lead is also using. So, so here, here, with this automation, we’ll be… we’ll be satisfying two of our signals.
109 00:13:35.510 ⇒ 00:13:37.889 Mustafa Raja: So, I feel… I think this is pretty good.
110 00:13:39.250 ⇒ 00:13:44.419 Uttam Kumaran: Yeah, I actually think, Lev, if we get this right, we can probably stop using their stack.
111 00:13:44.940 ⇒ 00:13:48.630 Uttam Kumaran: … You know, in some cases, because this will be more accurate.
112 00:13:48.940 ⇒ 00:14:08.940 Uttam Kumaran: Or use their stack as a backup, so… But again, like, I want to get something soon that you guys can use, and then we can continue to rerun the enrichment as we make version improvements to this pipeline rate. So, sort of balancing that, but I… but I do think that this is going to be worth it.
113 00:14:11.050 ⇒ 00:14:11.840 levkatreczko: Fool.
114 00:14:12.240 ⇒ 00:14:12.960 Uttam Kumaran: Dots.
115 00:14:13.650 ⇒ 00:14:17.160 Uttam Kumaran: … Okay, great, so…
116 00:14:18.160 ⇒ 00:14:23.500 Uttam Kumaran: And then on the next thing, so this… the next signal was, like, routing enrichment and gap in dual path.
117 00:14:23.580 ⇒ 00:14:32.870 Uttam Kumaran: So, yeah, we actually… we can totally detect dual-path conversions. So, like, are there multiple CTAs? Like, is there a demo CTA? Is there a free trial CTA?
118 00:14:32.900 ⇒ 00:14:50.789 Uttam Kumaran: We also have an enrichment present tool detection. So this is our prompt. I wanted to put it here, Lev, in case you wanted to check it out. I’ll make sure this is shared to you, by the way, but it’s also, you can go into play and see this, but basically, we’re like, if this is the prompt to find the enrichment tools.
119 00:14:50.890 ⇒ 00:14:51.890 Uttam Kumaran: …
120 00:14:52.100 ⇒ 00:14:59.950 Uttam Kumaran: again, everything is gonna get better, as we start to extract stuff from the HTML, especially enrichment tools are all
121 00:15:00.300 ⇒ 00:15:04.849 Uttam Kumaran: they track IP, they enrich it, they go put it somewhere, so it’s all gonna be on the front end.
122 00:15:04.990 ⇒ 00:15:07.030 Uttam Kumaran: But this was…
123 00:15:07.440 ⇒ 00:15:14.740 Uttam Kumaran: Fine to do, and then form friction score is also going to be based on if we’re gonna get the form friction… the forms figured out.
124 00:15:14.860 ⇒ 00:15:15.860 Uttam Kumaran: …
125 00:15:16.010 ⇒ 00:15:23.110 Uttam Kumaran: So I think we’re gonna have the form stuff in the next two days. I just needed to spend another year working on the spike for that.
126 00:15:23.500 ⇒ 00:15:32.809 Uttam Kumaran: … And so, kind of, like, probably next steps here, one is… We’re going to…
127 00:15:32.980 ⇒ 00:15:43.570 Uttam Kumaran: Mustafa, I think it’d be great for us to start putting into this Notion doc a, like, a cost section. So, like, cost per notebook.
128 00:15:43.680 ⇒ 00:15:45.150 Uttam Kumaran: Run, basically.
129 00:15:45.320 ⇒ 00:15:47.569 Uttam Kumaran: We can just maintain a table.
130 00:15:47.710 ⇒ 00:15:52.760 Uttam Kumaran: So that’s there. I know you sent it to me on Slack, so we’ll just put it in there.
131 00:15:52.970 ⇒ 00:15:54.700 Uttam Kumaran: Just anywhere here.
132 00:15:54.830 ⇒ 00:16:01.129 Uttam Kumaran: Second thing is we have this, like, form, spike that we want to do.
133 00:16:01.520 ⇒ 00:16:02.270 Uttam Kumaran: …
134 00:16:02.510 ⇒ 00:16:12.090 Uttam Kumaran: The other item is, like, now that I… I just kind of wanted to confirm, Brian, if we’re good, we can actually start to get all of these into Salesforce.
135 00:16:12.290 ⇒ 00:16:19.590 Uttam Kumaran: And then create some of the reports. Now that we have, like, both of these signals, we would look to basically
136 00:16:19.770 ⇒ 00:16:26.379 Uttam Kumaran: Create a report for each of these Signals, where, for example, the…
137 00:16:26.930 ⇒ 00:16:36.459 Uttam Kumaran: they fit the criteria. Like, the… I think a good example is, like, if we go to this, it’d be like, hey, the, …
138 00:16:36.710 ⇒ 00:16:39.229 Uttam Kumaran: The overlap score is above 3.
139 00:16:39.340 ⇒ 00:16:46.710 Uttam Kumaran: They have a high tool count, and they have the jobs openings. Great, like, put them in this report. …
140 00:16:47.010 ⇒ 00:16:51.710 Uttam Kumaran: the… probably the only question I would have is if we’re… if you have a…
141 00:16:52.190 ⇒ 00:17:00.489 Uttam Kumaran: list of priority accounts for us to run this for, or if you want us just to select at random, like, any amount. …
142 00:17:00.640 ⇒ 00:17:03.880 Uttam Kumaran: I know we presently just selected these
143 00:17:04.109 ⇒ 00:17:07.609 Uttam Kumaran: But, tell me which ones to… to start to enrich.
144 00:17:12.140 ⇒ 00:17:17.189 Ryan DeForest: Yeah, I’m down to see it a little bit in action, but, like, I’m still…
145 00:17:18.190 ⇒ 00:17:24.660 Ryan DeForest: cognizant based off of the one that’s still in progress, to be honest. That’s, like, really top of mind for me.
146 00:17:24.849 ⇒ 00:17:25.429 Uttam Kumaran: Okay.
147 00:17:25.430 ⇒ 00:17:31.849 Ryan DeForest: So maybe once that one is done, then we could just knock out all three at the same time, if that works? Okay, okay.
148 00:17:32.640 ⇒ 00:17:36.829 Ryan DeForest: I’d rather just make sure that that part is dialed in, and then… because then that can, like.
149 00:17:37.110 ⇒ 00:17:39.129 Ryan DeForest: Work alongside these two pretty good.
150 00:17:40.080 ⇒ 00:17:44.679 Uttam Kumaran: Do you have a sense for if you’re gonna wanna run enrichment on, like, a select?
151 00:17:45.070 ⇒ 00:17:46.889 Ryan DeForest: A couple of thousand….
152 00:17:46.960 ⇒ 00:17:52.349 Uttam Kumaran: or… like, you gonna do it based on, like, a geo, or what do you think?
153 00:17:54.100 ⇒ 00:17:57.329 Ryan DeForest: That’s a good question, and something that me and Lev…
154 00:17:57.840 ⇒ 00:18:00.000 Ryan DeForest: We can tag team a little bit next week.
155 00:18:00.180 ⇒ 00:18:02.479 Ryan DeForest: Cause I’m going into the office next week, too.
156 00:18:02.510 ⇒ 00:18:03.470 Uttam Kumaran: But, like.
157 00:18:04.000 ⇒ 00:18:05.820 Ryan DeForest: There’s gonna be parts of it, like.
158 00:18:05.880 ⇒ 00:18:23.339 Ryan DeForest: where we have, like, a thousand accounts that we want to outbound to, that we want to do… we want to do, like, something for it, right? But then we don’t want to do everything. And there’s, like, some accounts that maybe they’re signing up for this… a demo coming up next week, and we want to prep the… You’ll auto-hit them. Yeah, we want to prep the AE for…
159 00:18:23.370 ⇒ 00:18:26.279 Ryan DeForest: For, like, what the hell to talk about and why, right?
160 00:18:26.280 ⇒ 00:18:27.220 Uttam Kumaran: ….
161 00:18:27.220 ⇒ 00:18:28.200 Ryan DeForest: So….
162 00:18:28.200 ⇒ 00:18:28.950 Uttam Kumaran: I was like….
163 00:18:28.950 ⇒ 00:18:33.699 Ryan DeForest: I know it’s not a full answer to your question, but, like, definitely we want to have the power to do, like.
164 00:18:33.970 ⇒ 00:18:39.030 Ryan DeForest: one or the other are full, or fully baked out, but not fully on, you know what I mean?
165 00:18:39.260 ⇒ 00:18:45.440 Uttam Kumaran: Yeah, yeah, so there’s gonna be some that need to be auto-enriched, and then some that are tied to…
166 00:18:45.670 ⇒ 00:18:48.300 Uttam Kumaran: Maybe, let’s just say, an active campaign.
167 00:18:48.430 ⇒ 00:18:49.510 Ryan DeForest: Yep. ….
168 00:18:49.880 ⇒ 00:18:54.520 Uttam Kumaran: Yeah, I would… I mean, for me, I would just caution against doing it on the full thing, so if we could pick…
169 00:18:54.850 ⇒ 00:18:58.320 Uttam Kumaran: 500 to 1,000, it just lowers the risk of something messing up.
170 00:18:58.420 ⇒ 00:19:05.069 Uttam Kumaran: We can also do this once we get the cost, so I can kind of show you what’s… what it’s going to be estimate-wise, but…
171 00:19:05.190 ⇒ 00:19:07.650 Uttam Kumaran: Yeah, ideally, this is tied to some…
172 00:19:07.790 ⇒ 00:19:10.810 Uttam Kumaran: Campaign, or some active, like, segmentation?
173 00:19:10.940 ⇒ 00:19:25.860 Uttam Kumaran: Whether it’s, like, one salesperson, a GEO, or it’s like, hey, we’re going after everybody in, like, mid-market SaaS. Great. We have, like, 300 accounts, let’s run them… let’s run them through our…
174 00:19:25.960 ⇒ 00:19:27.010 Uttam Kumaran: our enrichment.
175 00:19:28.590 ⇒ 00:19:39.860 levkatreczko: Yeah, I’m also doing a lot of list building, for campaigns right now, and so I would probably begin by deploying a lot of these in sort of an ad hoc basis, like.
176 00:19:39.960 ⇒ 00:19:57.370 levkatreczko: I’ll spin up lists of accounts via Apollo or Clay, and we’ll be testing out my own campaigns, probably with some of these triggers as, you know, reason for outreach. And, you know, over time, that’ll limit the exposure to, like, potential issues.
177 00:19:57.830 ⇒ 00:20:01.640 Uttam Kumaran: So for that, like, are you gonna create those lists in Salesforce?
178 00:20:01.880 ⇒ 00:20:04.469 Uttam Kumaran: Or it’s gonna be created externally.
179 00:20:04.470 ⇒ 00:20:06.590 levkatreczko: Likely not, but… Okay.
180 00:20:07.180 ⇒ 00:20:18.320 levkatreczko: it could be worked out so that the… we have, like, the Salesforce integration in play, so I could potentially, like, run updates through that, route, but I’d love to hear, like, why you asked that specifically.
181 00:20:19.780 ⇒ 00:20:24.300 Uttam Kumaran: Yeah, right now, we just have this hooked up. I guess, like, for me.
182 00:20:25.360 ⇒ 00:20:30.089 Uttam Kumaran: it depends. If you’re doing sort of just, like, broad cold outbound, I probably wouldn’t…
183 00:20:30.570 ⇒ 00:20:33.609 Uttam Kumaran: do it in Salesforce. …
184 00:20:33.750 ⇒ 00:20:41.699 Uttam Kumaran: And what you can do is we can just clone this and hook it up to whatever, wherever you want, but those sort of, like.
185 00:20:42.620 ⇒ 00:20:49.319 Uttam Kumaran: If it’s purely gonna get created for, like, you’re gonna create an email in here, you’re gonna create a LinkedIn message.
186 00:20:49.460 ⇒ 00:20:53.909 Uttam Kumaran: I guess my only point would be, where does the enrichment get centralized?
187 00:20:57.860 ⇒ 00:20:59.430 levkatreczko: Oh yeah, that is a good point.
188 00:20:59.700 ⇒ 00:21:00.720 levkatreczko: …
189 00:21:02.430 ⇒ 00:21:07.649 levkatreczko: I’ll have to… I’ll have to ponder that. I mean, there’s definitely a workaround where I spin up a list, let’s say in Clay.
190 00:21:08.020 ⇒ 00:21:13.900 levkatreczko: I import Salesforce, and then… And then run it in the way that you’ve been running it.
191 00:21:14.830 ⇒ 00:21:23.800 Uttam Kumaran: Yeah, I mean, light version is, like, you could… you could put into here and get it out into, like, a spreadsheet or something if you need to upload that somewhere, but…
192 00:21:23.980 ⇒ 00:21:27.280 Uttam Kumaran: I do think that if it can get back into Salesforce.
193 00:21:28.150 ⇒ 00:21:32.799 Uttam Kumaran: and we can just… Salesforce has lists, we can just bifurcate that and put into…
194 00:21:33.070 ⇒ 00:21:36.679 Uttam Kumaran: put into a list there, and then it could get reused, right?
195 00:21:36.680 ⇒ 00:21:42.530 levkatreczko: Yeah. I mean, looking at this, sheet right now, how do you have this hooked up to the CRM?
196 00:21:43.500 ⇒ 00:21:51.499 Uttam Kumaran: It’s all… I hid the Salesforce, tab, but there… it’s… we bring in stuff from Salesforce, and then we write back to Salesforce.
197 00:21:51.910 ⇒ 00:21:57.300 levkatreczko: Okay, well then potentially I could just add that right to Salesforce step in most of the flows that I’m imagining.
198 00:21:57.460 ⇒ 00:22:07.489 Uttam Kumaran: Yeah, so I think if in Salesforce there is a list building. In fact, the way we got these 6 is we added it to our Brainforge test list, and then it gets pulled into here.
199 00:22:07.660 ⇒ 00:22:10.979 Uttam Kumaran: So, it’s the same flow in that you would create a list.
200 00:22:11.120 ⇒ 00:22:17.129 Uttam Kumaran: And then you can pull them into here, and then push… basically, we push the properties back.
201 00:22:17.570 ⇒ 00:22:23.939 Uttam Kumaran: I think it depends on, like, what the action you’re taking. Like, if it’s gonna be, like, hey, for example, we have this, like, more…
202 00:22:24.160 ⇒ 00:22:35.819 Uttam Kumaran: non-PLG, like a sales motion, right? But then, if you’re gonna send that to instantly for emails, or you’re gonna send that to somewhere for, like, messaging, I think it matters, like, where that is.
203 00:22:35.980 ⇒ 00:22:41.819 Uttam Kumaran: … But, again, like, I want you to be able to reuse all this logic, you know?
204 00:22:48.210 ⇒ 00:22:51.499 Uttam Kumaran: So maybe we test, like, when you have a list that you want to go after, maybe we test….
205 00:22:54.180 ⇒ 00:22:55.939 levkatreczko: Yeah, I think that’s the best route.
206 00:22:56.290 ⇒ 00:22:57.310 Ryan DeForest: Yep, agreed.
207 00:22:57.790 ⇒ 00:22:58.360 Uttam Kumaran: Okay.
208 00:22:58.820 ⇒ 00:23:05.820 Uttam Kumaran: Okay, cool, so… I think we’re good. I think probably the only other… …
209 00:23:06.010 ⇒ 00:23:09.409 Uttam Kumaran: We’re gonna just continue working through, …
210 00:23:09.730 ⇒ 00:23:17.209 Uttam Kumaran: these signals, and then, yeah, Mustafa, maybe let’s chat after this, and we can start working on a spike for the forms.
211 00:23:17.770 ⇒ 00:23:23.550 Uttam Kumaran: We… we… we can use browser-based, there’s a couple of other options for browser, for headless browser.
212 00:23:23.860 ⇒ 00:23:24.850 Uttam Kumaran: Yep.
213 00:23:25.700 ⇒ 00:23:30.060 Uttam Kumaran: Yeah, and we also have some past NNN workflows where we do that, so…
214 00:23:30.260 ⇒ 00:23:32.190 Uttam Kumaran: I think we should be able to get it done this week.
215 00:23:32.820 ⇒ 00:23:33.810 Uttam Kumaran: Yup.
216 00:23:35.750 ⇒ 00:23:36.540 Uttam Kumaran: Cool.
217 00:23:37.050 ⇒ 00:23:45.080 Uttam Kumaran: Yeah, and then maybe other updates. So, we’re meeting with Justin to start doing stuff on the product analytics side.
218 00:23:45.130 ⇒ 00:23:58.679 Uttam Kumaran: just basically getting a lot of the, like, basic metrics on the platform. So that’ll be something I think we’ll have something to show next week, and we’re… we’re also talking to Caitlin about that, so we’re synced up there.
219 00:23:58.870 ⇒ 00:24:03.589 Uttam Kumaran: … And then we can also start to use those, …
220 00:24:03.750 ⇒ 00:24:06.070 Uttam Kumaran: For signals, for the product event.
221 00:24:06.380 ⇒ 00:24:07.659 Uttam Kumaran: related signals.
222 00:24:07.820 ⇒ 00:24:08.720 Uttam Kumaran: ….
223 00:24:10.440 ⇒ 00:24:11.030 Ryan DeForest: Perfect.
224 00:24:11.600 ⇒ 00:24:12.250 Uttam Kumaran: So, yeah.
225 00:24:12.850 ⇒ 00:24:19.280 Uttam Kumaran: Cool, okay, so let’s try to aim, Mustafa, to get something on forms ready so that team can review this week.
226 00:24:19.510 ⇒ 00:24:26.520 Uttam Kumaran: And then, like, I want to try to drive towards getting this into Salesforce, you know, so getting your approval right whenever we’re ready.
227 00:24:27.210 ⇒ 00:24:33.000 Ryan DeForest: Yep, yep, and we’ll get a list of things that… ones that you can test out, too. More… more than just 10.
228 00:24:33.450 ⇒ 00:24:34.700 Uttam Kumaran: Okay, okay.
229 00:24:36.190 ⇒ 00:24:40.080 Uttam Kumaran: Okay, great. Anything else that we can help with?
230 00:24:42.000 ⇒ 00:24:44.080 Ryan DeForest: Not at the moment. I think we’re good to go.
231 00:24:44.800 ⇒ 00:24:46.290 Uttam Kumaran: Are you in New York, Brian?
232 00:24:46.670 ⇒ 00:24:49.309 Ryan DeForest: No, I’m in Reno, Nevada, so outside Lake Tahoe.
233 00:24:49.310 ⇒ 00:24:52.150 Uttam Kumaran: Oh, okay, okay, okay.
234 00:24:52.150 ⇒ 00:24:56.299 Ryan DeForest: I’m heading there on Monday, I’ll be there Monday morning, taking a red-eye.
235 00:24:56.470 ⇒ 00:25:00.189 Ryan DeForest: Sunday night. I’m too old for that shit, but we’ll… it’ll work. It’ll work.
236 00:25:00.540 ⇒ 00:25:01.670 Uttam Kumaran: Are you there all week?
237 00:25:01.830 ⇒ 00:25:03.899 Ryan DeForest: Yeah, they’re Monday through Friday.
238 00:25:04.680 ⇒ 00:25:12.480 Uttam Kumaran: Nice, okay, okay, great. Yeah, I grew up in the Bay Area, so I went to Reno a lot growing up. Oh, yeah. But yeah, that’s a brutal flight.
239 00:25:12.480 ⇒ 00:25:15.779 Ryan DeForest: Yeah, yeah, yeah, red eyes, red eyes are not fun, we’ll see what happens.
240 00:25:16.750 ⇒ 00:25:20.329 Uttam Kumaran: Nice, okay, okay, great. Alright, guys.
241 00:25:21.120 ⇒ 00:25:23.750 Ryan DeForest: Awesome. Well, I appreciate it. Looking forward to the next updates.
242 00:25:24.080 ⇒ 00:25:25.830 Uttam Kumaran: Okay, alright, thank you.
243 00:25:25.830 ⇒ 00:25:26.600 levkatreczko: This goes….
244 00:25:27.020 ⇒ 00:25:27.930 Henry Zhao: Byron.