Prospecting & Outbound Personalization w/ Yash Tekriwal from Clay - GTM Engineer School (Preview)

Today we'll introduce GTM Engineer School briefly and then Yash will share a bit about the workflows you learn in GTM Engineer School, which kicks off Tuesday, April 29th. So this is the agenda. Do a little intro, Yash is going to present, and then we'll take some questions. And I am recording this and I'll share the link to the recording a little bit later today. Okay. About your host. Well, yeah, like I said, I'm Jared, based in the heart of Silicon Valley. My background includes product growth and demand gen. I led growth at Auth0 and ran campaigns at Okta1. Constant throughout my career has been a commitment to continuous learning, both for myself and for my team. And I've gotten the most out of structured live learning environments where you can interact directly with experts. This is programs like Reforge Sales, Impact Academy Pavilion, and Winning by Design. They've all made a big impact on me and they, they've inspired how we've designed this course. Yash probably needs no introduction, but I, I will let you go introduce yourself and then I've got a couple more slides.

Yeah, I know. Let's do it. I'll keep it quick because I think people really want to get into it, but I think wide diverse background, obviously currently running the education department over at Clay. And I think the one anecdote that I'll share is my original background in education also comes from I was working with Alt MBA as a coach, and I think that had a tremendously powerful impact on my life because it was the opposite of every online learning experience I'd ever had where I would just go to Coursera, take a quiz, sort of get some knowledge, but never really build a skill. And instead we were spending 15 hours on Zoom a week for four weeks. I made some of the best friendships I've ever made in my life through that kind of interactive learning experience. And I also grew radically as a person. And so being able to do that in different contexts and ways is super exciting to me. And I love what we're doing with the GTM Engineering School.

Cool. Thanks, Yash. So the tech is f is is evolving faster than I can keep up that phrase. I've heard variants of that from a lot of folks. AI is absolutely transforming how go to market teams operate, and the ones that learn how to use it effectively, we'll pull ahead. But to be honest, the GTM tool landscape can feel overwhelming. And learning it all on your own is tough. And that's why we've built this program. We've assembled 30 top voices from founders to educators to seasoned GTM operators to help you through the noise. This isn't just lectures, it's project based hands on learning. You'll walk through 16 practical mini projects using five of the most powerful tools in the GTM stack. You'll get frameworks, case studies and best practices. But most of the learning happens through doing no prerequisites are required. It's beginner friendly and it's made for people in rev ops, sales growth content, PMMs and marketing ops roles. The course includes 28 hours of live instruction spread over four weeks and if you miss a session or two, replays are available. By the end, you won't just understand the tech you'll be building and we're giving you the tools to do it. So for example, three months of Octave Core, which would normally cost $450. Two months of Clay Pro which would cost you $1,600 plus credits from Cargo, Lemlist and Air Ops. Altogether that's going to be well over $3,000 of software included. Go briefly through the four modules and I'm going to hand it over to Yash. So in week one, messaging and prospecting, you'll define your ideal customer profile and turn that into a compelling message. You can operationalize and activate across channels. You'll walk away with a solid list of high quality leads and a process to score and prioritize them. Then in week two we get into cold outbound and personalization. And this week we'll build and automate highly personalized outreach. Then optimize deliverability so you can scale your outbound efficiently and effectively. Week three Agentic GTM Automation. It's all about building smart GTM agents, automating research, qualification and outreach based on product usage or deal lookalike so you can spend less time guessing and more time closing. And then week four we switched up a little bit and this is going to be about AI led content creation. We're going to wrap up by teaching you how to turn any topic and to optimize SEO content and repurpose it into short form and messaging assets, giving you a repeatable system for high leverage content creation. And the TAs are going to be in the zoom breakouts guiding you. Folks like Kellen Patrick, do you need something? If you guys aren't speaking, if you could mute mute your that would be awesome. Okay, so yeah, so these are some of the TAs that we're going to have helping you, guiding you. And these are, these are experts, practitioners, folks who run agencies Things like that. Okay, so today's focus is going to be in weeks one and weeks two that Yasha is going to go through. And I just put at the bottom here, and I'll put in the chat as well, this link, the YouTube link, will have replays of the webinars that we're doing this week and next week and the following week. And then this other link is, if you want, up for a couple of webinars that we have coming up. And with that, I'm going to stop my sharing and hand it off to you. Yash, you want to request access?

Let's do it. I think we should have it. Let's see if it works. All right, is everyone seeing my slides here? Awesome. So we're going to give you guys a little teaser, right? We don't have the time to do a full module. Just for context, like Jared was describing in the course, the way this is going to work is it's going to be something similar to what I'm doing here, where we'll share for, you know, 15 to 30 minutes what exactly we're about to do in the module. And I'm going to first demo it so that you guys know exactly what to expect and what you're doing. And then the part that we're not going to get into, obviously, live on the webinar is we're then going to spend an hour, hour and a half in small breakouts where the ratio is one person, one ta, one teaching assistant to every two students helping you guys actually do this process yourself. Because that's what Jared and I were just talking about, right? That's where the learning happens. Not from watching me do it, but from doing it yourself after having just seen how someone else does it and having resources to help unblock you. So what I'm going to preview is quickly a couple of frameworks around prospecting and Outbound that we love using here at Clay, a way that we think about Clay, broadly speaking, and then jump a little bit into Clay. And while I'm doing that, I'm actually just going to ask people, let's make this off the cuff interactive. Go ahead and drop your company domains that you work at into chat and or your LinkedIn profiles, and I'll just pick one at random when we get to the interactive part and see if we can do that live. All right, so first things first, I think before you jump into complicated software tools, it's really helpful to understand the way in which you're going to approach that tool and what you're going to get out of it. One of the biggest problems I've heard people come up with at clay and actually struggle with this is not unique to clay, though I think it's really true of any of these more complex GTM engineering tools is where do I start? Because I think a lot of times we're used to software where we just jump in. I think of Superhuman as a good example. The purpose of the software is explicitly clear. It's to help me get through my email faster. I just have to go through the onboarding, get better at the keyboard shortcuts, and use the tool the way it was intended. I don't actually have to go in and creatively think about how to use the tool myself. Whereas all of these GTM engineering tools, I think everything in the program Play, Air Ops, Cargo, Lemlist, are a little bit more creative in nature. The tool is only as good as the hypothesis, experiment or execution that you're going to run through it. And it can be really intimidating if you just enter a blank canvas, a blank workflow builder, a blank clay table, and you don't exactly know what you want to approach. So I think the best way to think about using these tools is to start with your end goal in mind. There's tons of end goals in clay. I'm not going to bother going through all of these right now because it's not the best use of our time. I think what's most important is that in clay in particular, there's always four steps, no matter what your workflow is. And using those four steps as cornerstone pillars of what you're trying to accomplish will help you understand what order you need to start adding things to your table. So we call this fet. Couldn't come up with a catcher one. This is the most memorable one because I think it's intuitive, natural and easy for people to remember. Step one of any data workflow. So this is clay. This is also just other tools. Frankly, I think if you apply this to Zapier, Air Ops is a good example. Cargo is a good example. You'll find that it's always the same process. You can't do anything without first having data. So step one is you need to find data. And in clay, the easiest way to think about that. And we'll cover the prospecting piece. I need data about companies, I need data about people. Any data about website visits, signals. There has to be a starting source. So you need to connect your source or you need to search for your source. And we call that find Then step two, once you have some initial data, I think everyone here is probably familiar with this. Initial data is never enough. You want more data about the people. Data you can use for personalization, for relevancy, for targeting. So that's where you need to enrich. Probably the most important enrichments are contact information. If I have your name, your company name and your job title. Thank you. Don't have your or LinkedIn. Right. So you can use waterfalls and clay to enrich. You can use AI in clay to enrich and then do a bunch of things from there. Once you then have all of your data, you can transform it so you can manipulate it, shape it, you can do summaries with AI, with AI formulas, clean it in the format that you need it for. And then the last step is you can actually run the whole thing. You can export or execute. I've seen people use both words here. Send that data somewhere else. Because you're probably living in a LEM list for sequencing, a HubSpot, a salesforce for CRM, using data in other ways in air ops and cargo, all of that good stuff. So I'm not going to spend too much time on each of these steps because I think frameworks are helpful. But also, again, learn from practice. These are just slides that we're happy to send afterwards if you want to know more about the many, many ways you can do each of these steps in clay. So this, remember, find, enrich, transform, execute. And then the one bonus I'll include here before we actually just jump into software is AI formulas, which I think is really cool. And we'll demo a quick example of as well. Now what we'll do is let's make sure. Can everyone still see my screen? Everyone see Clay now. Awesome. Yes. So let's actually apply this framework to software. I'm going to move this zoom window out of the way and I'm going to call attention to a couple of things. Right. I'm going to call attention to, number one, that first step, find. We have a bunch of different places we can find up at the top here. And I think one of the cool things about clay in particular is that you're not limited to the prospecting that I think all of us have been used to for the last two decades of work, which is I go in, I just find the list of companies, I just find the list of people, and then I enrich that data. Let's say your ICP doesn't live on LinkedIn. Then we could also try and find these SMBs through scraping Google Maps. If we wanted to find people based on their job changes or job signals, we could do it via job postings. We could even just export and download CSVs, we could scrape the web. All of these are things that we'll cover more in depth in the actual modules in GTME school. But for now, let's go through the quintessential example of finding companies and finding people. Before I jump into it, let's try this live. Let's see if we pull someone from the chat and make this contextual and relevant. So I see, wow, a bunch of companies in here. Let's pick speak.com because I don't know what that is and I'm curious about it. So even within the find companies example, there's tons and tons of ways that we can actually go about trying to create a list of people. So number one, I'm just going to show you, once we get into a workbook, I can add a source from a ton of different places. My recent favorite feature that's a little bit in beta, but I think is a cool one to just show off in terms of how AI is shaping. Go to market engineering is we can try and find prospects just based on your company website. You'll notice it's in beta. So let's see how this works live. And I think if we put in speak.com and then add, it's going to take a second to scrape and I'll explain what this is about to do. This is about to go use AI to analyze Speak.com's website, think about the competitors, the industry, the company description, what it does, and then use AI to auto generate a search for us that we can then modify and use to pull in a list of companies or people. Now my one thing that I tell people with AI is that I think the one worry I have with AI is that people get lazy and actually building the skills that you need to assess whether or not the quality of this output is good. So that's what we'll cover. A little bit more together is how do we think about narrowly building prospect lists? How do we narrowly think about ICPs? In the course itself, we'll actually give you a full rubric on how exactly to think about scoring your list so that it's actually narrowly fit and tailored and targeted. All right. Unfortunately, speak.com didn't work. I wanted to see if we could do this live. We'll come back to this in a second, but let's just do this from scratch then. So I'm going to actually go ahead and look@speak.com language learning app that gets you talking. Need to speak a language out loud to learn it. Speak users. Oh, this is cool. I like this. I have a bunch of friends who have gotten really into duolingo again recently and have been looking for other ways to be building their language skills. Okay, I like this a lot. Someone really you could talk to over AI. Okay, so I'm just gonna make this up for the sake of time on the fly. Let's say I am on the sales team at Speak and I am reaching out to employees at software companies with, let's say, headcounts greater than a hundred. Because I am making the assumption that if you work at a software company with a headcount over a hundred, there's a very high likelihood that you have access to some sort of L and D budget that you could use for a tool like speak.com and maybe if I wanted to get super specific on my targeting here, I would even try and look for companies where they actually care about training their employees in specific other languages. People who travel often, consultants, people who might be on the front lines in different countries. There's a whole industry of frontline workers needing language training. So let's think about that as our icp. Right. We're starting. Keep in mind that end goal of I want to reach out to people with a specific business objective. We're going to run through two steps just quickly here. Step one is we need to find the types of companies that actually have the criteria I was just talking about. So let's jump into find companies here. I'm just going to put in software development. I am going to put in company sizes. Let's go for just to keep it small. And we'll talk about this more in school as well. But the smaller your list, the better, in my opinion. Actually, I've seen this over dozens of different companies. Even if you have a large tam, you want to keep your list size typically less than 3,000, because getting it that small allows you to be more personalized in the way that you're reaching out. If you just grab a list of 10,000 people and there's nothing unique that separates them, you're going to have to struggle more for the creativity and what you do with your outbound. So we'll keep this to greater than 200 for now and then see how else we might want to narrow it. Let's keep it in the United States because that's probably where those types of L and D budgets are. Going to be unlocked. And then just to see how this goes, I'm going to test right this preview here and this is some of the art of prospecting that I like to remind people is they get a lot of stress when they look at the screen. They do it the first time. They're like, I have to get it right. You don't have to get it right. This is sort of a continuous learning process in its own right. Where okay, I can now see I've got 5,446 results. So I'm not quite at that 3,000 number. And if I look through this I can see that, okay, we've got a bunch of companies on here, but we also have still a pretty wide variety, variety of companies. I imagine reaching out to people at ServiceNow, Tableau, Freshworks, Databricks. These are some of the largest software companies in the world. That's going to be a lot of multi threading. It's going to be hard to find someone with the title that has the same level of power that someone with that same title at a smaller organization might have. For example, a VP of HR at Google is not going to have the same decision making power as a VP of HR at for example, Clay. We're 120 people right now. So I'm going to quickly update two things on this search actually. Let's take out the mega enterprises from this list and let's actually keep it down to just 200 to 500 and we'll even include the 51 to 200. So we're keeping this tight range of companies between 51 to 1000. Let's see how that changes our list first before I make any other steps and then while it loads. The other factor that I might play with here is the actual office location. Right. It might seem like a small detail, but for example, I'm based in New York, so if I'm doing outbound to a list of companies or people, then I might want to actually do a list of outbounds specifically tailored to people in New York and offer some time IRL to meet up. And I can tell you guys from hands on experience, I just came back from San Diego. People, especially these days in a post Covid world, love meeting up in person. It's fast, it's authentic, it's genuine and it creates connections in a way that is just not possible over digital. I was able to talk to three different vendors for our own education learning services that I had had email threads going with for the past two weeks. We got demos with all of them IRL on Tuesday. We got a sandbox for one yesterday, and we're probably going to implement it Monday of next week. So that kind of speed is what IRL can unlock. And then this search is taking a little bit of time. So what we'll do is refresh this page and then run the same search. So remember, we're doing software in the United States. 51 to 1,000 do. All right, so I'll come back to that find company search. We'll come back to software. We will come back to 51, up to a thousand, and then we'll do in the whole United States. Awesome. So we're actually seeing a larger list here. The reason why I just know where time pressed, otherwise I'd actually ask because I'd be curious if people could see the reason why is because we have a huge spread probably in that 51 to 200 category that we added that dwarfs the size of the larger companies. So just for the sake of time, we're not going to go super deep into only prospecting because I also want to show you guys some cool AI things I'm going to put in New York. We're going to hopefully narrow this list down to a couple thousand. There we go. 1,300. Import it to a new table and just continue for free for now. So now I've just got my list of companies. One thing that I could do to this list of companies is I could try and enrich it to see how important language learning is to their employees and or whether or not there's any information about whether or not employees will have budget to spend on something like Speak. So if I had a lot more time, I would write a better prompt. But I think what's really cool about AI today is we can use this clagent AI web researcher. And I don't have to be a prompting genius. Just to get started, let's use this help me button and I'm just going to voice dictate for a second. Help me write a prompt that researches a company website and figures out whether or not the employees at that company have individual or team budget for language learning or other stipends of their own. Do as much research as possible, execute as many steps as possible. Cost is of no object for this task. Also, just because I get asked this all the time, the tool that I just used is not clay. It is a tool called Whisper flow. Helps make dictation a lot faster. I'm now going to use this metaprometer and so what's going to happen is we're using AI to use AI, right? We're having AI take my prompt with some specific instructions and then go generate a prompt of its own and recommend the right model for us to use so that I don't have to become a perfect prompt engineer. Now, my caveat to this is that prompt engineering is still an important skill and one that will dig into more depth later. The quick preview I'll give you guys here is a lot of people I talk to don't use this format. Context, objective, instructions, examples. It's worth noting that talking to AI is still talking to a computer. And so this type of CSS HTML formatting where you put hashtags to separate sections of your prompt and you give the AI real context on what you're looking to accomplish really helps break down what we call zero shot prompting into manageable bite sized chunks. Because the big difference between prompting in any of These GTMe tools versus prompting an OpenAI and Claude is that you're not going to get multiple stabs at the apple to get the answer right. Right? I can't chat with the AI back and forth. I could try and replicate it here, but at scale that's not going to work. And it's also going to cost me thousands of additional dollars. And so getting that first prompt right so that I can nail the accuracy on the first go is really important. This metaprometer helps get me 80% of the way there without having to put in all the effort. But what we would also do in the program is iterate on this prompt because it's probably not going to be perfect. Just to show you guys an example of how this will work, we can look at the prompt here while it runs the language and you can see that it did a pretty good job, right? We're tasked with determining a company's website. Here's the objective. Employee benefits related to language learning budgets. Start by visiting the domain. It even knows which pages to go to, mentions of benefits if there's a search function, document any findings and then even give some examples. There's a few modifications I would make to this, even just looking at it, but for the sake of time and speed, we'll let this run and we'll see if it runs while we're waiting for it. And then the last thing that I will show you guys is finding people, finding contact details. And then I do want to leave it open to questions because that's probably going to be the best place for you guys to actually interact with some of the concepts that we're adding here. And we might have a bug this morning, so apologies for the live demo, but the last thing that we're going to add here is while we wait for these AI results to populate, we're going to go in and we're going to do find people at these companies because it's great to have the companies. I'm going to add some additional information from these companies about it. I can't reach out to a company, I need to reach out to people at these companies. If we're thinking about HR budget, if we're thinking about learning budget, I'm just going to look for, broadly speaking, people who are in human resources and recruiting functions or education functions at these companies. And let's include levels only of since we have smaller companies, we'll do C suite, we'll do VPs, we'll do directors, maybe heads as well, actually. And let's just preview this list. Notice I'm not putting in any of the other filters because I've already done a pretty good job on the company side. Right. These companies are limited to New York, therefore the people that I'm finding in this list should be smaller as well. That's a list of 800, which you can see at the bottom. So I feel good about that size. I'm going to go ahead and import this to a new table and then I'm also going to go ahead and add this work email waterfall and this mobile phone number waterfall off the bat. Now, I typically don't advise people to do this when they're first learning clay, because you're not really understanding what's going on here. But for the sake of a quick demo, I think it'll be helpful just to see how this works. And this is the last concept I want to show you guys before I show you one last table and then open it up to Q and A. I think we talked about AI, we talked about prospecting, we talked about web scraping. We've done fine companies, now we're about to do fine people. And what I think is really cool, the last step of play that really put us on the map, honestly, in a pre AI world is waterfalls. And for anyone who's been in the sales world for a while, I'm sure Jared can relate to this as well. When I was doing BDR AE work back in, gosh, what 2018, this didn't exist. We had individual subscriptions to five or six different data providers. And what I had built at the time was crazy advanced. I had Built a Google sheet that would automatically pull in the results via API from every data enrichment provider that we had. But someone still had to go in manually and then filter the sheet down to which emails we thought were valid because we had to use a separate validation service and then copy paste that data into the next tab in the sheet so that we were optimizing how much contact data we were finding without wasting all of our money finding redundant contact data on all the same people. The waterfalling concept takes what we used to do in like a few hours into a few minutes. Because now I'm going through Lead Magic, Find Email, Prospeo, Drop, Contact, Hunter, detagma, Wiza, and then only when we find the email and we validate it, we hide the validation columns just to reduce some visual clutter. But I'll show it to you live. That way we know this is a safe email to deliver to. And then we don't run anything else, we just throw it in the final column. And then we also do this for phone numbers as well. We have a few really cool like this didn't exist 10 years ago either. Phone number validation providers as well, so that you're not just picking up the phone and dialing random people to figure out whether or not they're actually real or in the worst case scenario, getting the company help desk. And so you can see that data populate automatically. And I don't know, I love the AI stuff. This is still the stuff that blows my mind the most. This was just not possible even five years ago to be able to instantly get contact data of this level across this entire scale. So that's almost everything I have to show you guys. I'm going to show you one more thing, right? For the sake of time, we're not going to build this out together live. Because AI copywriting takes a little bit of practice and purpose. And we use a concept I recommend everyone use, no matter what tool they're in, called AI snippets at Clay. The number one mistake I see people make with Outbound in AI is they're trying to use AI to write the whole email. Not a great approach. Here's why. Your best BDRs and SDRs are never rewriting the whole email. If you watch the process of your best BDRs and SDRs, what they are doing is they are personalizing the subject line sometimes most of the time. Then they're personalizing the open. They're adding some relevant context and personalization. They may fudge a few words in the ask, but the Ask is almost always the same. The ask is, can I have some of your time? The structure of that ask depends on the format of your email. And then maybe if they're going the extra mile, they'll add some context at the end that makes it feel really custom to an individual. And then they'll do that same thing throughout emails in a sequence if they have time. But they're taking a structured email and then modifying it to be contextually relevant to each individual. They're not rewriting the email from scratch every single time. And so we follow that same approach, just using AI to augment it in clay. And so I'll show you guys a table of how this can look as a final output. And this is the kind of thing that we're going to build together live in the GTM Engineering school. So first things first, just to make the PII of the data, this is only people at Clay. You can see here that AI web scraper that we use to go find the learning and development benefits. It's just going and finding LinkedIn profiles here. From those LinkedIn profiles, I'm able to scrape a bunch of additional information about these people using the CLAY database. And then what I'm doing is just number one, this is helpful information to a bdr no matter what they're doing. What are the most unique elements of this person's LinkedIn profile? I know their job title, I know the company, I know the headcount. I just want to know what makes this person special, what would speak to them if I were to reach out to them in an interesting way. And this is true for me. If you were to speak to. Maybe it's been a while now, but my debate experience, data science boot camps that got me into education, my alt MBA experience are all uniquely relevant things to me. I'll even go back to Everett, our head of GTM engineering. He talks about perceive all the time. He is such a nerd about computer vision. I love that AI picks this up because he talks about it so much on his LinkedIn. He's got a really great background. He's an engineer, a salesperson. He worked at some really old, boring legacy tech companies and some really exciting new ones. And then big, big fan of open source. So it's really cool that we're able to get all this kind of information unique to an individual just from AI. And then the last thing that I'll show you guys quickly here is we're using that AI web scraper to go get thought leadership about these people. One of the most effective outreach tactics I had as a BDR was if I could find a blog article, a webinar, an interview, a panel that someone had been on, I could go watch or read that information and then cite something specific from it. Response rates would be 30 to 50%, depending on the industry, if you could do a good job. Now, I think that tactic has become a little bit old, so don't expect those types of response rates. But the process is still hard. So I'm finding up to three you can see here. I'm finding podcasts, I'm finding webinars, blog posts, and then if I can, finding a third source of information. And then this is that AI snippets concept I was talking about. We're going line by line and crafting different parts of the email using AI so that we can be much more specific. And so, for example, I'll leave it at subject line because we're finding thought leadership. The approach for my copy here is going to be your insights on blank. And then I want to reference what the thought leadership is about. And that's what AI is able to do, is to take the context, take the content, and then output its own copy. And then we're doing that for first line, second lines, third lines, et cetera. And as a final element of this, just to preview what you're able to get to, I'm going to pick a funny one because I like the funny ones. You can obviously control these with voice and text in your prompting. Here's an example of what IT output for Scott, who's our head of content. Just listen to his podcast on the real stories behind business leaders success. Love this line in particular. That line could be improved, but we'll work on it long shot. Notice this middle part, right? Everything from long shot to the as well is boilerplate. We're not using AI for that. And then this P.S. here at the end is like a specific one that I've included in university videos. So also don't do this yourself. I think lots of people have tried to do it themselves now and it's become an old and tired tactic. But it's like, hey, notice this was your undergrad university. What's a specific unique tradition to that undergrad university? One of my best friends went there and it was really interesting. And they do actually all at Michigan State, right? Get together in the hallways and scream at midnight before exams to relieve all their stress. So like these small specific personal details are what help you stand out. The lesson here though is do not Just go copy what we're showing you here and what we'll teach you in the school as well. Right. Is be creative. Come up with these tactics and unique, relevant ways to the people you're reaching out to, and that will help you succeed. No matter what tools you're using, no matter how you're doing it, you can do it the manual way, but what we're going to teach you is the way to do it at scale, automatically, using all of these tools. So I took another minute extra than I wanted to, but let's open it up to Q and A. I'm going to stop the screen share for now. Feel free to ask anything. It can be about Clay. So I'll open back up the screen share and answer questions. But my advice to you guys would be ask questions about the program, about what you should be learning, about how this entire environment is changing every week. Like Jared was saying, there's too many tools to learn. And we'll. We'll answer regardless. Yeah, it's Harrison here. Am I okay to speak or you need to put the message in the chat? Yeah, Harrison, I appreciate it. Let's do. Let's first drop questions in chat, since there might be a lot of questions, but then we might have you come off chat just to come explain your question. Does that feel good? Sounds good. Thank you. Of course. Sounds good. And then also I'm going to catch up on the chat while people type some of their questions in. Yeah, Jorge. Talking about that prompt context and setting. So, so good. Jason. Wow. We got a lot of former Clay cohorts. People here as well. I love that. All right, Carl, let's look at the questions that we get, but then depending on the time, we'll potentially come back to the prompt specifically. Lydia. Yes, let's do this. I'm actually just going to share. A couple of people look like they're asking about the prompts from that table that I was just sharing. So let's make this easy. I'm just going to drop this link into the Zoom chat. If you go to this link, you will get a template of that table that has all the prompts in it. So you can go in, see how they're prompted, see how they're modified. I didn't have the time to spend, you know, diving into all eight of the AI prompts. But very, very useful. Okay, I'm going to jump through a couple of these questions. Rapid Fire Balaji is asking for a total beginner. How should you prepare ahead of the school? My two cents. I'M curious if Jared has anything else to add here. Look at the tools that we're using, look at the preview of the outline that we're going through and just try and get yourself familiar with these tools. Try and use them yourself for the things that you're most excited about. That will hopefully help you maximize the live learning time that we have. Because you're not going to be starting from zero, you're going to be starting from 10 or 20.

Yeah, yeah, I'd agree with that. And we're going to set our pre work ahead. It's not a lot of pre work, but it's going to be largely setting up your accounts and the tools and then pulling together some specifics about your product, your company, your audience, targets, etc. So that you can use them in the exercises in the course and have a usable output for your organization.

Yeah, exactly. All right, we're going to run through the best of these questions just because we were limited on time. Harrison interested to question the LinkedIn scraping with what happened Apollo, et cetera, recently. So here's an important detail. While we are getting LinkedIn profiles and using those profiles to reference more data about those individuals, Clay is not scraping LinkedIn. We have never scraped LinkedIn since the beginning of clay because of exactly what happened with Apollo and is another one that I'm blanking on. There are data providers, for what it's worth, that you can connect to in clay if you really, really want to directly scrape LinkedIn. Phantom Buster is a good example. You can use your Phantom Buster key in clay, but all the data that you get from enriched person or enriched company, while it looks a lot like the data that you might have on LinkedIn is actually not gathered from scraping LinkedIn. So that's a quick question to that. Let's keep going down the list. Night Thick has clay for job hunts, but unique tips or any creative ways to use clay? That's a fun question. There are lots of creative ways to use clay, right? Obviously our branding is around go to market engineering, but at our core this is a fun bit of clay history. It was a vertical sat or sorry, a horizontal SAS tool for the first five years of its existence. It was just like rose.com it was a spreadsheet you can connect APIs into. We focused in on go to market because it was the best use case for us to now have the market traction and pull that you guys have maybe been seeing on LinkedIn. But at its core the product is still the same thing it was seven years ago. We've just built a lot of specific go to market enrichments, but all those go to market enrichments are also data enrichment. So I'll, I'll name two off the top of my head, creative use cases that are not at all go to market related. Number one is someone actually internally for their onboarding projects two weeks ago built a clay table that uses scheduled sources. Every week on Tuesday and Thursday in New York, they release the new court times for tennis courts. And so it's really hard to have someone go and then book the courts for you and figure out what to book. So her table goes, finds which courts are available, figures out which times are available on her schedule for that week based on her Google Calendar, and then uses something like browser base or OpenAI operator to go and actually book two sessions for the course she wants to have. And I think there's a bunch of other people building bots as well. So it works like 60% of the time. But it's cool that she's able to do that. Another one is I actually built a career coach in clay where you can input your LinkedIn profile, you can input the next job you're looking for and the things you're struggling with, and then we're going to put it back through clay. Use AI in very specific ways to give you recommendations on what are companies that you maybe aren't looking for right now that you should be reaching out to, who are the people at those companies you should be reaching out to and what are some suggestions on your narrative or your LinkedIn profile that we would maybe add for the jobs that you're looking for at those companies. All right, busting down this list, Harrison, how does it all work? I wish I could tell you that part is proprietary, unfortunately. But we're happy to share some of the high level brushes and strokes, maybe. Clinton, do I see us getting to a place soon where we can personalize at a level where we don't need snippets? That is an interesting existential question. Potentially, I think the biggest thing that I don't really know where we're going to get to is agents. I don't know if anyone saw that Google released, I think yesterday, maybe two days ago, an agent to agent protocol, which means you can now have chains of agents talking to each other to execute commands. And so the, the fact of the matter is anyone who tells you they know what's going to happen in the next three months with AI, unless they're like Sam Altman himself maybe is probably full of shit. I have no clue what's happening. I am just trying to balance my philosophy here is balance the state of the art with the state of the practical. So, for example, the agent to agent protocol is what I would call state of the art. It's brand spanking new. People don't really know how it works. I played with it briefly last night and was like, okay, it's interesting. It's unlocking new capability. It is still rough technology. I'm not going to implement agent to agent protocol in my tech stack at CLAY anytime soon. Or maybe soon, depending on how it evolves. We're still going to experiment with what is really reliable to execute on today. So what I would say to your question, Clinton, is maybe I would prepare for a world in which you would need to personalize without snippets. You can personalize more in one call or two calls. And that's kind of why I like clay. I'm biased, obviously, but you could just take what I had as six columns and turn it into one. We'll do a good job of trying to keep up to date with those technologies and keep your stack flexible enough so that you can evolve with whatever's happening in the tech landscape. All right, blasting through questions. I know we've only got two minutes left. Doug, is the first tier still 149amonth or have we offered a less expensive one? Unfortunately, the first tier is still 149amonth. But like Jared said earlier, if you're a part of the GTM Engineering school, you'll get access to our pro plan tier features for two months. Included as part of the school. Seamless is the other one. Thank you, Sean. Heyreach is a LinkedIn tool that I do think. Actually, I'm not a lawyer, so I'm not going to speak to it. LinkedIn's terms and conditions are LinkedIn's terms and conditions. All I know is that we do not we comply with LinkedIn's terms and conditions. There is an official LinkedIn integration in Clay you can use as well and we can talk about that more in the program. It helps you monitor mentions of your company on LinkedIn. Diana. Local business workflows starting point is a website or domain, but you have a bunch of local businesses not listed on Google Maps. That's an interesting one. My quick hint to you, Diana, would be I would try Yellow Pages. You can access the yellow pages via Appify if you want to get that kind of a data source into Clay. Awesome. Some thanks. Thank you guys. Thank you guys for asking these questions. That's what makes these sessions exciting. Noel, I will think about releasing that career advice tool. I actually did release it a few months ago when I made it and then it burned through a bunch of AI credits for Clay and it was just for fun so I took it off. But I'll think about potentially re linking it. All right. A lot of people want the career advice table link. Okay, I will think about it. I will think about it for sure. All right, next webinar 149 confirmed. All right, there's a couple of questions I'm not getting to right now and maybe more that other people have, but if you've got more, feel free to shoot us a message on LinkedIn or message us on our Clay community, Slack, or reach out on the GTM Engineering School page as well. I think Jared posted another webinar. So if you guys want more of this type of content, attend. It's going to be fun. I think everyone teaching is going to be great and then I'll pass it back to Jared to close us out. Yeah.

Thanks everyone for participating for your questions. Thanks, Yash. Great session. And yeah, reach out on LinkedIn if you have any further questions about about the course, about anything. Looking forward to meeting you. And thanks everyone for attending. The recording will go out later today and that's all I got.

Sweet. Thank you guys. Have a good one. Enjoy your Fridays. Take care. Bye. Bye.