Name: From Event Signals to AI-Powered Personalized Touchpoints
Uploaded: 2025-09-24T18:03:32.512Z
Duration: 48 min 4 s
Description: From Event Signals to AI-Powered Personalized Touchpoints

Transcript for "From Event Signals to AI-Powered Personalized Touchpoints": Hi, everybody. Uh, welcome to our AI revenue playbook webinar series. I'm Kelly Cheng. I'm the CMO at Goldcast Events, and I'm very excited to have you join us and kick off, uh, for what's gonna be one of first a very exciting sessions in our a new AI revenue playbook series. Um, this series is all about helping demand gen marketers and revenue marketers figure out how to leverage programs, different tools, integrations, and AI to scale your revenue. Um, it is our goal to really cut the fluff and offer practical, implementable strategies and demos that help drive results. Um, and so today, we're actually gonna focus on taking your event engagement signals and turning them into personalized touch points with AI, and we're gonna be talking about this in partnership with a very, very cool company and product called Clay. I've been fangirling about Clay for the last couple of months, and they've definitely made some strides. Uh, before we jump in and before I go ahead and introduce our guest speaker, um, just a couple of, um, housekeeping items before we go in, uh, and and dive in. This session is being recorded. Uh, we'll share it out with everybody who registered within the next twenty four hours. Uh, and I love these sessions to be very interactive, so please do chime in in the chat. Um, let me know where you're you're joining us live from. It's always fun to see how global these webinars are. Uh, share your point of view as we kind of go through the workflow and the demo, and then also ask questions. There is a q and a tab on the, uh, the right hand side in your chat. We will have about fifteen minutes at the end of our webinar today to go through your live questions, so feel free to ask any questions you have about the tools like Goldcast Events, about Clay, about the integrations. And also check out the docs tab also nestled on your right hand side. We've loaded up some really great resources that I'll mention throughout our session today, um, and those are great takeaways for all of you. Um, before I go ahead and launch our poll, uh, our and introduce our guest speaker, I'm gonna launch our very first poll here. So, um, for those of you in the room that run events and webinars and digital, uh, events, um, I wanna know, um, what your biggest event follow-up nightmare is. So you should have just gotten a little pop up notification on your bottom left. Um, you can click on the polls tab now which has a a red dot on your right hand side. And let me know if, um, your as a demand gen marketer, revenue marketer, if your worst nightmare is a generic spray and pray type of email manually researching each and every attendee, figuring out who actually is interested to buy your product, um, and proving events actually drive revenue. So let me know in the poll, uh, what your biggest nightmare is. And while you're answering that, I would love to introduce, um, Ziqi Deng, uh, to join me on stage. She is a go to market engineer at Clay. She's gonna be our resident expert. Um, Ziqi, can you join me on stage? Hello. Such a thing Hi, Ziqi. How are you? Hi, Kelly. Uh, doing great. It's a nice day in your school here. Yeah. I'm excited. Awesome. Um, Ziqi, can you give us a quick background, uh, our intro and your background, and also tell us a little bit about play? Yeah. So I'm Ziqi. I lead our strategic partnership at Clay. Uh, I started off as a go to market engineer end of last year. Uh, before Clay, I was actually a founder for three and a half years where I started and scaled my own company. It's a fashion supply chain tech company. Uh, so I was working with fashion brand owner all the time, quite a different landscape in tech world. Um, but I scaled the company to 3,000,000, sold it, and, uh, joined Clay. The main reason is as I look back into my experience growing my own company, Clay was definitely a pivotal part of the success. Uh, I was a fan girl myself as well, uh, when I was joining. So the yeah. It's a crazy time from last November to right now. Um, I joined when we had 80 people. Now it's 200 plus. Internet is not so well these days because there's so many new people, uh, at the office. Um, but, yeah, super excited to, um, meet you guys here and share what I know about Clay. Awesome. Ziqi, I know that Clay just had your very first ever, uh, annual conference in San Francisco. Um, what was that like for you guys? Yeah. It it's really amazing. I think, um, it's one of those moments that you feel proud of your company. I think, like, I did not take any part of the planning, but the whole event was such a success, especially when you think about Clay being a SaaS company. It has it says, like, a a big music festival vibe at some point, and, also, people just really are very excited and, um, very hyped and great vibe. Uh, I I know people traveling from Antieta, traveling from Australia. It's really, uh, attendees from everywhere. Because Clay is a big community driven company, so it's it's a pivotal moment to really see it, uh, in in reality where, you know, all the folks who talk to each other, online on LinkedIn, on Slack community, uh, finally got got to meet with each other. So there are some really sweet moments when I was there. Awesome. I love the story of you being, um, you know, a founder and then selling your your company. Congratulations. Um, you know, I assume that your company was very much Clay oriented and and then joining Clay's full time. Um, Clay does have an incredible community. We actually, at Goldcast, work with a Clay agency, and I know that Clay has a a big, uh, portfolio of Clay certified agencies. And, um, the agency that we work with is called Growth Today. Highly recommend, um, any marketer out there that's looking for more support and expertise in, you know, using Clay. At Goldcast, we use Clay for a lot of our data enrichment. And over the last couple months or so, I've been, uh, I'm more of a novice in clay. There's always so much to learn. Um, and so I've been playing around with different use cases that can help me and my team in particular. Absolutely not an expert exactly why Ziqi is here, uh, but I'm excited to kind of dive in and showcase, um, the use case that I've, uh, found that have been really, really helpful for event marketers and and for our team in particular. Um, so I'm gonna actually go ahead and share that poll that I just shared. So these are really, really interesting responses, and we'll definitely keep keep these in mind as we continue to build out our, um, topics for our upcoming series. Um, so it looks like about, um, 17 17% of the attendees are, uh, you know, afraid of the generic spray and pray emails. About 13 are, um, you know, struggling with manually researching each attendee. A big portion of you guys, um, um, have a tough time figuring out who's actually interested, so really trying to understand intent signals. And then, um, finally, proving events actually drive revenue of, uh, almost 40% of you. So these are really, really helpful insights for us at Goldcast Marketing as we continue to try to create content, um, that helps, um, all of you all reach your goals and and figure out your help help you with your pain points. Um, alright. So I'm gonna stop sharing that poll now. Um, so I'm gonna dive right into a demo. So at Goldcast, if if this is your isn't your first event, you'll notice that we have a lot of different interactions. I just launched a poll. I'm gonna launch at least two more polls for for today. And we're collecting all these really, really great data insights that are first party. Um, you as attendees are answering these things, uh, based off of your own pain points and and real life experiences. And oftentimes, um, these data points just sit in your CRM and aren't getting used by your sales team. Um, it is a lot more challenging than it sounds to actually connect the dots to leveraging the data to actually using it for follow-up because you your sales team is doing this at scale. So I'm gonna go ahead and share my screen and walk through a demo now. Um, let's share screen. Okay. So if you are a Goldcast, this is gonna be a very familiar interface. If you are not a Goldcast, let me walk you through a little bit. Um, so this was an event that we ran about a month ago, August 27. Um, it's the top line. You'll you'll see some some some data points. Um, we have summaries, engagements, some post event content, uh, but I'm gonna dive specifically into the data. So in the this particular session, um, you'll see, you know, a a line chart of all of the the audience and that's in the room so you can really understand how long people are staying at any given time. Um, but, really, really, what I'm focusing on today is the attendee activity. These are all of the different great engagement points that we collect here at Goldcast. We sent here CRMs with the hopes that you know, go to market teams can leverage this data to help with their follow-up. Um, so as you can see here, there's a a question asked by somebody. Um, and when someone asks a question, um, what what a great way to to have an an an opening for a salesperson to come in and say, hey. I saw you asked this great question. Let let me help you more and open that conversation and be able to book a meeting. And so that's exactly what we're gonna be talking about today is how we can leverage, uh, first party data points like questions asked, like poll responses to really help your teams personalize your event follow ups, you get a stronger connect rate or a meeting booking rate from your post event outreach. Alright. So this is what it looks like in Goldcast Events. Um, if you are integrated with Salesforce, super easy. If not, you can download this raw data and also just import that directly into Clay, which I will walk through in just a minute. Um, I think Keith might be having a challenges hearing me. Is that, uh, is anybody else having a challenge hearing me, um, in the audience? Just just wanna make sure. Okay. Sounds like we are all good. Great. I'm gonna keep on going here. And so, uh, in Goldcast Events, we have some really, really great, um, integrations to different marketing automation platforms and CRM. And so we integrate with HubSpot. We integrate with Salesforce and Marketo. I'm gonna talk about Salesforce in particular here. So if I open up our Salesforce integration, all I have to do is connect my Salesforce campaign ID in here. And all this great data that I'm collecting, I think we have about 16 unique data points that we're collecting in every single event, um, including all the basics, like who attended, who did not attend, um, poll responses, if people up to updated their profile and changed their job title, if they've downloaded a resource, if they clicked on an event CTA button, um, all that data is being synced over to Salesforce, uh, one to one. So, um, if you don't have that integration, no worries. Again, you can download that manually within a raw table and then upload it into Clay. Uh, but if you do have this integration, uh, what's beautiful is that you can go right into Salesforce and build a Salesforce report. And so this is a a report that I built based off of this Goldcast spotlight event, um, and I'm specifically pulling out, um, questions asked. So all these questions are the real live attendees, what they asked in that particular, uh, event. So if you have any questions about the integration, uh, if you're a Goldcast customer and you'd like to use the integration or connect it, happy to, you know, make that make that happen for you. Just let us know after the event. Alright. So after you have your Salesforce report, um, you can directly go into play. And I'm gonna I'm gonna do two different workflows for you. One's one's a little bit more simple, and one's a little bit more complicated. Um, and this is where I definitely wanna make sure Ziqi is here to help me kind of walk through any more technical questions that I may not be able to answer. Um, but the more simple, um, workflow here is that I am, step one, importing that Salesforce report, um, into play. And then, uh, from that Salesforce report, I'm able to enrich different points to then have a very, very, very personalized output. Um, and so let me actually see if I can do a live demo of what this looks like. So if I add a new table in this workbook, um, I can go ahead and say Salesforce report. So import the records from Salesforce report. I'm gonna select my own account here. Really hoping that the Internet, uh, works with me. Um, so the the report that I wanna pull is is this Clay q and a report. So I'm gonna select Clay q and a report. I'm gonna submit it, and it's gonna start pulling in those records. It might take a couple of minutes for it to pull in, so we'll see, um, how long it does take. Let me minimize my bar here. Yes. So it it might take a little bit of time, but I wanted to kinda showcase how easy it is to actually pull a Salesforce report into Clay. Um, it's gonna take a bit of time. So I I've actually prepared this already here. And so here is the Salesforce report. Let me just double check to see because I I do wanna show how we can oh, there we go. Okay. Perfect. So we have the Salesforce records here. And if I click into these individual cells, uh, you'll be able to see all those fields, um, that are in the Salesforce report. Um, and so here, what I can do is create new columns. So it it matters to me what the event name is. I'll add that as a column. Um, it matters to me who the user is. I'll add that as a column. Um, it matters to me, um, the job title of the person. I'll add that as a column. The account name, just the same as company name, um, and the the question that they asked, And even their email address. Where is their email address? There it is. Email address is over here. Um, so these are all of the important, um, fields that I care about that is now directly imported into Clay. Um, Ziqi, anything you would wanna add here? Um, Yeah. I was just typing that you're very bold in doing a live demo. Something, uh, we're not advised to do, but great for you. Um, uh, I was just, uh, typing as well that, um, in terms of importing from Salesforce, it's actually quite simple like what you showed. Um, not only can you do a report, you could also do a list. I heard from my colleague that it could be quite challenging otherwise to pull a list from Salesforce. Um, you can also set up a daily refresh in certain plan can be hourly, so making sure that any newly added, uh, records to that report or list also get synced to Clay. So any workflows you build after having the records imported would also automatically get kicked off. Um, in that way, you you build a workflow and, uh, you don't have to worry about it. It will always run automatically in packet. Got it. We have a great question from Amber in the chat. I'm going to, um, share that onto the screen. Uh, so Amber is confused. Is Clay a Goldcast add on? Clay is not a Goldcast add on. Clay is a, uh, a separate tool that we use here at Goldcast. It is extremely powerful. Um, it Goldcast integrates with Salesforce, and Salesforce integrates with Clay. And so that is how we're kind of sending our first party data from events into Clay to be able to add enrichment, which I will kind of go through how we'll do. Hope that clears up your confusion, Amber. Alright. Uh, there's another We are gonna keep on going. So now that another question. Let's so Adam has a question for Ziqi. Does this process work the same with using the HTTP API versus the integration? Absolutely. So, um, you basically set up a I would I would say HTTP API when when it's, like, pulling information in, uh, you're using a webhook effectively. So how that does is you basically set up a webhook with with a, uh, source. For example, if it's your inbound form, if it's type form, from there, any net new added inbound leads will also get synced to Clay. And it should be API could also be added as a separate column. So if you are pulling information from another source, it's a post, uh, API call, for example, information also got synced into Clay. But that's more of a column level, um, enrichment. Cool. Awesome. Thank you for the question, Adam. Uh, please keep asking questions. Um, there's a q and a tab, um, on the right hand side, and we'll definitely make sure we cover any questions, uh, towards the end of our session as well. Um, alright. So as I I'm gonna keep on going. I'm actually gonna open my second poll now. Um, so I wanna know for the audience if if AI could help automate one thing for your event follow-up, what would it be? Would it be researching data enrichment, writing personalized messages, or prioritizing hot leads? Because all three of those things is possible with Clay. Today, we're gonna be really, really focusing on research and data enrichment and also writing personalized messages. Um, but all three of those things are possible. Um, okay. So in now that I have my my dataset in in play, what I wanna do is use AI. Um, so I'm gonna add a column here and I'm gonna add enrichment. And I'm gonna say, um, I wanna use GPT to help me create some text. Um, and what I'm gonna do, actually, is actually, this is taller than I'm gonna go add GPT, um, generate text, this one. Um, and I'm gonna go ahead and add a prompt here. So very similar to how you would prompt your regular GPT. And I already have a prompt that I've put together. I just copy and paste it here. I'll walk through what this looks like. Um, so I'm saying you're a sales assistant for Goldcast. Your job is to generate highly personalized follow-up emails to all the webinar attendees. Um, I am giving the, um, my GPT some event context on what happened in the event and what we talked about. And then here's where I need to map some data. So I have my attendee name. I'm gonna do a backslash and pull in, um, that person's name, uh, which is here, uh, the email of that person, um, which is over here, uh, the question that they asked, which is here, and then their company name. Okay. Um, so now that I have all this information, I'm giving them some research instructions. I want you to research the company name and look for, uh, uh, you know, any notable company news, any marketing strategies or industry focus. Um, I want them to, uh, really make sure that they tie in that relevant details from our webinar and then make sure that there is, um, the, uh, alignment to our Goldcast point of view, which I've listed out here. Um, and then I'm giving instructions on the output. Um, I only want I want two fields. I want a subject line, and I want a body copy that is personalized based off of what this person asked in this event. Um, and so down here at the bottom, in order to get those fields, I'm gonna add two fields here. So the first one here is the subject line, and then the second one is the body copy. Alright. So I am gonna go ahead and save this now and run on the first 10 rows, and we'll see what the output looks like. So now it's it's working. It's magic. AI is it's finessing. It's composing. It's fine tuning. Uh, we have some responses here, and I'm excited to show what these look like. Um, so let's let's take a look at, um, at this one. Um, so there's the, uh, subject line field that I I like I wanted to to create and the body copy field. So I'm gonna go ahead and add that as a column and then also add this one as a column. And so here, I can go ahead and open it up. Um, so this person in particular asked, um, if they could use an external four k camera, um, in our events. And so this body copy email is specifically addressing that particular questions to and and have a very, very specific call to action about having a brief call. And so you can, you know, adjust your prompting to, you know, match your goals on your event follow-up side of things. Um, but this this looks pretty good to me, and it also has a, uh, personalized, um, you know, subject line as well. Um, so that there's a this is a very, very basic example of how you could take, uh, different fields within a Clay table, add some enrichment in your email follow-up with GPT, and be able to parse it out into a subject line and body copy. Um, Ziqi, anything you would add here? Um, Yeah. I think, like, this looks great. Additional thing I would add is there are also, um, many ways you can incorporate additional data points before even drafting a follow-up email. So for example, in this in this situation, you might care a lot a lot about, uh, whether this uh, prospect is already talking to someone else on your team or whether they've gone to your website previously or whether there's engagement with LinkedIn. And, um, ideally, those data should be on your CRM Salesforce already. So what you can do is when you are drafting the prompt, you can tag, uh, those fields on Salesforce. So AI not only has contacts about who this person is, uh, which company they work for, and using general web research to come up with the, uh, actual subject and body line. But rather, they can incorporate your first party data. So even, you know, uh, the engagement of the person, uh, through intercom. That's also one thing you can directly pull into Clay. So, uh, I would just say if if you can get extremely creative, like, anything you have, uh, as first party data as well as third party data, you can pretty much incorporate all of those in a follow-up email. Awesome. Um, thank you for that additional color. So, uh, one thing that you can do, if you're you're happy with this, um, and, you know, we are in in the docs tab, we do have a resource that our agency, GrowToday, has put together to help templatize this. So it's very, very simple to replicate. Um, you can go ahead and actually click on actions and export it and, uh, send it to your your sequencer. Um, so so there's many different sequences. If your team is using Outreach, SalesLoft, I know that, um, Clay has their own sequencer that they've launched to help with, um, you know, this type of use case. Uh, but you can directly integrate with those tools, uh, and really, really be able to make these very, very, you know, one step simple. And it's not like a very, very manual process, which I know that a lot of marketers have to handle, um, usually. So there's a lot of different ways that you can send this data. Um, what we've done here at Goldcast Events is that we have created two separate fields in outreach, which is our follow-up tool. Um, we have a, uh, field that is clay subject line and clay body copy. And what it does, it just automatically sends, uh, based off of the email address in the campaign into that sequencer, and it automatically sends. So it's very, very simple. Um, that's the very, very basic way of doing it. But something that I have had challenges with, and I'll I'll spend the next maybe five to six minutes kinda talking about this, is many times at Goldcast Events, uh, we have very, very highly engaged audiences where they may ask more than one question. They may answer up to six or seven polls. And I wanna have one follow-up email that encompasses everything that they've done, not just a one off. Uh, you know, I saw you answer this poll. I saw you answer this poll. I saw you answer this poll. And so we also have built out a a more, uh, complex, um, workbook to kinda showcase what that looks like. Before I go and and and show that, uh, I wanna take a look at the poll results here. Um, so I asked the audience here, hey. I could automate one thing about your event follow-up. What would it be? Um, looks like majority of all of us, uh, are saying writing personalized messages, but it's very close to prioritizing hot leads. Um, that's a very interesting, uh, pain point. We'll definitely think through, um, a use case for prioritizing hot leads, um, at Goldcast and come up with a a session specifically to address that. Um, and then some of you, about 30 25% are having challenges with research and data enrichment. Um, so for those that are having challenges with research and data enrichment, um, I think this this particular upcoming use case is gonna be helpful for you too. Uh, so I'm gonna stop sharing here. We'll watch my third and final poll now. So as I'm going through this demo, please feel free to answer it. Uh, for your current events of pipeline conversion, um, would improve most with with what? Uh, better lead scoring, oops, better lead scoring, faster follow-up, or more relevant messaging, or actually knowing what people did at the event. Um, so if you head over to that polls tab, um, help me answer that question so we can tailor the conversation today or future sessions to address your challenges. Alrighty. So let's keep on going. Okay. So I mentioned that this is a a a little bit of a more advanced, um, advanced workflow, uh, to be able to address address not only, um, not not only taking large datasets where there might be duplicates, where one person, they've done multiple things and you might wanna have a single outreach that would, um, address all of the things that they've done, um, but also using a Clay to help you enrich data, do research on your company, on your your particular contacts. Um, so the workflow here is, again, importing that record source from Salesforce, the same thing that we did. This is a much larger date dataset you'll see here. We have over 2,000 respondents here. Um, and then the first step here is to, um, uh, clean up our input. So, uh, clean up the data. I'm not sure if anybody has this challenge, but we do at Goldcast Events CRM hygiene, um, and definitely making sure that we use Clay to help enrich to clean up so that our output is as as clean as possible. And then finally, the output, really taking, um, you know, the input that we've cleaned up, that we've deduplicated, and then, uh, creating very, very seamless output outreach emails. So I'm gonna dive into the input table here. Okay. So this particular, um, follow-up is for an event that we hosted about earlier this month. It's it's called AI School. It was a really, really big event. We had over, I think, you know, 2,000 people sign up for it. I think over a thousand people live attended. Um, and so the first thing that we did is we we did some, um, account cleanup. And so in this, um, Salesforce, we called it oops. We called it Salesforce ID and name. We wanted to make sure that we had, um, our account owners very, very clean as well as the IDs. Um, and so we have a a very particular prompt here that says that you are AI data cleaner. Um, a lot of our output was based off of an HTML anchor, um, string, and we wanted to clean it up. So for example, the, um, dataset for account owner typically looks like this, and we want it to just come out with the actual first name and last name of our account owner. Um, so that was the real goal for this column is to use GPT to help us clean up the data and spit out the output in these two fields. Um, the next is, um, for those that are familiar with Salesforce, um, it's really challenging when you have a Salesforce campaign and there is the concept of leads and contacts and there's the separation of those fields. Um, and so we have the same challenge here with, uh, account name and company name. Um, these are both the same but in in two separated fields. Um, so over here in another, um, like, we're we're trying to consolidate some of some of these account information into a single field, um, and we're doing that also with GPT. And then we also have as you can see, Nicole here asked one, two, three, four, five, six poll responses. Uh, we wanna make sure that we are now, uh, consolidating all of that into a single value. So we're not writing six different follow-up emails, but one follow-up email that addresses, um, her responses. Um, how we're doing that is essentially writing all of this data into another table. Um, Clay, do you wanna jump in and talk about what writing to another table means? Yep. Um, so as you can see, uh, Clay really lives in the spreadsheet table world. Um, in certain situations, you will find it helpful to break it down to multiple table when you are, let's say, cleaning account level data and then you have some contact level data you wanna work on separately. So what Kelly is doing now is to, um, move to another table and then continue the workflow building over there. Um, so the other table I'm I'm moving it to is is this particular output table. And so what I'm doing here is that before that table, you see there's six different Nicole's. Now there's just one Nicole, and it's it's really bringing it kinda like a v lookup to consolidate everything into one single row. Um, and so here, um, I'm gonna minimize and then meet these fields here so we can categorize the different marketing roles. Um, let me this is the challenge with live demos. Um, and so here, it's it's a classification assistant prompt. I'm I'm using them to really based off of their job title because everybody has very different job titles here. Um, help me basically classify them into different personas. Um, so this is the the prompt that I'm I'm specifically giving them instructions to. So there's the demand gen marketer, there's the marketing ops person, the marketing leader, um, and so that's all being organized cleanly here. Um, I'm enriching the company. So, uh, based off of the previous output table, um, making sure that the the company name is clean. A lot of the times in CRMs, it's a dead giveaway that it's an automated sequence when you have a company name like, um, Hatch, a a b to b software company. You really just wanna address it as Hatch. Um, and so that's how we're enriching companies to make it all super clean. Um, and then, uh, over here, uh, we're also looking at company size. So, uh, for us at Goldcast Events, we really, really care about, um, you know, what company size your your company is classified in, whether it's a growth stage company, a mid market company, or enterprise company, uh, because, uh, your pain points and your use cases might vary based off of the size of your organization. So that really, really helps the personalization of the follow-up. Um, so we're also using Clay and different prompting, um, to be able to help us classify the different, um, you know, sizes of organizations. Um, we're also using Clay to help us, um, um, do research. So extracting relevant business insights. Um, so really, um, taking a look at any kind of new news, like funding or new new go to market initiatives, new marketing campaigns, um, anything that that g GPT might be able to research on. Uh, we're having this row kind of come up with the responses, and you can see here, um, Prudential Financial ex recently expanded its essential term suite. Um, so really, really being able to pull that information and use that in your follow-up. Um, we're normalizing our poll questions. So a lot of the responses are, you know, in multiples. So there's a a lot of different questions that someone might have answered in the poll, and we're using this, um, GPT prompts to consolidate all of it and normalize all of it. Um, so then we can have different snippets that really address every single response. As you can see here, there's there's there's one response. A lots of, like and this is a pretty complicated one where we ask a lot of poll questions and we wanna use every single response in our follow-up. Um, and, ultimately, we are reaching this green section here where we've color coded as the, um, the event follow-up email. So, uh, for, let's say, Ashley. Um, she answered a bunch of poll questions. This is her subject line, um, and then this is her follow-up email. So notice your poll response about, you know, all these things, did some research on Prudential Financial. We thought this would be helpful. Uh, would you like to check about how other directors of GenAI are are are doing all this? Um, so this is a very, very personalized follow-up email based off of the poll question that she responded to. Um, and then we have a workflow that sends this directly into Outreach so that our sales team can automate this very simply. Um, Rishi, anything you'd wanna add in here? Kinda went through that really quickly because I wanna make sure we have time for questions. For sure. I think this looks great. Uh, I love how many New Year's research you are doing until, uh, you reach the actual generation email part. One thing I would call out, I know there are some questions already in the chat around data enrichment and the coverage rates. Uh, so the unique thing about Clay, as you can see, uh, extremely easy to build workflow and to leverage AI in the process, uh, as well as using your first party data. One other thing aside of everything else is Clay essentially is a data marketplace. So think of us as a, um, data clearing house where we aggregate 150 different data provider. Um, Kelly, if you you don't mind clicking actions, I just wanna yes. That is great. If you type work email or phone number, you can see that, um, for whichever data source you want, you can directly search in the in the chat, uh, in the search bar. It's actually a natural language processing. So, uh, if you search number, search email, uh, it will come up. And if you click work email, uh, at the top, uh, and click full configuration, We recently made the the view simpler because, uh, it's actually quite technical at the back end. You can see we have a list of seven different seven, eight different vendor, uh, already been configured to go through the email enrichment. And you only get charged the credits when there is a result returned. Um, in that way, you're really maximizing the coverage, uh, with the upside of, you know, having gone through every single vendor without the downside of procuring money or being charged by money. Uh, we also conduct internal audit ourselves in terms of provider coverage and accuracy and rank the vendor accordingly. So the best performing high quality vendor would be ranked the first. Uh, when you search for a phone number, for example, you could actually see that, uh, we have different region, uh, list out differently. The reason is different region, uh, different vendor perform differently. Uh, some provider is really good at APAC phone number, uh, some Japan focused data provider versus some other vendor is really good at, uh, UK data or France data. So that's the way that's called water for enrichment in a Clay world, but, uh, it's really a strong advantage when it comes to actually getting the best data and making sure that the email you send will not be boxed. Awesome. Thank you, Ziqi, for that. Um, so, yeah, this is a this is a use case that I've been nerding out about over the last, uh, month or so and and working with our agency growth today to really perfect. Um, if you are interested in this template, a resource called the webinar follow-up template that our agency had put together, um, for you to check out, um, I also highly recommend growth today if anyone's looking for an agency that is, you know, expert in play. And and there's, you know, many other agencies that play resources as well. But, um, growth today is is our partner here at Goldcast and and is a a big, big credit to helping me put this use case together. So if you're interested in this use case and you'd like more expertise in play, uh, do reach out to Brooke today. I highly recommend them. Um, let's go into some questions. I I see there are a ton of questions in the in the chat here. I'm gonna move them over to the q q and a tab, um, and I'm just gonna go through the q and a. Great. Um, let's see. So share. So Adam had a question. Wondering if there's a q and a process at the end of this workflow for catching any AI mistakes, wonkiness, and copy, and how that would be scalable with a large event list. Um, Ziqi Yeah. Um, can you hear me? I think the Internet is a bit wonky here. Yes. Okay. Great. Um, so Yes. I've actually, uh, created workflow where use AI to generate a first copy, and then you use another AI to do, uh, proofreading. So you basically have two AI agents. Uh, one agent has a role of copywriter. The other agent is, uh, copy validation, uh, at the end. Um, Yeah. That's it that's been working pretty well. So if there's any obvious and minor mistake, it will be it will be caused by the second agent. Cool. Um, Adam, let us know in the chat if that helps, uh, with your your question. Alright. Next question, uh, from Todd. Will you share the prompt? Um, I will definitely share the prompts the prompts that I've used in my follow-up. Um, they should also be in the template that is in the docs tab. Um, but my little trick that I use for prompting is that I actually use GPT for reverse prompting. So I'll actually tell GPT, I want you to help me write a clay prompt that will help me do x, y, and z. It'll come up with something really, really good. So that that's how I kind of gain the system a little bit, but I'll absolutely share some of the prompts that we've used in this work this particular workflow. Yeah. Uh, another thing to highlight is, uh, prompting very easy. You can use if you use Clay, you can write a very simple prompt and it might generate a long essay of prompt. Um, I highly recommend to give it a try. We also recently launched something called Sculptor. It literally was launched last week. So you can, uh, without building all the workflow yourself, you can chat with AI to directly tell AI what you're trying to do, and they will suggest table build. Pretty game changing for especially for, uh, beginners. Cool. Um, Ayushi, glad to see that you also use the same process. The reverse prompt generation is is definitely something that I've been guilty of doing. Um, okay. Melanie Question about play in particular. Uh, so Melanie has a question about Clay. I see on the website that Clay helps with CRM hygiene too. Can you help can you talk about that? Yeah. Absolutely. Um, so as I mentioned, uh, because of our integration with 150 different data provider, especially majority of them, you do not even need to have third party, uh, you know, procurement with those vendor. Uh, in the case of work email, you have, like, nine, uh, data provider at the at the back end actually enriching with phone number as well, and you have third party verification provider afterwards to verify the results. Because of that, people really see improvement with Clay's data. Um, we we've, like, helped OpenAI three times their enrichment rate and Endropic, I think, number is, like, double their enrichment rate. So, uh, the CRM hygiene is actually one of the biggest use case with larger company. When they come to us, we always recommend doing a data foundation, uh, fresh up before, you know, considering any more complex workflows. And and in terms of the client, it's it's quite straightforward. You directly import a list from Clay. You can keep the, uh, import process happening every every hour, every day. And then after all the enrichment happen, you can directly sync back to a CRM. So it's really seamless when it comes to, um, you know, having the reps using the CRM data. They do not even you could touch click. The data should be clean and, uh, updated. Great. Um, so our next question is from Ashish. If a all if an already converted customer ends up attending a webinar, how do we ensure the personalized follow-up is not like a prospect? Um, so achieve my prospect or a customer, um, and then make sure that you tailor your prompts to customize your outreach based off whether or not, uh, that field indicates it's a customer or a prospect. So you can be very specific and not treat a customer like a prospect. So that would be my recommendation there. Any anything you've changed, Ziqi? Yeah. Exactly. Like like I mentioned, the real magic of Clay is you can incorporate any first product that you have when it comes to drafting personalized copy. So as long as you have that indicated in your CRM, uh, you you you should, you know, make it clear to the AI agent who's generating copy, uh, to consider the role of the customer, uh, uh, or the prospect while you are where they're drafting the message. Um, question from Grant. Um, are there restrictions on which AI products Clay can use, GPT, Clod, etcetera? Um, not aware of any restrictions, but Clay connects to ChargeGPT, uh, openly and claw anthropic, which is clogged, uh, Gemini as well. So you have access to all of these through CLE credits directly. There's also Cheng, which is our own native AI agent. Um, with depending on the task or the, uh, prompts that you provide, we actually recommend the best agent for that particular task. And there's also different, uh, credit consumption depending on which model you use. It gets really complicated. Definitely a new rising expertise, uh, for whoever has it, uh, but we made it easy by automatically suggesting which one you should be using. And you can also bring your own API key. Cool. Very, very powerful. Next question from Todd. Um, this is actually gonna be the last question that I'm gonna be able to answer today because of time, Um, but we'll definitely make sure that if your question wasn't answered that we get back to you. Um, so from Todd, for Cheng, can you give an elevator pitch on the advantages of Goldcast? Todd, I'm happy to. Um, so if your team runs a lot of digital events, um, it is it it and and and events are a big part of your revenue strategy. Um, Goldcast helps make consolidating those data points really, really simply, especially use, uh, with the integration with Salesforce and being able to bring those signals to clay. Um, so we collect 16 different unique engagement points. Uh, you can find those engagement points in the docs tab onto the right hand side. Um, and being able to actually take those insights and make them actionable to help you drive those revenue, um, so you can cut it back to your event programs, uh, for where where pipeline came from. What's also really powerful about Goldcast is that a lot of content gets created in these digital events like this webinar. Um, Goldcast uses AI to help you take the that those pieces of content and repurpose it into other assets like blog posts, like short form clips, like your social media calendar. Um, and we are releasing a lot of AI agents to help you make video editing extremely easy. So being able to prompt our agentic editor to, um, like, with natural language to add zoom in, zoom out dynamic captions, background music, b roll. Um, so you, as a marketer, can also be a video editor and creator, uh, and be able to repurpose a lot of your content and, um, use it for a much longer shelf life. A lot of the time, this, um, these webinars just sit on a shelf for about forty five minutes and, uh, um, you know, who has time to watch forty five minutes of webinars. Um, and so we help you kind of parse that up cut up into the key moments and share it out to the world so people can consume your content and help build mindshare for your company in a very, very easy and simple way. Um, so that's a little bit about Goldcast, and, uh, really, really excited to that everybody is so interested in Clay here. Um, if you do want to learn more about Clay, um, reach out to Ziqi. She's an expert here. Um, if you head over to the speakers tab on the top of the navigation here, you'd be able to find her information. But, Ziqi, how how should people look for you? Yeah. Just welcome. There's also her email address. Are you on LinkedIn? Awesome. Um, so look for Ziqi on on over email or over LinkedIn. This was a really, really great engaging session. Uh, many more to come. We've got some really great sessions upcoming. Uh, let me just pull up what those are. Um, so the next two sessions are, uh, let's see, about that's, uh, dark funnel to close one. Um, so we're partnering with, uh, our partners at Vector, uh, to learn more about how we can, uh, de anonymize traffic to really be able to, uh, drive pipeline. And then, um, at the October, we've got one about, uh, registration revenue, uh, critical touch points. I think that one's with Sendoso. Um, so we're really trying to figure out how we can use gifting and engagement to really accelerate revenue. So thank you everyone for joining. Again, this is recorded. Um, we have the upcoming sessions, uh, linked at the bottom of the stage here, so feel free to click there and register for those sessions. Um, but if anybody has any other questions, uh, we will be available to you. Find us on LinkedIn. Um, but have a great rest of your day. Thanks, everybody.