AI Sales Automations with Zapier

[00:03] Hi, my name is Charley and I'm super grateful and excited to be sharing with you my passion for AI automation. Recently, I've been leveraging this incredible tool called Zapier. And wow, I must say it is absolutely amazing. From being able to rapidly deploy these automations to be able to create zaps and connect different tools and technologies, it is truly endless. Anything that you can automate, you can do it with Zapier. So with the initial zaps that I created was from the standpoint of a solutions architect. Typically when we go through and we are collaborating with account executives as well as end customers clients. We want to ensure that we are gathering the necessary relevant information regarding previous calls or what that company does or what their

[00:48] interest is as well as updating our CRM system and being able to communicate via email or notifications. So the initial first flow is from the standpoint of when you are out at a conference or you are collaborating in person and you come across someone that provides you their business card. And so when we receive that business card, we want to get that information into our CRM system as quickly as possible as lead. So we do this by initially sending a message in a Slack channel. This Slack message will go through and to post an image. As you can see here, we posted an image and we will upload this to our Google Drive. And so you see our Google Drive here. It

[01:33] uploads it to store it for persistent storage. And we leverage a tool called Nanonets OCR. So what this does is it performs an optical character recognition and extracts the relevant information and text from that business card. And we then leverage Zapier's code tool to write custom Python and or JavaScript code to format or do various things with the data. So in this instance, we are specifically extracting the email and parsing out the URL and utilizing that as the company domain where then we then search our HubSpot CRM system to see if we have that company in our CRM. If not, it will create that new company.

[02:18] We then want to go through and leverage the very helpful formatter tool. In this instance, we are splitting the name for the client into a first and last name by eliminating the space. We then leverage that relevant information to find the contact in our CRM system. If they are there, great. If not, it will go through and create a new contact. We then feed all of this information to our agent. This agent leverages its various tools to go through and perform an analysis and research on who this client is, what their interests are, what their company does, and it gathers as much information

[03:03] for us to identify who this person is, and we then update and create a note in our HubSpot CRM system for us to be able to then reference later. as you can see here. So this is super helpful. So the initial people who are leveraging the CRM system or doing the outbound engagement have information on who this person is and notes to be able to Then what we do is we are then adding this contact specifically to a HubSpot lead list. And so this lead list would then be leveraged for account executives or for the solutions architect or business development reps to follow up with that account. Once that account is being in communication with or the

[03:49] account executive or someone that is gotten in contact with them, they typically will go through and schedule a discovery call. So in this initial discovery call, we go through and utilize the Google forms. So all of the account executives can leverage this form. Specifically, they will go through and fill out a discovery request. And this asks them basic questions like what is the company's name? Who's the point of contact, when is the discovery call, and any additional information. We then leverage Zapier's incredible tools to go through and specifically filter out our results. So in this instance, we're filtering out to ensure that it was a discovery call. We split the name just like we did before. So we split out

[04:35] the user's name and we identify if this person and company are in our CRM system. If not, of course, we'll go through and create it. We then feed this information to an agent. This agent again utilizes its tools to go through and extract relevant information to be able to uncover what this organization does with the information that the account executive gave us. And it provides a deeper analysis on what that organization does, the things they care about, how they operate. And it does this by leveraging the different tools Once we complete that, the agent returns and we create that note in HubSpot as

[05:21] shared. And then we leverage the code by Zapier again to customize a specific function to be able to go through and leverage the initial start stop or the initial start and date time of the discovery call. And we convert it to a start stop time leveraging the proper time zone. This leverages information to go through and query our Google calendar where it identifies if I'm specifically booked for that time. If I am already booked, it will then send a email response to the account executive telling them to create a new discovery call at a different time. And if it I am not busy

[06:07] and available then it will go through and create an agenda where this agenda will then be attached to the calendar invite as we can see here. So this would be the initial discovery with our agenda that the client sees and they can of course join. super powerful to be able to develop the automation to promptly schedule the meeting as well as gain relevant information for what that client cares about. Next is typically when an account executive would bring a solutions architect into the call. So this one flow or this zap is more heavily utilized and this zap specifically is

[06:53] the demo zap. And within this zap we are of course utilizing the Google forms again and we are filtering the Google form to ensure that we are only using the demo intake. So with this demo intake, they will submit similar information like company, point of contact. This time they will include a meeting link. They will include any additional insight for the processes, the softwares and anything that they have knowledge on and schedule the demonstration. When we schedule this demonstration, we will of course curate information regarding information in our HubSpot and extract contact and company. Feed this to an

[07:40] agent where we will perform a research on that company, what they care about. It will specifically target and focus the responses that the account executive input to ensure that it is targeting relevant information. We then extract utilizing the code by Zapier to get the URL meeting link. specifically in this instance, I received this file from a Google drive where I had an audio recording. Of course, you can utilize a zoom or a gong recording for the transcripts. but we leverage the ChatGPT transcription to create that. And we feed all of this information to an agent. This agent specifically is doing a discovery call analysis. So, it takes in all of the

[08:25] previous research and information from our HubSpot and the meeting recording that was transcribed and performs a deeper analysis. We run some custom code to do the time start stop for the demo. And we of course check the calendar again to identify if I am available. If I'm not available, it will tell the account executive to reschedule the meeting. If I am available, it will utilize AI by Zapier. This is an incredible tool where you can leverage any large language model like OpenAI or Gemini. You can feed it relevant inputs and a specific prompt for it to output a specific response. And in this specific response, what we are having it do is

[09:13] create a agenda based upon the initial discovery call that was sent. That agenda is highly targeted then and leveraged to specifically create this Google calendar invite where it is then updated and the relevant agenda is attached. Once this demo is scheduled, we want to then create a deal in HubSpot. So this is to initiate the tracking of this opportunity now because there is a demonstration scheduled. So there typically a quantifiable number that is attached to that. So this is the setup with followup is the demo and we include the relevant information for this with

[09:58] the start and stop date. Once we do that, I send myself an email notification just for relevant information to be notified that hey, a demo has been scheduled for me. That's super powerful to be able to take in the requests from clients and the account executives to schedule a demonstration, be able to extract relevant information out on the web as well as from the account executives form and be able to put together a proper relevant demo communicating around the information that they need. Next, we then ultimately need to go through and follow up this demo. So when we follow up this demo specifically I would be entering in information and the intake form again

[10:45] especially in the follow-up section relevant name information I update the deal size with what I would expect it to be as well as the expected close date and any notes that I recorded and followup links or resources that the client may want. When I include all of this in the response, we have this run and of course we filter it to ensure that it is the follow-up form. We extract the meeting link, extract the audio recording and create a transcript from that. So we create this transcript from the demonstration that we gave and we then utilize our HubSpot CRM system to curate the information on the company and the deal. And we also

[11:30] look up data in our Google Sheets where we then feed this to an agent where it will create a draft message for a follow-up demo. And so this will then format the deal size. specifically when we were extracting this, the account executive put a dollar sign in front of it. So, we had to remove the dollar sign with some of the formatting structure and we ultimately needed to remove the comma because HubSpot does not want a comma in the number that it is putting for the deal size. So, for us to be able to do that, we utilize the formatter and code and we ultimately updated this deal in our HubSpot CRM system. So once we sent that follow-up message and we had the demonstration, we

[12:15] want to ensure that the deal is progressing and so we include an amount. We include a close date to that and we of course set the next steps to send a contract. Super powerful to be able to ensure that you know I'm communicating properly to the to the client and following up with relevant information on the conversation that we just had in the demonstration. Ultimately though, there may be clients that may specifically have some additional questions or requests before signing the contract. And so that is when I leverage the info zap. So this info zap does an incredible job of intaking specific requests from the account executive. the account executive will go through and fill out

[13:00] this form again. So they would fill out the info request and this info request similar type fields. specifically though this one is asking what their request is where they can copy and paste the question. Once we filter to ensure that the condition has been met, we go through and we feed this information to create a HubSpot task. So specifically in this case, I want to ensure that you know I'm staying on top of these tasks and they don't fall through the cracks. So I create a task in HubSpot to identify that I need to follow up with that person. So I specifically did that here and this specific task here is the one that was created and I can of course

[13:45] follow up with their question and provide a response to that. Then I want to identify and have it do some deeper research on you know what is this client asking what is the relevant information that they are requesting or needing. So, I give the information to an agent. It queries Google, searches our CRM system. It looks all of Zapier's documentation and help docs and the community portal to find the answer to that specific question. This will then generate a customer email leveraging AI by Zapier and the OpenAI large language model. we feed it a specific input and prompt and it will give us that pre-formatted

[14:30] email. Of course, we want to ensure that we have that checkpoint or that human validation in it. So, in this step, we specifically sent a request to that account executive or the solutions architect and want them to validate this email. And so, in this case, I did not validate it. So, I said, "Hey, I need it to be more technical." And so I declined it because it was not approved. If it was approved, it would go to the and send a specific email, but because we not approve it, it sent it back to do another rerun through the agents to find more relevant information to respond in a more technical or whatever the request was play. And from there, we leverage AI by Zapier again to generate the

[15:15] customer-facing email. And ultimately we do another step to feed in that email and generate a subject for that email. We then leverage the Gmail tool to draft that email for us to be able to review and put our time towards that for sending. From that standpoint, I've really taken you through the entire life cycle of being able to ingest a lead from a business card, be able to schedule in discovery calls, demonstrations, as well as being able to follow up these demonstrations and even send relevant information and answer questions upon their request. And so these specific zaps have been incredibly helpful to

[16:00] automate a lot of the communication as well as research and information gathering that is required for me in the solutions architect role. And it assists with the collaboration with account executives and updating the CRM system to move the deal across the finish line. And it really overarching helps the client and the end customer get the relevant information they need in a timely manner. And so truly Zapier is a fascinating platform. There's so much to do. I only touched the surface. There are so many things like custom actions. There's tables, canvases, a chatbot. The list goes on. And so I'm really excited to dive further into this and explore more. And I hope this video has been helpful. And I know it's been fun to build. So thank you so much.