hey guys in this video I'm going to show you an easy method to help you generate many leads using an open-source tool that works just like notebook LM notebook LM is one of the most underrated free AI tool right now because it allows you to do so many things if you have never used this before just go to notebook lm. google. com and then here click on create new and then you can upload many types of documents like PDFs audio recordings uh websites urls or even plain text and then create a super realistic podcast about this data or just query information about it in this video want to unlock a new way of using a free tool like notebook LM to generate leads so if you're a business owner a coach a consultant or any other professional that relies on leads for your business you are in the right place as I said notebook LM is a great tool it lets you generate insights and very good podcasts out of your own custom data I myself generated a podcast about the ROI of AI applied to businesses that I will soon upload on YouTube but the thing is that it's still not a perfect tool that can satisfy all types of business needs so if you go to notebook LM Discord server and then you click on feature requests you'll feel like it's endless I mean people want so many new features to be added to notebook LM there is even a guy here who used notebook LM to analyze the future requests about notebook LM for for me the biggest flow about this tool is that it doesn't have an API meaning I cannot integrate it to external systems like websites slack channels emails crms or even Ecom online stores if that was possible this would open up so many business opportunities I'm happy to say I found a free and open- Source tool that works just like notebook LM in the sense that it uses PDFs URLs and other types of input data to generate realistic podcasts in seconds that tool is called podcas fy.
a I think it was advertised by its developer on Notebook lm's Discord server his name is tharis Suza so shout out to him for creating this tool when I saw it in the Discord server I wanted to test it out by myself I mean to just judge how good it is the result is that it's a pretty solid Tool uh not only the generated podcasts are pretty realistic but uh it is customizable in so many ways let's listen to some AOS let's first hear the beginning of this podcast which is about Japanese art welcome to Japan Arts podcast because art is everywhere hello everyone and welcome hey everyone so excited to dive into some powerful Japanese art today absolutely we've got two incredibly evocative pieces to discuss that iconic Great Wave right hocus eyes Masterpiece oh The Great Wave it just grabs you it's so d Dynamic the way those boats are almost swallowed by the wave it reminds me of uh those over pretty good right now let's listen to a small extract about Taylor Swift and monaa imitates life you know and vice versa like look at Taylor Swift a modern day Mona Lisa oh interesting comparison uh elaborate well both of them right they're iconic the Mona Lisa this Timeless figure of beauty and Taylor She's the voice of a generation you know Millions adore so now you understand why pod cify is so powerful because first it allows to create you a realistic podcast just like notebook LM but also you can customize it in so many ways you can change voices uh or even voice providers and use 11 labs for example if you want you can integrate it pretty much anywhere else you can even change languages if you want I know that's a pretty popular feature request now let's talk about the actual methods to generates the leads if your business website has visitors already and you feel like you could do better at capturing the name and email of your visitors to transform them into clients later on pod cify can help you do that exactly this method is very simple what I thought of is to add a little form in the corner of your blog posts that says no time to read receive it now as a podcast the visitor who doesn't want to read your blog post or simply can't read it now if they are driving or doing some chores would want to receive it in the form of a deep dive podcast in their inbox and then the lead will click to open the form to enter their name and email not only will they receive it in their inbox but at the beginning of the podcast it will mention the name of the lead for example if the lead's name is Jackie one of the speakers at the beginning of the podcast will say this welcome to Tech crunch a podcast exclusively prepared for Jackie to be honest I'm super excited about this idea because we are effectively turning every single blog post into a customized lead magnet I think this method of generating leads has many Pros first it's Unique I have never seen something like that before and tools like notebook LM and podcas ify unlock this new possibility second we create a stronger connection with the lead also we know the lead was interested in some specific content because he can be tagged and we know he comes from a specific URL and Via a podcast meaning we can further nurture the lead with the content of the same type later on and finally the people who do the effort of filling name and email to receive something show a stronger intent than other visitors now even though I'm super haed about this new method of generating leads I think it's adapted to only some specific types of businesses I think the ideal type of business who will benefit from this method is a business that has a website with a blog post section and this website needs to have at least 5,000 visitors per month the whole purpose of this method is to capture the information of some visitors and if nobody visits your website well I think it's not worth using this method here is the list of business types and professionals that I think will benefit from this method the first one media companies I think it's pretty straightforward those type of companies need to capture attention so most of their content are block posts so I think this is a very good use case for them you have S businesses software as a service most of them leverage SEO with the blog section then we have real asset agents with strong online presence most of the realtors that have a strong brand create a lot of content all those blog posts about uh taxes or remodeling or even I mean real asset issues uh they can be turned into leads the same applies to coaches or Consultants who also have strong online presence for example nutrition coaches uh who create blog posts about the perfect diet during summer well they could turn those blog posts into leads the same thing applies to smas and also Ecom stores who leverage SEO and email marketing also think Home Service travel and education businesses could all profit from this way of generating leads okay so first let me show you the end result then I will explain to you how to do the same first we open up this little box here we write a name and an email here the lead will receive a podcast in his inbox and if we listen to it welcome to paddle USA a podcast exclusively prepared for a smile today we're diving deep into the world of hybrid paddle rackets on top of that we have the lead recorded in a Google sheet okay so the first thing that you want to do is to go to my Discord server then click on the resources Channel and then here you should find a link to a Google drive folder where you'll be able to download all the files it is not there yet but as soon as I will upload the video you will find it the next thing that we want to do is to install python so here in the prerequisites it says uh we need python 3. 11 or higher version for this project I personally use the 3. 1.
5 version of Python and then to download and install python it is very easy you will just uh Google download Python and then here uh you will click on the first link here you will find other versions of python but uh just to be on the safe side uh just download the same version uh that I'm using which is this one then after you've downloaded python you need some extra packages to make it work so here we should right click on the folder where you downloaded your files and then here you will click on open in terminal and then here you should write pip install dasr requirements. txt and here are the packages that will be downloaded uh so it is fast API podcas ify of course requests and UV corn the next step is to prepare our API keys because we need to write them in this EnV file so first you need your open AI API key if you have it already just space it here if not uh you should go to platform. open.
com here you should click on your profile then click on user API keys so you can see my API key here but you can create your own by clicking on this button the second API key that we need is the one of 11 Labs so you'll go to 11 Labs website once you log in you will click here and then you will click on API Keys here you'll just click on create API key it should pop up and then you will just copy it and then paste it in the EnV file finally the last API key that we need is the gini API key so here you should just uh Google get gini API key you will click on the first link and then here there will be a button that redirects you to uh Google AI Studio then click on create API key it should pop up and then you will copy it and paste it in that file the main file that will be needed is the this file that I wrote uh uh it's called podcas ify to. py Uh but before that we need a way to run python files so here I like to use vs code as the IDE you know to run python files and we need the right extension to run python so click here on extensions and then you will just write uh here Python and then you should download the python and python debugger extensions once they are installed you should have a play button Buton here uh which just means that you can uh run Pyon file I will go quickly over the code just to explain to you how it works um basically here we have a first function that gets the last podcast that was created because all the podcast that will be created will be stored in a folder in data/ audio so here this function will just return the file path of the latest uh podcast created here I created a dictionary that uh does a mapping between uh the page that will be loading and the URL from where the podcasty will fetch information and create the podcast so here I created those keywords just to inform podcas ify about the name of the podcast so here for example if we create a podcast from this URL at the beginning the podcast will say welcome to Tech Crunch and then I created those different end points this one is the homepage just to test whether the server is working or not and then those are the three different web pages that I created I mean I did not create them I just uh took them from an existing blog post and I adapted those web pages so that they will fit uh you know this lead generation method so if we create the server and then we go over Local Host SL food article it will load a page about uh the top 10 Nutrition trends of 2024 and this is the main function that does the job of uh capturing the lead information creating the podcast and then sending it to make. com so that I will be sending an email and recording the new lead information into a Google sheet the good thing about podcasty is that you can customize so many things so here you can see that the podcast name changes whether uh we load a page about food or about AI and then you can see that I made the tagline Dynamic meaning that if the lead name is Jack here at the beginning of the podcast it will say a podcast exclusively prepared for Jack and here this is the actual line of code that creates the podcast you can see that the same line of code is commented out here and there it's because it is using a different model in terms of uh text to speech I found that 11 Labs is actually the best model model to uh generate podcasts in terms of realism but you also have the option to generate podcasts using open AI by using this line here or to do that using Gemini multispeaker model by using this line so for example let's say uh you want to test open AI model you will just comment this line here and then uncomment this one here and finally this part of the code sends uh the audio data as well as the lead information to meg.
com web hook which means that this URL that you can see you need to change it and meg. com will generate a new URL for you you'll just paste it here so you can see that in addition of the audio file that we sent to make we also send the name of the lead as well as his email and the URL now before we run the python script we need to create the meg. com blueprint so here the first module that we need is a web hook so here select custom web hook click on ADD and then save here this link you will just copy it and then paste is here first we need to send some data to me.
com so that the web hook will know uh the data that will be received basically because we need to use that data later on meaning we will do a first run of the script but uh the entire automation will not be complete yet now let's run the script by clicking on this button here if every everything has worked well you should see a message UV corn running on and so on if you need further double check to see whether it is working just go to the homepage and to do that just write Local Host colon 80 and you should see here working now let's test the three articles that I've told you about and see if they are loading uh properly to access the first one you should just uh write slash aore article and then once you hit enter here it is it is actually a block post from Tech crunch that I saved so you can see that here we have this little Banner if we click here we have the name and the email so in order to receive the data in make. com let's run the scenario here in the form just write a name and then an email and then click on yes please here if we go back to uh vs code you should see generating podcast okay it should take some time you know to generate the podcast at just measured it took uh 1 minute and 42 seconds so don't worry you know uh if that loading Circle keeps you know loading uh it's it's perfectly normal and then here you should see latest podcast generated and then you know uh the path of the podcast also once the podcast is generated it should take some uh extra time you know to send it to uh make. com I mean unless you have a very fast upload speed now let's see the data that was received by make.
com if we click on this bubble you can see here uh the name of the lead uh his email and the URL source and here you can see its data so what you see here is called hexadecimal you don't need to know that uh just know that what you see is uh a preview of the audio data that was sent and received by meg. com the next step is to send that data via email so we need a new module the Gmail module more specifically the send an email action here you need to create a connection between uh mega. com and your Google account so that mega.
com can send emails on your behalf and if you have not done this yet uh you should just click on this online help here it says click on create a connection then select a Google restricted and then once you click on Save uh popup should appear and then you'll just click on allow basic basically uh you will allow me. com to uh get the permissions then here click on add a recipient for the email address we just select email here for the subject let's write hey and then um the name of the lead comma your podcast is ready and then for the content I wrote here is your freshly baked podcast I hope you enjoy it and then very important click on add an attachment because we'll be attaching the audio file to our email here for the file name I like to add some extra customization we'll call it podcast 4 and then the name of the lead mp3 it is very important that you write MP3 otherwise it will be um the file sent uh will not be readable basically and then for the actual data that will be sent you will select this data here and then you will click on okay finally we need a way to to record that lead into a spreadsheet so we need a new module here click on Google Sheets and then here click on add a row so this is the spreadsheet that I prepared for this tutorial here we have six columns we have the columns date time the name of the lead uh the email the URL as well as the tag so you will click here to choose the file then you will write its name select sheet one here of course we should mention that the table contains header so just keep it as yes here okay so for the date and time we will format them in a very specific way but uh will be using mega. com builtin functions to do that so for the date you will click on this little calendar icon here and then uh you will click on format date so this function takes two parameters the first one is a date and then the second one is the parameter that we want to use uh to format that date into a specific format so the first parameter will be now it is a variable that is accessible in me.
com it means the current date and time and here for the format I will use the US format which is day then month then years so here you should write dd/ mm sl4 * Y and then for the time we'll do almost the same thing but here instead uh of this format we'll just be selecting uh the hours and minutes so you can see that here if we want the US you know time format we should be uh writing HH lowercase colum mm then space uh uppercase a okay once we're done with the time let's click on name here and then select the name of the lead for the email let's select the email the URL the same thing and for the tag I just like to write podcast and then click on okay okay perfect we have the m. com automation now let's test everything end to end first let's run the. com scenario and then if you haven't done already just reload this page then let's write uh the right data and then click on yes please okay so now uh it has started let's wait for me.
com to receive the data okay so in the meantime meg.