Python for Marketing 101

So for every entry in the end um in this response, hey please print me out the name of the keyword and um the popularity like the keyword volume. So fingers crossed I hope it works. So we are seeing that here on the right. So it's doing something and voila. Whoa. Awesome. It's working. So and here we see um fitness has a volume of 61 for example. Hi everyone, welcome to another episode of Tangent AI 101. This is Roman from Tangent and today I'm joined by Swen Jurgens. Swen, how is it going? Uh, awesome. Awesome day. Thanks for having me. Fantastic. I'm Swen does UA and he will tell you more about himself, but he also does all sorts of micro tools on Python. So that's was the reason why we decided

to have this uh recording because that's very untypical topic for marketers to discuss. Uh maybe when you can start uh about talking about yourself maybe a small intro. So yeah I'm doing now um model marketing for roughly 10 years already and um I started with Python I think seven years ago. So somewhere 2017 18-ish I would say. Um because back then um yeah we had a BI um and uh he was super smart in doing all these crazy things uh and I wanted to do it the same honestly right um that's the very reason for that and um yeah over the years I was then sticking to Python and became better obviously after some time and um yeah you can do all sorts of crazy things with that uh especially also for marketing and for automation reasons And um yeah, I'm happy to share some of that

today here on the show. Super. Uh, of course, I asked Swen to prepare some slides. Uh, I haven't seen them, so this is going to be also a surprise for me, but I trust Sven so much. So, whenever you're ready, Sven, uh, please share your screen with the slides. Yes, I would do. I would do. So, just give me a second here. Um, I will then share that here with you. So you should see my screen I hope. Yeah. Yeah. Yeah. Maybe. Awesome. Cool. Then I can uh just go here in the presentation mode. Awesome. Um yeah. Um as I said before, I'm preparing uh some uh quick tools uh for you there. Um so that you can actually use um in your day-to-day marketing. Uh we will also share that later on on the blog and uh in in the video section here I think. So

um the idea is how can you actually save some money on some of these expensive tools out there for ASO and Aspire right um so that what we are going uh through today. So uh quick introduction of Python. Um, this is not a like a fully fleshed out beginner course here, right? Just uh uh to to wrap your head around that even if you never have written a good line of code. And then we go through some uh quick examples here. Um so two that are actually helping me every day um to uh fetch some uh keyword volumes on Apple App Store and um also um this is maybe the more important one even that I'm spending quite some time on to see some competition ads on meta and uh you can actually build that yourself um with just a few lines of code. So um intro to Python maybe a quick question because people are

watching us and thinking oh why Python I I can wipe code anything now very good question you didn't prepare that I have a side for that um very very good yeah um yeah when I started out so maybe one step back um so when when I started uh a few years ago um with with uh Python and programming. So I have a marketing background, right? I studied that stuff. I uh never written a line of code in in my studies in university or anything, right? So for me it was really starting at zero and back then honestly I was like hey what can you actually do with that right? So, and the simple answer is a lot obviously. Um, but uh when you have some of these feelings, right? So, um maybe hey you need some some Monday morning reporting every day or um you are paying lots of money for some software tool or some BI or dev guy

never has time for you, right? um then you might want to maybe tip your toe into that and uh start learning how to code because you can save so much time and um obviously also progressing your career. So for me it was a big helping. Um coming back to your question and this is not prepared. So um I had that here um because of all these platforms out there right so hey you can today vibe code um uh as uh some some people call it um so there's like lovable and and bold and all these platforms right so do you need to learn how to code? Um I would say yes still I mean um I've tried some of these tools. um so lovely and and bold and um I'm not bashing against that. So um it's it's cool for what it is, right? So um when you want to have like a prick prototype on something um it can actually save some time, right? So um I also have some some deaf friends

um who use these code uh tools or um some cursor AI or or some other LLMs. Um it makes you way faster when you know what you're doing. But um the thing is when you have never written one single line of code, it's still overwhelming, right? And um I think I I maybe I'm I'm wrong, but I think that at some point you might be trapped because um bug fixing is part of the process and um these tools can only bring you so far as of today. Maybe in five years it's different. No, fine. um always hard to predict the future but um so as of right now I would still say hey it's making sense um to learn hot code and um yeah it's also for you and your career right so I mean just open LinkedIn or stepstone or all these platforms right so people and and companies um still hire people um who have some sort of skill and um not hey

are you good at V coding so I don't see that on job descriptions these days right maybe changing I don't know But as of today, I think um it's still making sense to to learn that. Um so to dig a bit more into that. So what can you like really do uh in in mobile marketing? So in what we are doing here um there's also a great book um automate the boring stuff. Um so I can recommend that by El Swagert. Um also good starting point for me back then. Um and um yeah so what what I usually do um so um automating like workflows right so what what I need to do as a marketer um so you want to know which ads are working for your competitors um maybe you need some sort of uh connection to your reporting API from I don't know Tik Tok or meta ads right so uh Monday morning reports I think we've all been there to to

summarize something for for the CEO or investors. Uh where are we right now? Do we hit the KPIs or not? Right? Um uploading, I mean, we've all also been there, right? So, um most of the UIs are quite buggy sometimes. Uh you upload a video, then it's breaking, then you need to reload, then you upload again, right? So, it's a tedious process. Um ASO text analysis. Um so, you can do a lot of things um with that. Um and also, yeah, user review mining. what they hate, love about the app or don't like, what you could implement maybe in your own app, right? Um to to kind of carve out your niche. So, just a few examples of what you can actually do and it's um yeah, quite a lot. So, every time when you think, hey, um I do that more than maybe once a month. Um you might maybe want to think about that. H um it's tough. There's there's no easy way to

say that. Sorry, but um it is tough. Uh so um I'm doing this now for for many years, right? And still struggle sometimes because I run into bugs and it's not working and then I'm kind of literally biting my key keyboard because it's not not working, right? Um so but still it's it's getting easier and easier, right? So with with all of these tools that we have today, right, Chat, LLM, whatever, some sort of courses out there. Um I think honestly today it's easier than ever before. I mean just think about that how they uh people uh learn stuff 20 years ago right so literally reading books right today you have all sorts of interactivity and and courses and whatnot um so I think today um to to maybe yeah tip your toe into the water uh it's it's quite easy uh to do that and um yeah it's as always right human motivation that uh in your way um but when you um have your first um yeah kind of sense of of achievement and success um then it's also way easier to

stick to that and maybe today is a starting point for some of you would be cool um yeah honest example here um I was sharing that sometime on LinkedIn um that was I found that on my on my um on my Google drive somewhere um that is literally the code that took me a weekend so um down here it's just a few lines here and um as you see in the URL here It's connecting to Google Play Store and um it's searching ROM back then. I was with the card games company. Um so it's um imitating the the search query that you type in in the UI, right? So and to figure that out without chatbt with without all these fancy tools took me a weekend. Um today might be two seconds, right? So um even deploy Python like on your machine. Yeah, even that, right? So it's it's

already a struggle to install all that, right? Because uh it's it's so I'm honest here. I was super new to all of that. I I've never written a line of code. Yeah. Right. So how do you get that? Um how can you install that? How do you go through the onboarding of that? Right. So what is a parser? What is an IDE? All of that stuff. Um it's it's um it's a barrier. It is for sure. But when you crack that, when you when you go into that and you really have at some point this ah moment, right? So ah I I know that piece of thing here. Um it's it's it's awesome, right? It's it's it's great. Um yeah, today it's it's quite easy, right? So you can ask an LLM to to also write way more complex stuff. Um but the drawback here is you don't know it, right? So it's for you still like gibberish, right? It's some random random stuff that you don't know. So I I will still uh encourage to to learn these things. How do you get all of the

ideas like you had this long list of things that you can do with uh yeah Python? Is it like really what you do on daily or weekly basis and then you're like ahuh I think I can do this despite what there is any other it's it's it's really like that. So it's it's for me um and I can say that from a marketing position because I'm a marketer um it it was opening a door to to a world that I didn't see before. So um you will see that later on in in my little environment. Um before the recording I was closing I think 50 tabs in there um because um every time I see something hey I can automate that hey I can do that quicker hey I can connect to this website hey I can scrape that data here make something here connect that here and there um so it was helping me a lot to think in different ways and um also these these areas that you're touching are so connected right so it's

at some point you need to kind of you force yourself to learn also about um websites and HTML and um bit of JavaScript here and there, right? Because maybe you want to connect to some sort of uh I don't know tool out there that is displaying some cool data and you want that and you need to figure out a way how you get there and how you can scrape that out, right? Um and you are kind of forced to yeah not only learn like maybe Python or your language that of choice but also look left and right right so same with marketing right so then you also want to know a bit more about design right and bit more about copyrightiting right and um yeah it's it's kind of opening a whole new world for you and um you start thinking at least for me um I was starting to think in a different way to say hey um how can I look under the hood of thing and how does it work and how can I make it work and how can I get that to work

right so um yeah it's it's I think a super valuable addition to your to your expertise and and to your career and for me I can just talk about myself here um it was advanced my career by a lot and um I think I wouldn't have um landed these kind of well-paying jobs that I had in the past without that skill Um yeah, it's it's super valuable to have that. Yeah, you're a real ambassador of Python. I try. Um yeah, and I'm sharing that for free, right? So um uh also later on on your blog, um you can then click on the link here. So um that's just a link here to my to my Oh, I've checked it out. I loved it. I will leave in the description for sure. Um but yeah, I loved that you had tasks right from time to time. Yeah. Yeah. There there are some exercise in there. Yeah. Um that's also like a starting point for you. Um or also today

um right so when I wrote that stuff out for some friends some years ago um there was no chat, right? So and um so today you can have your own kind of learning schedule, right? So just ask chat, hey, I'm a big beginner. give me 10 task with solutions. Um obviously don't look look at the solutions otherwise you don't learn and um then you can yeah kind of sketch out your your learning journey today right so with your favorite LLM out there so it's um yeah today it's crazy so uh it's it's really cool and if you need a selling point take this one here it it's free I don't know 100 slides or more I don't know um and um yeah going through all the the the building blocks all the installing uh from from a kind of marketing perspective. Um, so because I know how it feels when you sit around that and have really no clue what you're doing. Um, so really step by step, easy, easy, easy. And um, yeah, take that one as a

starting point if you want. It's free. Good. Okay, now we get actually to the to the meat of that here to um the very uh first um small tool. Um but um it's uh still solving some some cool issues here. Um so I didn't found a better logo here. Now it's Apple ads, but you get the point here. Um so we want to read out the search volume um from Apple search ads. Um here are a few slides also that help you to get going. Um again for someone who never did that or these kind of things. Um so it's a kind of a walk through um of first of all how you get there and what kind of tokens and funny numbers you need and um later on I will also walk you quickly through this code. Um yeah to give you a starting point right um if you're starting with um zero knowledge. So um what do you need

actually? So you need an Apple ads account, right? So otherwise it's not working. Yeah. Um also all the big guys out there like Appre and others um I think they not disclose that but I think they have like hundreds of accounts um somewhere to to um read them out because they have like a millions of keywords in the database, right? Otherwise it's not working. Um you need a a kind of big distributed network of accounts uh to read out all of the data. Um but for us um so for the for the smaller team um who wants to um learn Python, learn programming and maybe build a small tool um one account is totally fine. So um you need that obviously then you need um Python installed. Um you can uh go through my presentation. It's also walking you through the installation process and how you get Python installed on your machine. And um yeah, you need something to to say, hey, um this request that I'm making from not the

website is the website. So you need a cookie and um lots of funny security tokens in there. Um sounds sounds tough, but I will walk you through that. It's literally just a few clicks. So in 5 minutes you can get up and running. Swen. Um, does it make sense? Uh, when I like played a bit with Python, I use this Google uh environment. It was web- based. Yeah. Does whatever gets you going. Um, it's it might be a bit of an issue um because it's a web based as you said um when you play around with some uh tokens. So, um many of these tokens are unique, right? So, they are bound to your login um when you're candy crashk king.com, right? So it's it's bound to your uh environment and your company and um you want to double check what you put on a server. Um that's the only security reason. Otherwise um these tools are great and um you can also plug in some I think external resources there or buy some machine etc. um to have some more

power. Um not needed for that case. It's it's quite easy. Um but yeah you can do that. Only thing is security and um uh privacy in the end right? to secure your RB tokens. That's the only thing that you got. And for the cookies, that's basically we'll be simulating that we are real person, right? We're not using the API. We are doing like sort of parsing. Exactly. Exactly. So great has an API um where you can um I mean also other tools are leveraging that API, right? So within their UI you can change some keywords and then they are pinging to um Apple but and that's um yeah big bit of a drawback um also why um other tools out there exist um you cannot read out keyword data from official API so that's the only thing um but also very interesting thing right so you want to

know hey what what sort of search volume has maybe my set of keywords there right so that you put your uh the the best keywords in your store descriptions and the app name, etc., etc., right? Gotcha. Gotcha. So, okay. Um, how do you get there? So, the starting point is quite easy. So, um, just how you actually set up your campaign, right? So, um, for every marketer out there, this is um the the the usual stuff um that you do. Um so you log in, you um create your search um campaign, um choose your geocation, add CPIs, etc., etc. Then on the next tab, you're clicking on here um add the keywords to this group. So this is what what every marketer out there is doing. So no big no big deal here. Um uh what you then do is to I'm uh just putting that here. Here um you are um opening the dev tools. Um also some years ago I

never opened that before. So um so web devs know that in and out. Um I did not. Um for me it was quite new. Um you can do that with every browser. Um so Chrome, Firefox, whatever. Um when you press F12 or right click on any website somewhere and um click on inspect. So then you see some dev tools. Um so it looks something like this here. Mhm. And here you want to click on the network tab um because here you see every call um of this website. So it's quite interesting also when you first open that you might not get something out of that because lots of stuff is happening. Um so u I also didn't think about that before right so when you just load an ordinary website doesn't matter what google.com nike.com whatever right um you will see hundreds of calls in there. So this is quite crazy in the end um because um yeah a lot of stuff is happening in the background right so um

everything that you see needs to be stored somewhere needs to be fetched somehow right needs to be displayed somehow so every logo every piece of word every everything right um and you see all of these calls here that's quite interesting um so you want to open that and here click on the network tab and then um you already see here something and that something is uh what we need to inspect a bit more and uh wrap our head head around that. So now when we are here and this is quite straightforward but until then it's it's one click right. So um when you now type in any keyword here for example in the official UI um for example fitness um then you might uh know that um there's like a suggestion popping up and the popularity with these funny little dots here um and you also see hey there are more calls now being added here right so and that is the interesting part so and um also there

are I think some um browser extensions that you can use uh they are there for for many many years um that you can actually turn these funny dots into actual numbers. That's already better than having dots because Apple is hiding that way. Um and you can um read out everything that you see in the UI here in a structured way. So um then you might see something like this here, right? So app ads.com blah blah blah RP blah blah blah recommendation. So then you have your unique Adam ID is it here um that is bound to your session and um the keyword that you're searching for and um this is actually everything right so until now we have I think three clicks to get there so um and in this call is everything that we need right and if you see already hey um fitness got a popularity of 61 right and um workout of uh 48 for example Right. Um

so that's quite interesting and we just need to grab that somehow and we do it here. So um we want to have that. So in the next slide we can um then do that here. We click right click on this very call um that we want to have. Then we go to copy. Then we want to copy as curl uh as bash. Then we go to this little website here that I have open constantly I think. um it can convert um this command to actual Python code and maybe here we have now six clicks right so it's it's still quite easy when you do that the first time obviously you need some minutes for sure um but when you do that uh on a regular basis it's it's a minute maybe it's it's a few clicks only um really then you want to copy paste that here and here down below you choose Python and um you have code it's working right and with that it's already everything you need and it's already

emulating um the real request for you and that's quite powerful. So um now we get to [Laughter] code. So I will stop that um presentation here. So um I hope that you see what I see. Yes. So you should see the IDE here. Good. So um I was collecting that because there a lot of stuff is happening. So, um, when you ever watched some headers and and cookies, so there's a lot of stuff going on there. Um, most of this you don't really need. Um, some you need, but it's hard to figure out. Um, so easy advice, just copy paste it. What is shown from this converter tool and put it here and um, now fingers crossed it's live. I hope it's working. Always works. Yeah, I will click on go. So um what we do here now is uh we are emulating um the call

from the official UI from from Apple app store and um we are searching now for fitness and um if you've never written a line of code I will walk you through that quickly here. So um you can change this keyword to everything what you need. So health, fitness, game, candy crush, whatever um doesn't matter. Um then it's calling this URL here with lots of stuff that Apple wants to see, wants to have. Um many funny tokens in there, many funny security thingies in there and um then we are iterating over that. So we are getting this back as a response here and um we are iterating over the response. So this is what we do here. So for every entry in the end um in this response, hey please print me out the name of the keyword and um the popularity like the keyword volume. So fingers crossed I hope it works. So we are seeing that here on the right.

So it's doing something and voila. Whoa. Awesome. It's working. So and here we see um fitness has a volume of 61 for example. Um fitness app knows it's free now. um 42 etc etc and um you can do that um with uh every keyword that you want right so we can just do another one here and um again this code will be online on your blog later on so um we can do I don't know football if you wish and then put it here let it run for a second so and there you have it so we have here then football quite good keyword I would football games for kids, soccer games, etc., etc. And um also all of these kind of related tools um sorry related keywords um that are coming out of that and um also some yeah lower keywords um that you might maybe not want to put

directly into your descriptions. And yeah, with a bit more um then again this is a starting point, right? So with a bit more of uh code um that you um want to maybe add there um then you can uh also do that not only for one keyword but also for list of keywords right and iterate over that and um yeah that's it in the end right so with what do you I don't know let's say 10 clicks and um bit of code here now we have like here um okay it's 100 lines of code but this is actually not really code code that you that that you have written before but all sorts of cookie because it's it's super long. Um but with this kind of stuff here you can connect two episodes ads and read out every keyword that you want. So quite powerful and let's reiterate so the the the keywords uh that we got we could not get them in the

Apple search ads. There'll be UI that be hidden somewhere. Yeah. Um it's um kind of hidden um by default um as let's go back here. So the official UI So that's a screenshot from really like the the official um episode here. Um so they are um hiding that away already. So um you as don't even see the real number. Um and um some tools are there um that you can install as a as a Chrome extension or browser extension um that can also they are doing the same right? So they are reading out this structured data here um to um tell you more about that right. So for me it was never enough to say hey it's four dots so what on earth is four dots is a good okay okay that's an added value and um that's already helpful um but um you don't have to

um fumble around with the UI anymore and um then you can uh do that um yeah with with every keyword that you want to check and um with uh then you can add a list there maybe you have like I don't know 50 keywords you want to check 100 whatever right you can do that with every keyword. Um, so obviously it's a bit clunky for sure, right? So I mean that's why these tools like app tweak and all the other ASO tools exist. Um, but the point here is it's free and you can do it, right? That's the point and uh with a bit of more iteration um you can make it uh work better and uh personally it's just my opinion here but um I don't need a fancy UI um so I just need the insights. So that's what I want, right? So, and for the stuff like cookies and was like I think three parameters to authenticate is it like normally what you use when you try to do

something like this or you kind of play with it and try to understand how I mean um how can you connect to a service? So it's um uh I either they officially give you an API. So they say, "Hey, this is our endpoint. You can connect to that. You can authenticate." Should be straightforward. Everything is there, right? So cool. Um then you need to set maybe set up a deaf account. Um then you get your your um API token. It's like a key to lock this door or unlock that, right? So it's it's official solution. Everything works. Awesome. Um that's the not official way, right? because um Apple is not offering um this keyword data um on uh their official API. So that's a good point. That's another advantage of method, right? And uh you get it free. Um so you can I mean you your you your intern whoever um can sit in front of the UI

here and copy paste that out and format that etc etc. So works right but it's kind of tedious and um with these few lines of code um you can do that in a structured way then you can save it as a CSV in Excel whatever and um work your way through that um because yeah other tools are obviously paid tools and there's nothing wrong with that right so when you have a great service you can pay for that for sure but um um sometimes it's also bit of ridiculous honestly right so um when you say hey this CSV export is behind a payw wall, right? So, um yeah, I mean you can do that yourself, right, quite easily. Um or you can also right click on this tool that is knocking the CSV away and see if you get there and surprised often times yes. So, you're hackers, man. I'm not there yet. I I wouldn't say that, but um there are many ways to do that. Um so

um yeah and that's also like another advantage right so um uh you you're not kind of trapped in inside of maybe a I don't know SAS tool that's uh monetizing like a simple CSV export right um got so yeah yeah at some point you see how this stuff works and then you can fumble around a bit play around a bit with that um obviously this is not officially documented because it's not wanted that way So you need a bit of yeah this kind of Sherlock Holmes approach if you will. So dig into that and then figure out how that works. Um but at some point you get it and um it's uh way easier, way faster and saving money obviously. Mhm. Yeah, sure. Good. Congrats. So um we've done that. So, we've been through this process and um now we got our keyword data and um from there um I'm going to full screen

here again and um from there we can do more. So again it's a starting point right? So um we already found the volume so that's already cool right? So um like I said before, you can put that in a list as well, right? So right now we had like uh one keyword, click on go, get the results, another keyword, click and go, get the results, right? You can um do that in a more fancy way that you have a list of 50 keywords and iterate over them. And then um you have uh lots of responses because every keyword also gets you all these rates keywords from Apple. Um so you have a yeah good kind of mini database um for your keyword research. So quite powerful that you have the volume here. Um so then you can also um connect like in this very first example that took me a weekend um to the stores and read out their description text right um from your competitors um

then you can say hey um which keywords are you using and um search for them on Apple right so and then you have another piece of uh piece of the puzzle um for your keyword strategy and the volumes um if you've never done that before here with data muse. Um it's a great tool. It's free. Um can all can do all sorts of of um related keywords and suggestion keywords. So um it's it's in the end uh great for everything text analysis I would say. And um here you could actually um also play around with that. So I'm doing that at at this very moment um to and this is the the reason why these are bold here. So um when you're ASO guy or or keyword guy um you might know this uh three three um ideas here of keyword volume relevancy and competition score

um you can actually build that yourself right so um it is said that uh the the best keywords are the ones with high volume high relevancy and low competition right so and when you can find your uh yeah your your uh space your your niche in there um you might have a good set of keywords and you can build it out stuff. So, um obviously it's way easier to pay some money and go to the big bonds. Um but when you don't want to and or don't have the money, um there are ways to to to go there, right? And honestly, what is relevancy, right? So, um I got too uh too fast here. Um also relevancy um you you need some sort of of uh um of scoring for that and data muse can do that. So they have this um related tag here that you can use and um push for every keyword possible and um they have a big database. They read out all Wikipedia pages, all books, everything and um can say, "Hey, um maybe workout is connected to fitness

and car driving maybe not, right?" And this way you can maybe create a quite good little algorithm to say is that relevant for you, right? Is this keyword relevant? Um because also as as we uh at least think to know um how these algorithms work on Google Play and Apple. Um so the more related the keyword cluster is the better. So right otherwise it's kind of keyword stuffing and and uh if you're like a football game you don't want to put can you crush in there because it's not related right it's totally different. So that's one way to to check for that. And another one over here is data forzo.com. um is also a great um service provider. Um they um have that yeah as a as a service as an API um if you don't want to fumble around with some other um calls to the store. Um there you can perform a search

like an organic search right when you pick up your phone and say hey game and uh there's uh these search results um you can get these search results here from a fraction of a cent per search and um quite powerful. So with that you can then create your own competition score right and honestly sometimes it's it's it's you don't need a PhD in mathematics to do that. So uh sometimes you can just uh yeah use your gut feeling or your your best guess when you have a search result where every app has millions and millions of downloads then you can say competition is high right and when your app is quite small and maybe has only a handful of downloads then your competition score is quite low right um you can build that yourself and with that you have then volume competition score and relevancy score and that's building blocks of your ASO tool You can do that. Quite cool. Good. Um then we go to the next

one and um the the second um example. So this is um my favorite my go-to um the MA ads tracker. So um yeah due to some some regulations Meta was uh putting some more data within their Facebook ads library API and um you can use that for your advantage right that you know what your competitors are doing there like ads related and you can read them out and um yeah see which ads are working for them. So super um so right now I mean it's it's not a new idea right to to create a spy tool. Um so um I've tried some of them but um for many again not bashing here but some are not that good. Um it's it's just another endless list right often times it's just another list that you need to go through and scroll through and um they're still hiding the good stuff away. So, um, often times it's kind of

overloaded, right? So, they show trillions of ads. So, um, I still I as a marketer still need to go there, scroll through, read it out and see, okay, what's going on there? Literally, what is the best ad, right? So, this is my one question as a marketer, right? So, what's working for them? So, and some are quite expensive as well. And, um, you can do it yourself. So um here the process is a bit longer but again um this is taken as a starting point. So I've written everything down here in the presentation. Um what you need is a Facebook ads library access. So here's the official document. Um you need a Facebook app. So it's uh don't be confused with your app on the store. It can be totally different thing. It's just kind of holding your keys, right? So it's just an entity. Think about it as as that. So it's not your your app on the app store and you need a Facebook

page ID. Um so that's the easiest part um from any advertiser who is doing some ads in uh Europe. So this is what you need and we go through that. So um it's a bit of a bumpy road to get there but for me it's it was quite easy and um uh yeah within a few minutes usually. So my mine was really 20 minutes. Um you have your access um because they want your ID and passport or passport, right? You need to authenticate um yourself there. Um that not everyone is getting access to that. Um so it's kind of hidden and and locked away. It's not the perfect process. I don't like that. But um it's how they do it. We can't change that. We can just follow the rules here. So, uh, when I was uploading my passport, um, in 10 20 minutes it was there. So, there's a a good confirmation process. Um, you upload that there and then they do some, uh, I don't know, background screening. I don't know what they do. And, um, then

you have your access. So, that is this thing. Um, and then you need a Facebook app. Again, don't be confused with the app that you have on your app store. Mhm. It can be different things. um it's just an entity that is holding your access rights in the end to this API and um you can go here uh you might maybe need to click um like a checkbox hey I'm a Facebook dev um I read all the terms condition stuff click on that you're done obviously read it for sure what they write in there um but also that is quite easy and um they have a yeah quite straightforward process for that so when you then click on create an app. Um so this is like the app creation process here. Um you give it a name, your email address and um then you just click on on next in the end and um there you then have already two things like the app ID it's unique you' be bound to your one

app then and an app secret it's it's a password in the end right? So it's a key to this lock. So this is what you have then and then um this might be a bit confusing but again presentation is here. You can reread this every time um that you want. Um then you need to create an access token. So it's just another website also officially hosted on Facebook. It's their tool. Um then you can select your app that you create here. For example, this one here was API test um that I did. And then you can um uh fill out um this URL down here. So you need then um the app ID, the app secret and your access token that you get here. I was cutting that away for good reason. Um but you will see that immediately when you go here. So um and then you can fill that out and uh put that in the URL field here. It's looking then like this here. Again you can fill it out. It really copy paste looks

intimidating I know but it's literally copy pasting um looks like this here you find that on the website I showed before it's a graph RP explorer um then you just fill it out here and you click on submit and that's it and then you have your app ID app secret and the access token which is then uh valid for 60 days um if you don't do this process you still have an access token but it's I think gone in a few hours so Uhhuh. And then you have these three things and with that it's it's totally fine. Um again you can revisit all these slides later on. Um it's uh quite straightforward. Um but yeah the first time um it was for me a lot of googling around how that works um because um the documentation is not perfect on that. So take these slides here maybe as a starting point. Good. Um that's the easiest one. Um you need a Facebook page ID. Um so um you can use facebook.com for that um any competitor. So here we go again in the

fitness space n um and then you go to page transparency and here is the page ID. So and here it's very comfortable that you already see hey this page is currently running ads. Okay cool. Uh great starting point. So they need obviously um to to run ads. Good. Okay let's switch again to the code. Okay, let's open that up and then we go to this one here. So, and here I did exactly that. So, um we have then this access token um and the Facebook page ID from Noom here. And um yeah, this is pretty much all we need here. And again, take this as a starting point because also their documentation is not perfect. Um it's um yeah uh also took me some time to to figure that out. So take this code here as a starting point. Um just really quick um so what it's

doing it's the same idea of saying hey um we are a valid user um this time we are actually so because we registered for this API access and um we can authenticate ourselves with this access token and we say cool um we have this Facebook page ID um and our access token and then we can connect actually to Facebook ads. So um this is what what it's doing here. So this is from a documentation um this is the URL that they uh want uh us to use um to um yeah fetch these ads here and um here you can also filter for any um country that is in the EU and um yeah every parameter so every kind of piece of information that you want from this endpoint. Um so I was going for them here. You can go crazy with that but um we keep it simple here. Um and um yeah this is then everything that you need in the end right. So a

password in the end. So you can think about it as a password and um your Facebook page ID and um some sort of settings what you want the API to return and um then you do that. So and this is all you need here. So this is like the call obviously within your access token. Um and then you're good to go. So, we are scrolling down here. Fingers crossed. Uhhuh. That it works. And we're here. Cool. So, um that we see something. Um I was printing that out here to the screen. So, um we are just uh going through the very first five um ads here. So it's creating a response that is holding all sorts of information from this API and um we are saving this in this variable here and then we iterate over that and say hey for every ad that you found here in the in your result um please give me the the

ID of this um of this ad um the page name the EU total reach. So this is where the interesting stuff happens um because the total reach is attached to every ad and um for example in sides here um the creative body so the the the text that they're using and here we have then um the first ad from n got a reach of 7k uh people and here is the German part um it's kind of random because I was asking for every uh geo but here um it's it's German right now when you're printing out more stuff could be also English or or French or whatever, right? So, and here for example, again, quick example, right? But um here 16,000 um people reached um seems to be way better than this one with 700 people reached. And then you can uh look it up. You can say, "Cool, this is the Facebook page ID. I will exchange this URL pattern here."

And then we can go to the Facebook 8 um library and watch it. And as you see it's copy pasting. It's a manual process right now. Um you can go more advanced again starting point. Um but um this is then that and then you can read out and see cool what's in here right how does look like um all sorts of things right? So do I have reach on that uh UI page at all? Yes. Yes, you have. Um it should be somehow matching. So 16K for that and you should see that here as well. So when you see add details then you see transparency blah blah blah blah blah blah blah 16k. Yeah. Yeah. Then if it has reach then does it order by reach? I'm just trying to pinpoint the advantage of

using the API versus so in the in the UI not. So that's the point why all of these other tools exist. Um so here you have just an endless list, right? So it's the official S library here. Um it's it's not filtered or sorted by anything. Um I mean you can filter but it's kind of stupid. um here not um here in the official response also not but in the code obviously you can still sort of for so um this is also what I do in in my code um so um there I'm reading out the the I don't know last month last half year right and then um it's in the end it's it's two lines of code right and um then you can uh iterate over that and say hey what is the best two ads in last 30 days Right? So this is the big advantage that you can have with the code here. Um but as you see um we are just printing out the first five and this is kind of random right because here it's 7K 700

16k um the response is not sorted but then later on you can sort obviously for that so you can iterate over everything um like wish and um this is the super big advantage um versus using the UI um which is um it's it's cool it's that it's there but in the And um it's it's an endless list, right? So you as a user, you would need to see see at details. Click on EU transparency. Ah, cool. Not it down, right? Click the next one. Not it down. Cool. Right. So it's endless. Too much work, too much Yeah. clicks and scrolls. Yeah. And here we just in in Germany, right? So um I mean for for every um every app out there, it's way bigger. And then uh I mean look at Shiin or TU. I mean, uh, they have hundreds, thousands of ads, right? So, you you can't do that manually. Got you. So, with that then, um, yeah, you have,

uh, text in there, you have the URL in there, you have the links in there. Um, so also here, you can be a bit more advanced then. And, um, yeah, you see literally what's working for them and you see um, the biggest reach. Obviously, you don't have like revenue data attached to that. Um so it's just an assumption that um when the ad is running for a long time has reached many people um that it's hopefully working um but it's a way better estimate than just guessing right so way better for sure. So and um then when we are back here so what can we now do with that? So um then hey we've made it uh we've uh written out two tools and um then we can yeah uh also add a bit more here. Um so uh again starting point right um but uh you can add more access tokens. So um I'm I'm I

think I'm at 10 right now. Um so that you are not running into some sort of rate limits that often. Um the API is quite buggy. So sometimes they are just timing out. Um so you need to uh handle that. But when you have like I don't know uh one two three competitors and you just want a starting point um this code that I shared is totally fine. Um when you want to bit uh have a bit more and be more advanced um then you need a bit of tweak and add more stuff here. Um but um then you can also put that on a server somewhere and say cool um hey do that daily do that on every Monday and uh just give me the best ads from the last week of this competitor works. So quite powerful. Nice. Good. That's it. Yeah. Very nice. Very motivating I should say. I swear. Awesome. That's good. That's good. So if you need

something or have some issues or questions about Python, getting into that, um, ping me. Uh, right. So email or or LinkedIn, doesn't matter. So feel free to reach out to me. Yeah, we'll also leave those links in the description of the video. Awesome. Um, if you have any questions uh about what we've talked about in the video itself and you want other people also to see the answer, please uh leave your questions in the comments. We'll ask when uh to answer those. Um, yeah, it was great uh talking about this and it is I'm so happy that I've did a bit of prep work so I could like ask questions that genuinely interest me about this. Um, yeah, thanks a lot. Thank you. Thank you. Until next one. Yeah. Cheers. See you. Bye-bye.