Use ChatGPT for Backlink Analysis and Visualization

Key Takeaways

hello everyone today we will talk about jgbt and how to use it for search engine optimization the specific example will be about backlink analysis and how we can actually compare two different websites to each other according to their referring domain count or traffic amount of these very referring domains or how many backlinks these domains actually have that's why we will use scmrush and the a chefs together I share this specific to width before and also I shared it in our public holistic SEO communities and many people even if the things are simple many people actually ask me how to do that then I decided to create this blue head series you see a blue head at the left bottom corner it means that this video is very simple quick and if you are following me for over three years this video might be really simple for you but in the holistic SEO communities 10 years plus experienced seos agents or owners and beginners let's say two months of experience seos they all learn together and that's why I am really excited about my come community at the same time you can always expected r as your newsletter you can join to the art communities and just yesterday I started a group coaching and it was really enjoyable and also successful we actually had over 30 people and after announcing that we got really good amount of applications we will be opening the coaching session applications probably for three four days and if you want to join that you can reach out to me or my team or you can maybe use the website to be able to join us as well if you wonder the previous blue hat videos check this specific case study and and again this is a blue hat case uh video too basically we give three suggestions for increasing organic clicks 10 times I did not hide the website name you can directly see it the full case study is here it is being written and it will be published and this is the first jgbt SEO that we did and now we can move on to the second one so in this example here too I have got actually two different exports these exports are from directly the hrefs and our usual I use these type of borders in my prompts to be able to use just a single prompt as I said in the first blue hat chegebt for SEO video two I can tell that you can actually use multiple prompts but I like using only one because when you use multiple prompts it's harder for this specific dialog based interaction machines or text formatting machines it's harder for them to actually take the context from the previous discussion and make it carry to the next one now that you give everything as it should be with high level of confidence the machine actually will be knowing what they should be doing at the beginning and they will be just doing that so in this case here this is the introduction of the prompt and I directly tell that this is a visualization prompt based on a data set to compare two different websites backlinks overall quality and prom has five sections and every section has differentiated from each other with the specific question sorry quotation mark borders if you see these borders it actually tells the chat gbt that this is the first section and you should evaluate this section inside that there's an independent but also dependent section just a phase or a stage of the prompt every prompt section has one definition one purpose and one main instruction then here we tell this is the first one and it is for defining the data set to you or we can directly tell dataset we don't have to say to you here the data set has two different files and both of the files have the same columns but different values the data set is about temp covered and ensure daily named website's backlink data both of the websites are from United Kingdom and from the insurance especially the temporary insurance industry we increase the traffic really well for one of these companies as well so you can check it I might also put it in one of the case studies maybe if you hear my cash share look sorry for that but he might be naughty sometimes both websites have different type of backlinks from different web sources the purple series visualization of both of the website's backlinks by creating bar charts both of the website's backing data should be visible at the same time easily we should be comparing them to each other basically here we tell that we will be actually putting the all the backlink data to the same plot that's why we tell that both of them should be visible in terms of comparing to each other at the same time and we will need to be focusing on the backlinks mainly when I castled in this area we move on to the second Pro second section of the prompt again we tell that this is for helping for visualization and in this part we tell that the data set has domain column basically we start defining our data set in this area which is from this part with not this that it is from sem rush we tell that we have this domain coding and we Define this by telling that these are the names of the domains that are linking one of these specific sections then we Define the DAR I did not use the follow or no follow or traffic type of sections in this area and I guess the HFS has given uh yeah actually it's true and at the same time I use the traffic section too I will explain it later as well and in this part I also tell that to be able to understand which data belongs to which website use the file name if the file name starts like that it is about time covered and at the same time we tell that colorize colorize or colorize both of the domains backlink data in different colors so that we can see it easier Dr means domain rating and it is one of the columns and this section is really important okay create bins for bar charts by starting from one and increase the beams by 10. for example the domains from five six seven eight Dr will be counted together and they will be between 1 and 10 bins if the Dr is above 10 but below 20 it will be between about 10 and 20 bins the important thing here is that you should always give examples of the gbt it will make everything easier for it and the third part here one more time we give the definition then we actually tell that the line chart will use traffic column it is to see which domains link which one of these two websites and what are their traffic value put the traffic value of domains to the line chart and correlate it with the B means that we use for bar chart this section is complicated for jgbt because until now we just say that we will use actual bar chart and there will beans in the bar chart all right now we are talking about the line chart which is an extra part but at the same time if we come to this specific section or in this area we told that we will be comparing the backlink data and at the same time we defined our data set with different type of sections which means that when we upload our data the change gbt will directly be looking for this specifically named columns and since we mentioned a second column or second column name and a second type of chart it means that actually these both of these things will be going to the same chart by using two different columns and this fourth section it is about helping for the visualization because if you don't help for the visualization things will be random you can see stack charts or you can see that actually barge or line charts are in a different chart but chart might be in different place too and it might be a little complicated but if you help it for like this the this part is mainly for actually defining the concepts I would also suggest you to define the concepts while writing a prompt so that you can act this jgbt can directly understand what you are talking about what is the context if as in data science if the data scientists actually know what they are visualized visualizing it means that they will be able to visualize it faster and better year two I just give simple definitions for the concepts that I am using in the top sections and in the fifth section and basically I give the specific last instructions by saying that use this name for the title of the bar plots and write the number of referring domain domains so the y axis label this section is important because I give two different names to the same axis y-axis again the y-axis and this is another one more time this is an annotation because I use a travert traffic I use the word domain here which means that since I use traffic column and the domain column and since I am giving actually multiple labels to the y-axis it means that we are using multiple y-axis and this way the fifth section will be connected to the previous sections of the prompt in an easier way too then we give some other instructions by telling that use different colors for two different websites that to compare them to each other and here too we tell the second y-axis label is for the amount of traffic we make it more even more explicit together with the label name together with the column name and together with the purpose then directly the chair GB starts to do these things the first thing is that always remember ashcraft's exports they are in x actually utf-16 and they use T as a separator I did not state it here the switch gbt had to understand it by itself but it was able to solve the issue in this area then we see that actually this is a temp covered data for backlinks this is the ensure daily data again for the backlinks or referring domains and from this part then it directly gives some information as mentioned in the prompt we will create beams starting from 1 and increase by 10 each time then we will visualize both of the websites backlink data on on the same chart bar chart we will use different colors to distinguish between the two and in this area they start to create these Beams I can also show you the codes in this area it's not that hard actually basically we use numpy in this part and I will suggest you to in the first video too I told you to actually create a a Kaiser folder in your desktop then actually or put all your python codes there in Jupiter notebooks or IPython notebooks in the vs code and use them whenever you need them because as I say chegebi code interpreter is for interpreting the code not for writing the code so you can copy all these codes and give them a new shape and you can even automate these processes as you wish and this part will use PD cut function to be able to create certain type of bins and we include to even the lowest value in the bins as well it is one of the important things that we should be doing too and in this area too since we say that increase the bin numbers by 10 by 10 and in this area we are creating any kinds of uh it is actually from numpy we were calling them array yeah we are creating an array by starting from 0 and it will be going to the actually max value of the Dr and in this area the max volume will be coming to the Dr column since we don't have 100 let's say we have 97 it will be going until that point by starting from 0 by increasing 10 by 10. 10 times and and after that we see actually these specific bins basically in this area for temp cover and also for the ensure daily 2. we see that actually we have 460 referring domains and the domain rating usually are for these are between 0 and 10 when we look at the Domain rating between 90 and 100 it is just two and this should be insured daily when we come this area this number becomes six and this number become like 12 over 12 hundredths and then the 10gbt gives these things to us in a readable way even reading these things like that might be helpful for you to see some statistical data the reason that I am publishing this quick research and let's say quick suggestion videos in the blue hat mindset is that actually not just for showing how to use JG between automating stuff also giving some simple information first of all Dr or domain rating is not that much an important metric but still from time to time you can use it for statistical approaches to analyze a few things in a quick way and at least you can see where the backlink structure of both of these domains are differentiated from each other in this area we can see that especially between 70 and 80. this one has 50 domains and this one actually have just 17 and especially this Zone might be the might be creating a really good amount of difference in terms of indexation or in or prioritization for rankings as well too you can even check the average query count per URL from this and this domain to be able to understand that actually according to the URL count if one of the domains have way much more queries that they are ranking for search engine might be tolerating them even further even if the content doesn't have that much details and finally in this part we see that actually we have a certain type of let's say visualization the important thing here is that this is the visualization of only the bins that we actually have and we are rotating these specific X labels and then it says that here is the bar chart comparing the average domain rating of domains for two website stamp cover ensure daily each pairs of the bars represent a bin of Dr values the blue bar represents stamp and the other one represents the other one and in this part the important thing is that this jgbt still remembers our prompt which is really important to show that actually sectioning or segmenting your prompt by giving clear definitions is important because it did not start as you see since the beginning section by section they continue for every section they give an output according to the my prompt and here also they say that time cover has significantly more backlinks from these lower rated domains compared to the insured daily for the higher Dr beans both websites have similar numbers of the backlinks with temp cover having slightly more in most cases and here we then we talk about traffic column because we will need to create line chart and line chart y-axis will be coming to this area and third of the line that goes over these bar plots and in this part too as you see we actually use a different part again we create the bins and then we match these specific let's say traffic data by summing it first we filter the data with group Pi function of the pandas and with some entire traffic value for that and we reset the index so that we can actually according to the index of these part we can merge them one more time and then in this part you'll see that actually total traffic value of the of these domains that are linking one of the temp cover or the initial daily problem this is the time cover and this is the visualization here why this is important because as much as I can remember temp cover has just six referring domains from the R rating between 90 and 100 and the traffic value here is I guess this is I am not able to read the normal complete about this might be maybe 20 million even and sorry yeah yes and this is the issue daily and this is around I guess the 11 million the purpose here is actually seeing that yeah we have lots of referring domain but how many of them really have the real traffic if you are taking a link just because you have high level that domain has high level Dr it doesn't mean that actually it will be helping to you you should focus on the index session the rankings and the reactions of the websites or the spam updates if they are being affected from content spam update or product review updates or links One updates stay away from these type of domains and at the same time if they don't have traffic why do you get the link from there it won't be showing you that much further there are some exceptions or government websites or scientific educational websites the situation might be different because no one ex expect to have lost millions of clicks to scientific PDF but still it is important and unauthorative so in this way actually Microsoft being defines Authority and Page rank differently too sometimes Google mentions it but they did not disclose that that much but in the patterns we have lots of different things related to the page rank like trust rank or merge rank as well or ranked merge so basically here you will see also like this ensure daily has just 32 thousands of total track value for these amount of domains let's check how many domain is that so basically 400 domains are able to generate just 32 000 clicks which means that most of these domains are garbage when it comes to here the number is way much more let's check the value for here 61. so I built can 61 right yeah so I can tell that this is even worse than this one so most of these domains probably are not even properly indexed or they are just single page or some repetitive sub-domains they can even be when it comes to this area the numbers in this part are way much higher too which means that if you are linked from some of these domains it's possible that more users will be coming to your website with a referring click which means that actually these links are more important because it is where the users are so let's visualize it it becomes like that you can see the same data in this here here too we have a line chart as you see and basically the blue section here represents the uh time covered referring domain count according to the r range between 0 and 10 and at the beginning as you see it actually rotated this but later it did not because it used that specific section of the prompt only for the first visualization but we are able to still understand which one is which one the important thing here that is let me draw it I guess I'm able to take the recording properly while drawing the important thing here is that for every section actually the temp cover has very much more referring domains but the important thing here is that for really important amount of referring domain ranges according to Dr the traffic per domain is higher for the insured daily referral domains because in this part as you see we have two lines one red one blue even if we have a bigger blue bar chart every time the blue line chart is very close to the red one which means that most domains that actually link to the temp cover doesn't have appropriate traffic in it but it changes tremendously when it comes to the domains that have over 80 Dr and this might be the main section that creates a difference when it comes to the prioritization in terms of the linking so it means that this might be the more important area that you should be focusing on in your backlink work or maybe even semantic Network content Network Creations as well too the important thing here is that the chegebit also continues and give us some suggestions in this area since my end the answer was really long it stopped creating the answer then I said continue where your left off and it's continued there and then it gives some explanations here before the definitions it says that both websites have the majority of their backlinks from domains with ADR of less than 10. however Time cover has significantly more backlinks from these lower rated domains compared to insurer daily for the higher Dr beans both websites have similar number of backlinks with temp cover having slightly more in most cases in terms of traffic your domains with a higher Dr contribute more organic traffic for both websites this is especially pronounced for temp cover which receives a significantly higher amount of traffic from domains with TR of 80 or more compared to ensure daily it is interesting not that all the insured data has fewer backlinks from domains with at the art of less than 10 these domains contribute a substantial amount of traffic so the thing here is that in my prompt I did not tell jgbt that these traffic are for the domains of these specific domains I actually say that too but the 3gbt assumes that this is a traffic that actually comes from these domains to these targeted websites so you can make things more explicit too but you don't have to our purpose here was actually a visualization sample so this was from actually hrefs so how about actually a CM rush I use the same prone director for data and it did not work directly as you see here the change gbt is stopped why because in sem rush we have domain Authority score and domains and we don't have the traffic column which means that I will need to Define new Concepts and I will need to help the hgbt to understand the system here and I did that I changed the data set and then I also change the definitions and the sum of the code names and as you see in this area we created a similar result in a really quick way the important thing here is that when you use actually both of the tools you will realize one more thing because in the sem Rush export we have backlink amount a referring domain so here we have traffic of domains that link one of these and here we have the backlinks of the domains that link one of these websites which means that actually even I can merge all these data I can even create a more let's say a comprehensive visualization in this part and if you look at in this part 2 we will directly see that actually temp cover which is the backlink count in this area it actually says usual the backlink but the bar plot actually shows the amount of domains that are linking Time cover or insured daily and in this case ensure daily is the red one and again we see that actual temp cover has more referring domains the X labels here are according to the authorities course of domains according to semrush data and when it comes to the backlink count it is for the domains or the referring domains in this part the thing here is that for one of the websites here the authority score is between 30 and 40 and here one of the websites Lynch maybe 400 000 time to ensure daily we don't know which website is that we can check it because the back total backlink count increased in a really really uh tremendous way and in this case here you see the numbers are closer to each other and in this part too the church GPT gives some simple definitions to explain these things directly too maybe next time we can also show things over the sem rush or HFS 2 to be able to be helpful and give some quick suggestions and basically we will we will continue to give this type of quick researches and the quick tips for search engine optimization by using artificial intelligence we will even create videos for large language models natural language processing and we will talk about the future of the search engine optimization still I am trying to complete this specific case study for a really long time over one year and I will be publishing it as well too I will suggest you the content to follow us and join our communities I'm happy to see you all and I hope you are safe and happy love you all and see you later [Music]