PR on Habrahabr. Statistics, expectations, reality and surprises
On June 23, I wrote in “I am PR” an article about the inadvertent brainchild of a small group of programmers — another alternative quote book . A month has passed and it is already possible to sum up some results, which, I think, will be quite useful to those who want to promote themselves through Habr and are not without interest to those who are going to do some sort of promotion.
The main flow of visitors went, of course, on the day of the post - the 23rd. The next day, there were already a lot less new visitors, and after 4 days the flow of new visitors from Habr almost disappeared. In total, about 3,000 visitors came from Habr, of which about 150 remained to periodically visit the resource, that is, only 5%. Of these, 46 people registered on the site, with 8 people via VKontakte, 10 via Google, 2 via facebook, 2 via twitter, 2 via yandex. The remaining 22 registered in the usual way. That is, more than 50% of registered users made it through OAuth. That once again underlines the importance of this mechanism. Yes, through classmates and mailra - not one went.
Here was the first surprise - I expected a greater response, and a greater percentage of the growth of regular users. But if the response is what it is, there is one, then the second number can be analyzed a little.
Before the start of the promotion, I installed an openstat counter on all pages of the site. Why openstat? For nothing. Yes, I am aware of the opinion that this is not the most convenient tool for web analytics, and I am not going to argue with that. I just had some experience with him, and I decided to take advantage of what I was familiar with.
As you can see, the failure rate was very high. And the percentage of failures on the first day of the action was much higher than the average . In fact, 3/4 of new visitors closed the site, casting one look at it. Why?
I see only one answer - design. The article with which the transition to the site took place, honestly and in detail explained to the user where he was going. The user consciously and voluntarily went to the site, but immediately closed it. To see the functionality of the site in 10 seconds is unrealistic, which means that the user simply “did not like the site.” And this is about the design. Probably, many will simply grin at what they consider to be commonplace truths, but for me - a hardcore coder - it really was a revelation. The main task of the design is to keep the user in those first seconds when he knows nothing about the site. Convenience and functionality of the design, quality of the content and everything else is later.
Further - the most interesting reports. Since I investigated the relationship of various parameters with the number of failures, this indicator appears in all reports.
Here for me two surprises appeared at once:
Firstly, a broad 1920x1080 confidently broke into the leading position. For me, with my 1280x1024, it was a surprise. (Let me remind you, I am a coder, and not a designer near).
Secondly, the holders of this permit were also in the first place in terms of the bounce rate. I rushed to watch the site, confident that the design was leaving at that resolution — nothing of the kind. Yes, the letters on the 17 "monitor become a bit small, but what's the point of putting such a resolution on a regular monitor? Rather, there is a confirmation of the already put forward thesis about the design. The owners of large widescreen monitors do not like the site at first glance. It does not match. From the fixed sizes of fonts I put in the plan.
')
Chrome in the first place did not surprise me. But the undoubted dislike for the Opera users' site was rather surprising, and even a little offended by its groundlessness. Having spent a lot of time on the cross-browser design and, most often, remembering the unkind word of the opera, to get such black ingratitude from its users ... for what? No, everything is in order with the design of the site under the opera - it was checked and checked constantly. I do not understand.
Surprised by the presence among the visitors already five users of IE6. Fu on them! I am afraid to even imagine how our site looks like under the sixth donkey - they did not put a stub, believing that you wouldn’t find living IE6 users.
The bounce rate on smartphones is generally above average, which is not surprising. But especially the site did not like the owners of the Samsung. Who has Samsung - tell me, please, is this a valid claim or something as subtle and elusive as in the previous paragraph?
I expected and foresaw the spread between failures on Windows and other systems, but I thought it would be stronger - in my opinion, Linux users should like the site noticeably less than Windows users, but no.
And finally, two capitals became champions in the percentage of failures (from which the main stream came). I think this is also about design. The pace of life in the capitals is much higher, and few people decide to spend precious time studying a site that didn’t like at first sight. By the way, in vain. Of those ~ 700 people who did not close the site in the first 10 seconds, ~ 150 remained regular users, i.e. 21%, and this is a very high percentage.
On this with all analytics. The remaining reports did not seem to me deserve special attention, or - strongly tied to the structure of the site and uninteresting to anyone except the developers.
Strongly recommend this item to read to anyone who is going to be promoted through Habr. This term arose in the process of discussing the PR topic more than once, but every time it turned out to be a false alarm - the server successfully kept the increased load, and not to say that it was so big. The surprise turned out to be different - it was worth only to look in the fail2ban logs. Suspicious requests, attempts to find a password for all services, tapping all ports - hail fell simultaneously with the growth of normal traffic. Periodic messages about the inaccessibility of the site became clear - a firewall every hour a ban of 3-5 subnets. On the 46 registered "honest" users, more than a dozen accounts were identified (and deleted with a ban on IP) that were obviously created for something bad — with suspicious iframes in the real name and other fields of the profile (html tags are cut from me only before outputting ), with explicit sql-injections and other bjakami. The site allows downloading custom css files. The users used this feature 3 times, two of them, instead of styles, the code for the deface was loaded (although the attackers failed to execute it).
Well, and so on.
That's not too lazy for people :)
On the one hand - thanks, of course, for extreme testing.
And on the other hand, it’s necessary to warn :) Okay, we are not writing the code for the web the first day. And what kind of young-naive PR will be promoting, and his commercials will be muzzle to the ground. Not good :)
If you want to promote yourself here, I recommend in the text of the article tearfully asking for errors to be sent to the PM. And that will be like mine. Thank you very much alexglue for the response, but after he published a small list of vulnerabilities right in the comments - people rushed to check the list with enthusiasm, the right is worthy of a better application :) Okay, there was nothing critical.
In general, before you promote yourself here, pay special attention to security.
I did not count on a large influx of visitors. But there was a rather big hope that among the thousands of specialists who read the post there would be a couple of crazy people who would like to join our team. At least part time. Alas. Nobody even hinted: (And the designer would be very useful to us - Sergey Burnakov, the author of both the old and the new designs, (having given out the design promised half a year ago) announced that he was putting off the responsibilities of the designer. Alas, we are miserable. We'll have to learn the wisdom of Photoshop and Illustrator.
Number one - letters from spammers offering to psite the site. And one offered to send 2 million addresses absolutely for nothing. The rest of the letters were shallowed with anger, but we are still hesitating about it - we hate spam, but ... freehand :)
Number two is ambiguous. What finally made me bring all the above (and below) to the public. For, even though everything looks fine, I feel some kind of catch.
Initially, there was a letter with the proposal to "make money on the site." I suggested that we are talking about the placement of viral and other ### koy advertising (yes never!), But still entered into a correspondence. As a result of which it turned out, in brief, the following:
A certain Soft-Integrator LLC with a legal address in Novosibirsk is ready to pay money for unique humorous content. 300 rubles for a short form (up to 2 thousand characters), 500 for a story and 1000 for a story with a unique photographic material or a video clip. Content must be selected from at least X (for each type its own number) of applicants by open voting in the internet for at least one week, must receive at least Y unique votes and the coefficient "for" / "against" should not be less than Z. In confirmation, I must provide the lists of those who voted IP - they, de, will make control votes and check the presence of their IP in the list. Content authors must be non-anonymous; before entering the competition, they must tick the box under the agreement. I get paid money by transfer, but I’m free to do what I like with it - STE LLC recommends me, after paying taxes and cutting off some% for myself, to pay them to content authors, which actually attracts them. This is exactly what I think I’d do, and I don’t even take a percentage to myself - that there is a trifle, but the people can certainly attract such an action. I have already signed the contract (since he does not oblige me to anything) and now we are working on the code, but there are some doubts.
There are a lot of misunderstandings. Of course, I was in Skype communication with their manager, Svetlana, trying to clarify them, but did not work out to the end.
Q (I): Why do you need this content?
A (Svetlana): We are going to soon bend under us the entire comic sector of the runet. And so we need (at least for the first time) quotes, jokes, stories, pictures, etc.
Q: Why is an unknown site with a dozen visitors, and not the same bash - it is unlikely even they would refuse such a freebie.
A: Firstly, the bash (and the other pillars of humor in runet) is our future competitor. We are afraid of badly selecting content once, we don’t want to promote a competitor - two and, most importantly, three - it is desirable for us that this content - when it goes to the people - is fresh and unknown to the people.
Q: I have to mention on the site the name of the action “Make a joke and be richer”, mention its external sponsorship, but I am not allowed to mention the sponsor himself. Why?
A: We have not yet decided on the trademark, the main domain, etc.
Q: I am forbidden by the terms of the contract to transfer the domain on which the site that collects content is located to another person during the term of the contract. Why?
A: Because. Do not like it - do not sign.
We four times read the contract a hundred times, trying to find a catch. Not found. In general, there is only one point on it, which may require something from me - if later it turns out that the content is not unique and not authorial, then I must return the money received for it in twofold size. That is, before sending the winning content to them, I have to make sure that it is the author. With pictures and videos, here I foresee problems, so, most likely, there will be no competition for them on our site. And with the texts, what's the problem - 10 minutes in Google and everything is clear - the author's text or copy-paste. And the requirement is quite understandable. So it’s as if everything is in chocolate, the work is in full swing and, approximately in time, we will start paying the authors of the winning quotes and anecdotes fees:) ... but doubts are gnawing. Maybe someone has already come across something similar?
ps Yes, that Svetlana asked me if I have acquaintances who hold small sites of humorous subjects, to which one can fasten such a vote, give her their contacts. If there are such here, I can give your contact. At your own risk :)
Expectations and reality
The main flow of visitors went, of course, on the day of the post - the 23rd. The next day, there were already a lot less new visitors, and after 4 days the flow of new visitors from Habr almost disappeared. In total, about 3,000 visitors came from Habr, of which about 150 remained to periodically visit the resource, that is, only 5%. Of these, 46 people registered on the site, with 8 people via VKontakte, 10 via Google, 2 via facebook, 2 via twitter, 2 via yandex. The remaining 22 registered in the usual way. That is, more than 50% of registered users made it through OAuth. That once again underlines the importance of this mechanism. Yes, through classmates and mailra - not one went.
Here was the first surprise - I expected a greater response, and a greater percentage of the growth of regular users. But if the response is what it is, there is one, then the second number can be analyzed a little.
Before the start of the promotion, I installed an openstat counter on all pages of the site. Why openstat? For nothing. Yes, I am aware of the opinion that this is not the most convenient tool for web analytics, and I am not going to argue with that. I just had some experience with him, and I decided to take advantage of what I was familiar with.
Analysis
As you can see, the failure rate was very high. And the percentage of failures on the first day of the action was much higher than the average . In fact, 3/4 of new visitors closed the site, casting one look at it. Why?
I see only one answer - design. The article with which the transition to the site took place, honestly and in detail explained to the user where he was going. The user consciously and voluntarily went to the site, but immediately closed it. To see the functionality of the site in 10 seconds is unrealistic, which means that the user simply “did not like the site.” And this is about the design. Probably, many will simply grin at what they consider to be commonplace truths, but for me - a hardcore coder - it really was a revelation. The main task of the design is to keep the user in those first seconds when he knows nothing about the site. Convenience and functionality of the design, quality of the content and everything else is later.
Further - the most interesting reports. Since I investigated the relationship of various parameters with the number of failures, this indicator appears in all reports.
Screen resolution.
Here for me two surprises appeared at once:
Firstly, a broad 1920x1080 confidently broke into the leading position. For me, with my 1280x1024, it was a surprise. (Let me remind you, I am a coder, and not a designer near).
Secondly, the holders of this permit were also in the first place in terms of the bounce rate. I rushed to watch the site, confident that the design was leaving at that resolution — nothing of the kind. Yes, the letters on the 17 "monitor become a bit small, but what's the point of putting such a resolution on a regular monitor? Rather, there is a confirmation of the already put forward thesis about the design. The owners of large widescreen monitors do not like the site at first glance. It does not match. From the fixed sizes of fonts I put in the plan.
')
Browsers
Chrome in the first place did not surprise me. But the undoubted dislike for the Opera users' site was rather surprising, and even a little offended by its groundlessness. Having spent a lot of time on the cross-browser design and, most often, remembering the unkind word of the opera, to get such black ingratitude from its users ... for what? No, everything is in order with the design of the site under the opera - it was checked and checked constantly. I do not understand.
Surprised by the presence among the visitors already five users of IE6. Fu on them! I am afraid to even imagine how our site looks like under the sixth donkey - they did not put a stub, believing that you wouldn’t find living IE6 users.
Smartphones.
The bounce rate on smartphones is generally above average, which is not surprising. But especially the site did not like the owners of the Samsung. Who has Samsung - tell me, please, is this a valid claim or something as subtle and elusive as in the previous paragraph?
Operating Systems.
I expected and foresaw the spread between failures on Windows and other systems, but I thought it would be stronger - in my opinion, Linux users should like the site noticeably less than Windows users, but no.
Geography.
And finally, two capitals became champions in the percentage of failures (from which the main stream came). I think this is also about design. The pace of life in the capitals is much higher, and few people decide to spend precious time studying a site that didn’t like at first sight. By the way, in vain. Of those ~ 700 people who did not close the site in the first 10 seconds, ~ 150 remained regular users, i.e. 21%, and this is a very high percentage.
On this with all analytics. The remaining reports did not seem to me deserve special attention, or - strongly tied to the structure of the site and uninteresting to anyone except the developers.
Habraeffekt
Strongly recommend this item to read to anyone who is going to be promoted through Habr. This term arose in the process of discussing the PR topic more than once, but every time it turned out to be a false alarm - the server successfully kept the increased load, and not to say that it was so big. The surprise turned out to be different - it was worth only to look in the fail2ban logs. Suspicious requests, attempts to find a password for all services, tapping all ports - hail fell simultaneously with the growth of normal traffic. Periodic messages about the inaccessibility of the site became clear - a firewall every hour a ban of 3-5 subnets. On the 46 registered "honest" users, more than a dozen accounts were identified (and deleted with a ban on IP) that were obviously created for something bad — with suspicious iframes in the real name and other fields of the profile (html tags are cut from me only before outputting ), with explicit sql-injections and other bjakami. The site allows downloading custom css files. The users used this feature 3 times, two of them, instead of styles, the code for the deface was loaded (although the attackers failed to execute it).
Well, and so on.
That's not too lazy for people :)
On the one hand - thanks, of course, for extreme testing.
And on the other hand, it’s necessary to warn :) Okay, we are not writing the code for the web the first day. And what kind of young-naive PR will be promoting, and his commercials will be muzzle to the ground. Not good :)
If you want to promote yourself here, I recommend in the text of the article tearfully asking for errors to be sent to the PM. And that will be like mine. Thank you very much alexglue for the response, but after he published a small list of vulnerabilities right in the comments - people rushed to check the list with enthusiasm, the right is worthy of a better application :) Okay, there was nothing critical.
In general, before you promote yourself here, pay special attention to security.
Cheated expectations
I did not count on a large influx of visitors. But there was a rather big hope that among the thousands of specialists who read the post there would be a couple of crazy people who would like to join our team. At least part time. Alas. Nobody even hinted: (And the designer would be very useful to us - Sergey Burnakov, the author of both the old and the new designs, (having given out the design promised half a year ago) announced that he was putting off the responsibilities of the designer. Alas, we are miserable. We'll have to learn the wisdom of Photoshop and Illustrator.
Surprises
Number one - letters from spammers offering to psite the site. And one offered to send 2 million addresses absolutely for nothing. The rest of the letters were shallowed with anger, but we are still hesitating about it - we hate spam, but ... freehand :)
Number two is ambiguous. What finally made me bring all the above (and below) to the public. For, even though everything looks fine, I feel some kind of catch.
Initially, there was a letter with the proposal to "make money on the site." I suggested that we are talking about the placement of viral and other ### koy advertising (yes never!), But still entered into a correspondence. As a result of which it turned out, in brief, the following:
A certain Soft-Integrator LLC with a legal address in Novosibirsk is ready to pay money for unique humorous content. 300 rubles for a short form (up to 2 thousand characters), 500 for a story and 1000 for a story with a unique photographic material or a video clip. Content must be selected from at least X (for each type its own number) of applicants by open voting in the internet for at least one week, must receive at least Y unique votes and the coefficient "for" / "against" should not be less than Z. In confirmation, I must provide the lists of those who voted IP - they, de, will make control votes and check the presence of their IP in the list. Content authors must be non-anonymous; before entering the competition, they must tick the box under the agreement. I get paid money by transfer, but I’m free to do what I like with it - STE LLC recommends me, after paying taxes and cutting off some% for myself, to pay them to content authors, which actually attracts them. This is exactly what I think I’d do, and I don’t even take a percentage to myself - that there is a trifle, but the people can certainly attract such an action. I have already signed the contract (since he does not oblige me to anything) and now we are working on the code, but there are some doubts.
There are a lot of misunderstandings. Of course, I was in Skype communication with their manager, Svetlana, trying to clarify them, but did not work out to the end.
Q (I): Why do you need this content?
A (Svetlana): We are going to soon bend under us the entire comic sector of the runet. And so we need (at least for the first time) quotes, jokes, stories, pictures, etc.
Q: Why is an unknown site with a dozen visitors, and not the same bash - it is unlikely even they would refuse such a freebie.
A: Firstly, the bash (and the other pillars of humor in runet) is our future competitor. We are afraid of badly selecting content once, we don’t want to promote a competitor - two and, most importantly, three - it is desirable for us that this content - when it goes to the people - is fresh and unknown to the people.
Q: I have to mention on the site the name of the action “Make a joke and be richer”, mention its external sponsorship, but I am not allowed to mention the sponsor himself. Why?
A: We have not yet decided on the trademark, the main domain, etc.
Q: I am forbidden by the terms of the contract to transfer the domain on which the site that collects content is located to another person during the term of the contract. Why?
A: Because. Do not like it - do not sign.
We four times read the contract a hundred times, trying to find a catch. Not found. In general, there is only one point on it, which may require something from me - if later it turns out that the content is not unique and not authorial, then I must return the money received for it in twofold size. That is, before sending the winning content to them, I have to make sure that it is the author. With pictures and videos, here I foresee problems, so, most likely, there will be no competition for them on our site. And with the texts, what's the problem - 10 minutes in Google and everything is clear - the author's text or copy-paste. And the requirement is quite understandable. So it’s as if everything is in chocolate, the work is in full swing and, approximately in time, we will start paying the authors of the winning quotes and anecdotes fees:) ... but doubts are gnawing. Maybe someone has already come across something similar?
ps Yes, that Svetlana asked me if I have acquaintances who hold small sites of humorous subjects, to which one can fasten such a vote, give her their contacts. If there are such here, I can give your contact. At your own risk :)
Source: https://habr.com/ru/post/146976/
All Articles