There will be silence after the stats speak in this thread

[SwirlsGirl]

Greetings friends....

Instead of me going on a rant...I will let my stats rant for you...learn to speak the language of stats, and you learn a lot.

cumulative JOIN FORM SUBMISSION % for the month of APRIL for the last 8 years...

April 2005 form submission 26% approvals 79%

april 2006 form submissions 38% approvals 79%

april 2007 form submissions 43% approvals 88%

april 2008 form submissions 39% approvals 77%

april 2009 form submissions 27% approvals 74%

april 2010 form submissions 20% approvals 90%

april 2011 form submissions 37% approvals 52%

april 2012 form submissions 8% approvals 74%

I think its safe to say we know a little bit about our stats and form submission percentages and averages...

I think its also safe to conclude that what Beaner has said about our site being the best converting adult site he has ever seen has some merit.

We currently have the best tour, best content, best affiliates, best traffic, of all time and you want me to believe that our form submissions are at an all time low...just because

There is something causing the disconnect, and since our form hits are still there we cannot figure out why such a disconnect with form submissions.

March 2012 = record high month for SaraSwirls.com

April 2012 = Record low month for SaraSwirls.com

1 person please come in here and tell me they also had a record high month in march, then turn around and have a record low in april...

anyone,someone, bueller?

[arock10]

you have more random people entering fake info to your signup form?

[iSpyCams]

I also had a record high in March and april is lower, I think its pretty simple.

Payment processors were feeling good in march and had the scrub and fraud checking set fairly low. The increased sales volume got them feeling good enough about cash flow to run up security and this limits fraud joins but also produces quite a few false positives.

One they (not us) feel the squeeze from the reduced income they will turn it back down and we will all be trying to figure out what we did right again.

BTW Sara I don't know what you are referring to when you talk about form submit percentages. I get the approvals, I just don't know what you mean about join form percentages. Percentage of all visitors that visit your site that end up filling out the form? Percentage of uniques that convert to form submits, what?

[Sly]

Set up your own presubmission form so you can see what information people are submitting. Easy info to gather with a setup like NATS. Seeing names and e-mail addresses will help you determine how many of your submissions are actually legitimate versus garbage.

Your numbers certainly look "off", but without knowing exactly what people are typing in this month compared to last month, you don't have a full picture.

Furthermore, I'm not sure what the processor has to do with people filling out your form. Approvals, sure, filling out the form? Where is the connection? Assuming it's the same exact form.

[SwirlsGirl]

Originally Posted by pompousjohn

I also had a record high in March and april is lower, I think its pretty simple.

Payment processors were feeling good in march and had the scrub and fraud checking set fairly low. The increased sales volume got them feeling good enough about cash flow to run up security and this limits fraud joins but also produces quite a few false positives.

One they (not us) feel the squeeze from the reduced income they will turn it back down and we will all be trying to figure out what we did right again.

BTW Sara I don't know what you are referring to when you talk about form submit percentages. I get the approvals, I just don't know what you mean about join form percentages. Percentage of all visitors that visit your site that end up filling out the form? Percentage of uniques that convert to form submits, what?

I actually just demonstrated that by posting my join form submission stats for the month of april for the last 8 years.

Notice how the form submissions percentages were always double digit.... well thats with different amounts of traffic, different content, different tours, different affiliates, over the years

the point here is consistency in the numbers...when their are *HUGE* unexplained spikes and drops beyond a normal spike or drop then one seeks answers that make sense not answers that blow smoke

with all the changes over the years the form submission percentages were always respectable double digits!

Now to see such a drastic reduction or what should better be called falling off a cliff....that to me is beyond making no sense...

to me and to anyone who studies stats that would be a huge red flag and would warrant some serious digging to find out why the disconnect.

Especially if the goal of all parties is to make money and grow our business.

In the real world you do not have record high home and auto sales in march only to have record low home and auto sales in april

If there was no disconnect somewhere we might see a drop in form submissions for april but not a record drop in april after a record high in march....that is unacceptable

[SIK]

% are kinda useless without the actual numbers

[signupdamnit]

Is the first percentage Join hits to Join Submits and the 2nd percentage Join Submits to Joins?
If so it might be helpful to share the number of join hits for each period so that it can be tied into the data. If you see a swing from 37% join submits to 8% then that would be one thing if the join hits were nearly the same but it might mean another if the join hits during the period where you see 8% were 5 times what they were in the period where you saw 37%.

I see oddities as an affiliate even so I'm definitely interested, but I think it would help to provide some more data if you can.

[SwirlsGirl]

Originally Posted by SIK

% are kinda useless without the actual numbers

I would consider posting more data but it is of no use when so few understand and even actually pay attention to form submissions percentages and underestimate the value of the stat.

What I have demonstrated was that over a long spell of 8 years and a wide variety of tours, affiliates,traffic, content....the form submission percentages were pretty steady and stable for the month of april until this april 2012.

I can understand a drop off from march, but we are talking a drop off of such huge magnitude it totally defies reason and probability.

The form submissions fluctuate way more than any other stat and I assert that if you see wild form submit fluctuations then you should see wild fluctuations with traffic, approvals, and every other stat.

I am not alone with my sentiments as my account has been looked at by professionals and they also conclude that the numbers appear to show a disconnect at times.

Just wondering if anyone else shows a record drop off in form submissions for april while all other stats are normal?

[Dirty F]

God, you are a freak. Just stop fucking posting and go see a doctor. You fucking psycho!

[SwirlsGirl]

Originally Posted by signupdamnit

Is the first percentage Join hits to Join Submits and the 2nd percentage Join Submits to Joins?
If so it might be helpful to share the number of join hits for each period so that it can be tied into the data. If you see a swing from 37% join submits to 8% then that would be one thing if the join hits were nearly the same but it might mean another if the join hits during the period where you see 8% were 5 times what they were in the period where you saw 37%.

I see oddities as an affiliate even so I'm definitely interested, but I think it would help to provide some more data if you can.

Thanks signup for the input as I really respect your analysis and take on a variety of things.

Here is the missing data and I would really welcome your input after you check it out...

april 2012- 589 form hits 50 form submissions 8.49% form submit percentage 37 approvals 74% appr%

april 2011 - 275 form hits 101 form submissions 36.73% form submit percentage 53 approvals 52% appr%

april 2010 - 462 form hits 89 form submissions 20% form submit percentage 80 approvals 89% appr%

april 2009 - 899 form hits 240 form submissions 27% form submit percentage 178 approvals 74% appr%

april 2008 - 184 form hits 71 form submissions 39% form submit percentage 55 approvals 77% appr%

april 2007 - 58 form hits 25 form submissions 43% form submit percentage 22 approvals 88% appr%

april 2006 - 532 form hits 203 form submissions 38% form submit percentage 161 approvals 79% appr%

april 2005 - 1040 form hits 267 form submissions 26% form submit percentage 212 approvals 79% appr%

Now look closely and you will see form hits all over the place but form submission average percentage for 2005-2011 was a solid and respectable 32%

32% is the average form submission percentage! This is a crucial number to understand...as it shows regardless of traffic, affiliates, content....the average over 7 years for the month of april was 32%

So the monthly average form submission % dropped off a cliff for the month of arpil 2012 right after a record march 2012

again 32% average for april dropping down to 8% is not a *normal* drop based on the historical data...anyone coming in and telling you it is...well they could very well have other motives

let me also add that subscription price was posted clearly before the join form so you do not have to click to the form to get to the price... Something caused surfers to like the tour and price enough to go to the join form, however once on the join form something made the surfers change their mind for the month of april.

Maybe sloooow loading join forms for the month of april...I don't know but something does not add up

[CaptainHowdy]

[BV]

After checking out archives of your site over the years, what stands out to me is now you disclose the price before having to load the ccbill join form. Where as in the past you had to load the form in order to find out the price.

[raymor]

Originally Posted by BV

After checking out archives of your site over the years, what stands out to me is now you disclose the price before having to load the ccbill join form. Where as in the past you had to load the form in order to find out the price.

Very interesting. I wonder what would happen if it was put back like it was.

Sara, thanks for the additional numbers. Percentages alone left one huge question unanswered. If our Clonebox order form got ten page loads in April and two sales, that's 20%. Six page loads in March and two sales would be 33%. If you just look at 20% vs. 33% it looks like a big difference, but thre's actually no significant difference at all - just four more hits from spiders or whatever.

[Dirty D]

Originally Posted by SwirlsGirl

I think its also safe to conclude that what Beaner has said about our site being the best converting adult site he has ever seen has some merit.

We currently have the best tour, best content, best affiliates, best traffic, of all time and you want me to believe that our form submissions are at an all time low...just because

Looks like you only had 37 sales so far this month.

This would not be the best converting adult site by a long way.

[Dirty F]

Originally Posted by Dirty D

Looks like you only had 37 sales so far this month.

This would not be the best converting adult site by a long way.

Ofcourse not. That psycho lives in some kind of fucked up fantasyworld. Completely fucked in her head.

[epitome]

Threads like this always interest me because our trends demonstrate the opposite.

CCBill is secondary in the cascade and nothing looks astray. Ratios are great, transactions are going through and rebills are taking place.

With close to 60 sites we probably have a much larger data sampling, but everything is fine.

[arock10]

Originally Posted by SwirlsGirl

Thanks signup for the input as I really respect your analysis and take on a variety of things.

Here is the missing data and I would really welcome your input after you check it out...

april 2012- 589 form hits 50 form submissions 8.49% form submit percentage 37 approvals 74% appr%

april 2011 - 275 form hits 101 form submissions 36.73% form submit percentage 53 approvals 52% appr%

april 2010 - 462 form hits 89 form submissions 20% form submit percentage 80 approvals 89% appr%

april 2009 - 899 form hits 240 form submissions 27% form submit percentage 178 approvals 74% appr%

april 2008 - 184 form hits 71 form submissions 39% form submit percentage 55 approvals 77% appr%

april 2007 - 58 form hits 25 form submissions 43% form submit percentage 22 approvals 88% appr%

april 2006 - 532 form hits 203 form submissions 38% form submit percentage 161 approvals 79% appr%

april 2005 - 1040 form hits 267 form submissions 26% form submit percentage 212 approvals 79% appr%

Now look closely and you will see form hits all over the place but form submission average percentage for 2005-2011 was a solid and respectable 32%

32% is the average form submission percentage! This is a crucial number to understand...as it shows regardless of traffic, affiliates, content....the average over 7 years for the month of april was 32%

So the monthly average form submission % dropped off a cliff for the month of arpil 2012 right after a record march 2012

again 32% average for april dropping down to 8% is not a *normal* drop based on the historical data...anyone coming in and telling you it is...well they could very well have other motives

let me also add that subscription price was posted clearly before the join form so you do not have to click to the form to get to the price... Something caused surfers to like the tour and price enough to go to the join form, however once on the join form something made the surfers change their mind for the month of april.

Maybe sloooow loading join forms for the month of april...I don't know but something does not add up

your volume is far too low to draw any conclusions.

[Dirty F]

Originally Posted by arock10

your volume is far too low to draw any conclusions.

She doesn't give a shit. She is actively looking for conspiracies. She is a truther. Behind everything is a possible conspiracy. She thinks people are after her. She opens a new thread about her stats almost daily. Completely obsessed. Complete nutjob. Like all other truthers. Mentally ill.

[Dirty F]

Originally Posted by epitome

Threads like this always interest me because our trends demonstrate the opposite.

CCBill is secondary in the cascade and nothing looks astray. Ratios are great, transactions are going through and rebills are taking place.

With close to 60 sites we probably have a much larger data sampling, but everything is fine.

She ignores messages like this. She thinks people are targetting her specifically. That's how fucked up she is.

[WarChild]

It shouldn't give me so much pleasure knowing you make less sales in a month than I do in a good day but still it does.

I wonder how come nobody successful is ever a truther.

[EddyTheDog]

"86% of stats you see on the internet are made up on the spot by people who have no idea what they are talking about".....

[Jel]

not much silence in this thread, I call false advertising

[SwirlsGirl]

Originally Posted by Dirty F

She ignores messages like this. She thinks people are targetting her specifically. That's how fucked up she is.

Hey Dirty F and Warchild can you stay the fuck out of my threads? ALl you do is come into the thread and say she is a crazy bitch, she is mental, she is retarded...


You never address the anomalies in stats...you just attack attack and attack...its clear I am not crazy and its clear I make some serious points...

why do you search every thread of mine to say I am crazy...

32% average form submissions for every april in 7 years but this april form submissions are down to 8%

Come on gfy you can do better give me something tangible come on...show me someone else who has a 25-30% point drop in form submits...

no you cannot all you can do is say she is crazy....don't you know that evey genius has been called that at some point...

You call me crazy 1000 times...let me clue you in on a secret....your efforts are valiant...but you yelling crazy crazy crazy 1000 times does not explain the disconnect in form submits...

8% form submits for april 2012 when the average is 32%...well that tells me that my revenue for this month could potentially be 4 times what it was....LOL

But I get it...I have affiliates working their asses off...I have a partners working their asses off...I am working my ass off... and i am supposed to sit here and shut the fuck up about an obscene drop in form submit percentages...24% drop

The most I can get from gfy is a bunch of ridiculous grade school insults...

Can you tell me any other industry where you have a record high march for the year followed by a record low april....

All the work, all the sweat, all the tears put into this operation...and I should accept an all time record low of 8% form submits for the month!??!?!?

When I know my stats better than anyone...and I know that most members of my team have given all they can give....and everyone is doing their part to get sales...

I do not enjoy being called crazy, and looney but if that is what it takes to get people thinking, and to get to the truth so be it...I am crazy like a shit house rat.

the only thing crazier is this anomaly....

to have 8% form submissions for the month of april and 856 from hits....means all those visitors liked the tour enough to click to the join page....liked the price enough on the join page to click to the join form.....

but then they get to the join form and have a *sudden* change of heart....LOL comical...

Thats not how it works and you guys should know better....what do you think motivates me...You think I could do this if I was not 100% sure of the things I speak?

856 form hits after seeing the price for membership... with only 8% form submits that is one huge fucking disconnect and as far as I am concerned there has to be a better way of monetizing those clicks!

those are valuable clicks going to the join form after seeing the price....LOL there must be a way to monetize that traffic even if they do not get a membership....

[SwirlsGirl]

crazy bitch is mental...

ugly ass stupid bimbo...

retarded dumb fucking cunt

stupid ass wipe woman

mentally retarded abused cunt...

okay I helped you guys with a few insults....now that we have enough insults can we please address the *reasons* for the disconnect?

Sales do not explode during march and implode during april all by themselves

I am not unreasonable where I expect the same amount of sales month after month...I know better I understand fluctuations quite well...

We are talking a drop off a fucking cliff....

If you want to make me shut up about the threads then how about someone somewhere be honest with the adult community and acknowledge that there are vertical limits on sales regardless of what any merchant does

Just admit there are limits so I can say I told you so....the numbers show there are limits, and people tell you there are no such thing as a limit or ceiling....

I have shown you how I have doubled my form hits from april 2011 and the form submits for april 12 just evaporated into nothing.

What was the point of increasing traffic, and form hits?

What was the point of gaining new affiliates?

What was the point of putting in a ton of work to have form submits drop to a record low 8% for the month of april?

I showed you that every april we have been in business we have had respectable double digit form submissions....

April 2012 is an anomaly!

[Dirty F]

Holy shit! This is getting nuts. You are totally obsessed! Scary. I wonder don't you have friends or family who can take care of you? I know if i had friends with a severe mental sickness i would force him to get help.

[Dirty F]

And on topic: what fuck do you want exactly you fucking psycho? You won't accept any other explanation other than there is some kind of conspiracy going on. That's the only thing you want to hear and will accept. Because you are a truther and live in some kind of fucked up fantasy world. And every day you will open the same fucking thread till someone will say yes it's a conspiracy and they are after you. You fucking sick idiot.
Amazing how beaner said there is nothing mentally wrong with you. Even an 8 year old could spot you are fucked in your head.

[Socks]

Franck really livens this place up. :P

[Dirty F]

And did i read this correctly. Did you just call yourself a genius? Please tell me i understood that wrong.

[RyuLion]

Thanks for the stats!

[globofun]

Christopher Columbus never discovered America....... it's a conspiracy invented by the American Government ........like Apollo moon landings, JFK, 9/11 etc

[SwirlsGirl]

Originally Posted by globofun

Christopher Columbus never discovered America....... it's a conspiracy invented by the American Government ........like Apollo moon landings, JFK, 9/11 etc

There were Mayans, Aztecs, Incas, and Olmecs before Columbus came to the americas...

For historical accuracy Columbus never stepped foot on the American main land

Sure I can dance around in history class who wants to dance? We can certainly show you how deep the rabbit hole goes...

Why would the same 2 or 3 gfy posters take it SOOOOOO personally that I ask questions about stats anomalies....??

You guys dirty F, warchild, and the other 1 or 2 ass hats take it sooooooooo very personally when no one else takes it that personally I wonder why?

[Dirty F]

Piss off you dumb fucking nutjob. Go get help.

[globofun]

I'm an asshat now......FUCK OFF!

[SwirlsGirl]

Originally Posted by Dirty F

Piss off you dumb fucking nutjob. Go get help.

You never heard of Olmecs either? you really do have no shame in demonstrating your ignorance...

Alas the dumb fucking nut job knows more history than you....LOL

[Dirty F]

Yes, that's because you are a genius with the best converting site on the web. Or maybe because you live in a fantasyworld because of your mental illness. Hm, which one is it??

[livexxx]

A graph might help. What changes are you implementing on these forms. What is your ongoing QA/refactoring? Assume you are doing PDCA. How come it took 12 months for the last figs to spike low, or were they a monthly one off? If you have some A/B or MVT perhaps that is skewing things, or some load issues.

[Dirty F]

Originally Posted by livexxx

A graph might help. What changes are you implementing on these forms. What is your ongoing QA/refactoring? Assume you are doing PDCA. How come it took 12 months for the last figs to spike low, or were they a monthly one off? If you have some A/B or MVT perhaps that is skewing things, or some load issues.

Don't waste your time on helping that freakshow. She doesn't want answers. She just wants to hear there is some kind of conspiracy going on. She is a truther. This is how they work.

[ilnjscb]

Originally Posted by SwirlsGirl

You never heard of Olmecs either? you really do have no shame in demonstrating your ignorance...

Alas the dumb fucking nut job knows more history than you....LOL

Sara -

1. I think the sample size is a little too small. What appears to be anomalous may not be in context; at a minimum you'd need a few months to tell.
2. You have a hot ass, please post more nudes.
3. Your overall signups are the crucial figure, right? April 2012 doesn't even appear to be an outlier from that perspective.

These are all just opinions I thought I'd share with you.

[bean-aid]

Originally Posted by BV

After checking out archives of your site over the years, what stands out to me is now you disclose the price before having to load the ccbill join form. Where as in the past you had to load the form in order to find out the price.

Which should make your form submission percentage go up, not down. The thinking would be less clicks to the ccbill form because they already know the price, and a higher submission because they are more targetted clicks.

It's been live for 3 months... issue started on the 2nd of this month. It has returned to normalcy the past 5 or so days.

[Colmike9]

March always sucks, spring break time with less horny college kids at their desks.

I can make up any reason for it, but the point it when sales are down, just work harder. Problem solved.

[Dirty F]

Originally Posted by beaner

Which should make your form submission percentage go up, not down. The thinking would be less clicks to the ccbill form because they already know the price, and a higher submission because they are more targetted clicks.

It's been live for 3 months... issue started on the 2nd of this month. It has returned to normalcy the past 5 or so days.

Please tell us again how she is perfectly sane

[SleazyDream]

I like the site and concept and think it would sell quite well from what i see

[SwirlsGirl]

Originally Posted by ilnjscb

Sara -

1. I think the sample size is a little too small. What appears to be anomalous may not be in context; at a minimum you'd need a few months to tell.
2. You have a hot ass, please post more nudes.
3. Your overall signups are the crucial figure, right? April 2012 doesn't even appear to be an outlier from that perspective.

These are all just opinions I thought I'd share with you.

Many thanks and believe me I take all input and shared thoughts seriously.

I have to answer to affiliates, and my partners, and if we can not identify the source of the form submission disconnect then all we can do is come here and share and see if others have similiar disconnects

When I come in and put my reputation and sincerity on the line it is for my team...as far as I can tell their efforts have been commendable and even the ones counting our form clicks have said they can see something is a miss.

7 years in a row of double digit form submission percentages for april and now april 2012 is sitting at 8%

There should be 1 person at least somewhere who also had a record month of march for 2012 then followed by a record low for april 2012

Show me that person and I disappear quietly into the night...until I see that account or person I have no choice but to ask questions on how to improve my form submissions.

Time outs, slow page loads, glitches,bugs, are some of the things that can cause join form submission percentages to vary from site to site

The fact that I have such high form hits even after the price is shown tells me all I need to know about the general interest of my site, and satisfaction of our tour.

[SwirlsGirl]

Originally Posted by SleazyDream

I like the site and concept and think it would sell quite well from what i see

Thanks Sleazy very kind of you to make a comment like that in this climate Much appreciated

[SwirlsGirl]

Originally Posted by livexxx

A graph might help. What changes are you implementing on these forms. What is your ongoing QA/refactoring? Assume you are doing PDCA. How come it took 12 months for the last figs to spike low, or were they a monthly one off? If you have some A/B or MVT perhaps that is skewing things, or some load issues.

Hey Thanks for the graph LiveXXX that is very helpful and give us another perspective to consider
many thanks

[ilnjscb]

I guess this means no pics

[EddyTheDog]

Originally Posted by SwirlsGirl

Many thanks and believe me I take all input and shared thoughts seriously.

I have to answer to affiliates, and my partners, and if we can not identify the source of the form submission disconnect then all we can do is come here and share and see if others have similiar disconnects

When I come in and put my reputation and sincerity on the line it is for my team...as far as I can tell their efforts have been commendable and even the ones counting our form clicks have said they can see something is a miss.

7 years in a row of double digit form submission percentages for april and now april 2012 is sitting at 8%

There should be 1 person at least somewhere who also had a record month of march for 2012 then followed by a record low for april 2012

Show me that person and I disappear quietly into the night...until I see that account or person I have no choice but to ask questions on how to improve my form submissions.

Time outs, slow page loads, glitches,bugs, are some of the things that can cause join form submission percentages to vary from site to site

The fact that I have such high form hits even after the price is shown tells me all I need to know about the general interest of my site, and satisfaction of our tour.

What are you doing again?

[signupdamnit]

SwirlsGirl, I did some number crunching for you and with the limited data I see it does seem as if April 2012 is a bit abnormal and cause for some concern based on the form submissions (but approval % seems fine and please keep reading!).

What I suggest is to take the last 3-5 years of monthly data and do something like I did below except use all the data you have. Enter it on a site like this http://easycalculation.com/statistic...-deviation.php. That will give you the mean (simple average) as well as the Standard Deviation among other things. With the standard deviation you then add and subtract it from the mean (see example below). That will show you the limits for data to be within one Standard Deviation. Typically in about 70% of cases on average in a normal distribution data should be within one deviation. To see what is REALLY abnormal (usually less than a 5% chance of it happening) go out two deviations.

Your April 2012 form submission % is outside of one Standard Deviation so it is a little abnormal. But it's also only a low sample size I have and April isn't finished yet.

Keep in mind that it would be better if somehow you could eliminate traffic from tier 3 (and maybe tier 2) traffic from all these stats as well as other bad traffic like the google bot. Ideally these stats would track uniques and not raws too. If a spider or a bunch or surfers from a Korean forum hit your join page 1000 times in the month then that can mess up your stats unless you eliminate it to prevent that.

Here's some calculations I did using only the last five years values for April that you gave :

Mean is simple average. Very easily skewed. Don't rely much on it.

Standard deviation is a measure of how much the data varies. It's still subject to skew but much less so than the Mean. The more data points you have the more reliable it tends to become.

"Normal" is just what I call the range of one Standard Deviation from the Mean. To get it take the Standard Deviation and add it to the mean. That's your top range. Now take the Standard Deviation again and subtract it from the original Mean. That is your bottom range. On average about 70% of all data should be between this normal range. But that can vary depending on the data and the sample size. It suggests something is off when outside this range but it's still probable that it can happen. Bad luck.

"Somewhat normal" I didn't do but it's just like what I call "Normal" but with a range of two Standard Deviations from the mean. Typically 95% of data points would be within this range. Being outside it suggests something is wrong but it doesn't mean impossible. I would be looking to see what is wrong pretty hard if I saw a month outside this.

Form Hits

2008 - 184
2009 - 899
2010 - 462
2011 - 275
2012 - 589

Avg Form hits: 482
Standard Deviation: 282

"Normal" (approx. 70% of time on average with large sample size and normal distribution) : 200 - 764



Form Submissions

2008 - 71
2009 - 240
2010 - 89
2011 - 101
2012 - 50

Avg Submissions: 110
Standard Deviation: 75

"Normal" : 35 - 185

Approvals

2008 - 55
2009 - 178
2010 - 80
2011 - 53
2012 - 37

Avg Approvals: 81
Standard Deviation: 57

"Normal" : 24 - 138



Form Submission %

2008 - 39%
2009 - 27%
2010 - 20%
2011 - 36.73%
2012 - 8.49%

Avg Form submission %: 26.24%
Standard Deviation: 13%

"Normal" : 13.24% - 39.24%


Approval %

2008 - 77%
2009 - 74%
2010 - 89%
2011 - 52%
2012 - 74%

Avg Approval %: 73.2%
Standard Deviation: 13%

"Normal" : 60.2% - 86.2%

This is of limited use because it's only five data points and because of what I said. So the Standard Deviation is going to be huge and almost useless. It would be better like I said if you did every month for the last three or five years. ;)

You might have someone make a Excel spreadsheet for you so you can plot this stuff and have a nice graph showing all the data and see where the "Normal" ranges are and maybe get stuff like the margin of error. I bet something like this already exists somewhere. I hope this is helpful. Good luck!

[SwirlsGirl]

Originally Posted by signupdamnit

SwirlsGirl, I did some number crunching for you and with the limited data I see it does seem as if April 2012 is a bit abnormal and cause for some concern based on the form submissions (but approval % seems fine and please keep reading!).

What I suggest is to take the last 3-5 years of monthly data and do something like I did below except use all the data you have. Enter it on a site like this http://easycalculation.com/statistic...-deviation.php. That will give you the mean (simple average) as well as the Standard Deviation among other things. With the standard deviation you then add and subtract it from the mean (see example below). That will show you the limits for data to be within one Standard Deviation. Typically in about 70% of cases on average in a normal distribution data should be within one deviation. To see what is REALLY abnormal (usually less than a 5% chance of it happening) go out two deviations.

Your April 2012 form submission % is outside of one Standard Deviation so it is a little abnormal. But it's also only a low sample size I have and April isn't finished yet.

Keep in mind that it would be better if somehow you could eliminate traffic from tier 3 (and maybe tier 2) traffic from all these stats as well as other bad traffic like the google bot. Ideally these stats would track uniques and not raws too. If a spider or a bunch or surfers from a Korean forum hit your join page 1000 times in the month then that can mess up your stats unless you eliminate it to prevent that.

Here's some calculations I did using only the last five years values for April that you gave :

Mean is simple average. Very easily skewed. Don't rely much on it.

Standard deviation is a measure of how much the data varies. It's still subject to skew but much less so than the Mean. The more data points you have the more reliable it tends to become.

"Normal" is just what I call the range of one Standard Deviation from the Mean. To get it take the Standard Deviation and add it to the mean. That's your top range. Now take the Standard Deviation again and subtract it from the original Mean. That is your bottom range. On average about 70% of all data should be between this normal range. But that can vary depending on the data and the sample size. It suggests something is off when outside this range but it's still probable that it can happen. Bad luck.

"Somewhat normal" I didn't do but it's just like what I call "Normal" but with a range of two Standard Deviations from the mean. Typically 95% of data points would be within this range. Being outside it suggests something is wrong but it doesn't mean impossible. I would be looking to see what is wrong pretty hard if I saw a month outside this.

Form Hits

2008 - 184
2009 - 899
2010 - 462
2011 - 275
2012 - 589

Avg Form hits: 482
Standard Deviation: 282

"Normal" (approx. 70% of time on average with large sample size and normal distribution) : 200 - 764



Form Submissions

2008 - 71
2009 - 240
2010 - 89
2011 - 101
2012 - 50

Avg Submissions: 110
Standard Deviation: 75

"Normal" : 35 - 185

Approvals

2008 - 55
2009 - 178
2010 - 80
2011 - 53
2012 - 37

Avg Approvals: 81
Standard Deviation: 57

"Normal" : 24 - 138



Form Submission %

2008 - 39%
2009 - 27%
2010 - 20%
2011 - 36.73%
2012 - 8.49%

Avg Form submission %: 26.24%
Standard Deviation: 13%

"Normal" : 13.24% - 39.24%


Approval %

2008 - 77%
2009 - 74%
2010 - 89%
2011 - 52%
2012 - 74%

Avg Approval %: 73.2%
Standard Deviation: 13%

"Normal" : 60.2% - 86.2%

This is of limited use because it's only five data points and because of what I said. So the Standard Deviation is going to be huge and almost useless. It would be better like I said if you did every month for the last three or five years. ;)

You might have someone make a Excel spreadsheet for you so you can plot this stuff and have a nice graph showing all the data and see where the "Normal" ranges are and maybe get stuff like the margin of error. I bet something like this already exists somewhere. I hope this is helpful. Good luck!

WOW!!! Hey Signupdamnit....firstly I know that took some precious time to layout and I understand how precious time is...so please accept my sincere and profound thanks to you for that lengthy and detailed response!

I found it to be very insightful and intriguing as what you laid out about standard deviations and how to interpret statistical data.

I really appreciate the fact that you went out of your way to help me understand some things with more clarity as opposed to insulting or flaming me for these types of questions.

I think I will go back and input much more of the data into that website and see what more number crunching can be done to give us more accurate data to analyze.

We can not thank you enough for that contribution and I will keep you updated on our findings.

I feel somewhat relieved that some one of your credibility has gone public and agreed that there does *appear* to be an *issues* with april form submissions based upon the small sample of data provided.

[signupdamnit]

You're welcome I know how it feels to have to worry about all this stuff so I wanted to try to help. The 8% form submit you are seeing is outside one Standard Deviation so it appears a bit abnormal. When you figure in all those other points check to see if it is still outside of one SD or maybe even outside two Standard Deviations which I suspect it will be. It doesn't necessarily mean something has to be wrong but it suggests so. It could be CCBill, a slow loading form, hosting problems, error on the site, influx of abnormally horrible traffic, many things but it suggests something is up somewhere.

Here is a visualization of Standard Deviation which might help. It's not using your exact data but is just from Wikipedia.



The dark blue areas (labeled 34.1%) are what I call "normal". That is one Standard Deviation from the mean which is the center. 70% on average will fall within this.

The light blue areas (labeled 13.6%) are what I call "somewhat normal". That is two standard deviations. 95% on average will fall within this area statistically.

When you start seeing values outside the dark blue and light blue areas then there is likely a problem somewhere. If you suddenly start seeing a pattern recurring of months outside these areas then that would be very abnormal and suggest something has changed somewhere. This is one way to find abnormalities statistically in a proven way.

If you were to enter all the monthly data in for the last couple years then you'll get a completely different set of standard deviations and ranges than I have here and your data would be much more accurate with 36 - 60 months instead of my mere 5.