Every couple of days some one is asking a question dealing with, in some form or another, statistics. One of the most common is, "How long should I run a campaign?"
The answer is simple . . . if you know graduate-level statistics. 
I put together this tool which gives you all the information you need to know how to determine, given your current campaign stats, the probability that campaign will be profitable.
It is a DRAFT, and for STM members only. It works on a Win7 Excel 2010. That's all I can guarantee at the moment.
I would love your feedback, comments, requests, ridicule, etc. I will then clean it up and provide some decent training along with it.
Here is the download link:
VERSION 3: ROI Calculator v3
I also did a 15-minute walk through of the file. Apologies in advance, I got way too geeky. But, you'll get the point.
http://www.screencast.com/t/HmVZAI4CeeO
Update: Version 3 Video - http://www.screencast.com/t/cGZgdV3cGc << Still Geeky
Let me know what you think! Useful?
Thanks,
Walter
well damn...you just git on my awesome list
all i see is the #NAME error and nothing shows up.... probably way over my head..
@vidivo - Probably because you're running Excel 2007? I need to make it backwards compatible. It's on the list.
You really don't have to understand what's going on behind the scenes to use it effectively. Just plug in you campaign statistics, and it will tell you how likely you are to be profitable (and how profitable) for the campaign.
It will make sure you don't waste money over-testing a campaign, but it will also make you a ton of money by NOT letting you end a (potentially winning) campaign before you have enough data to make an informed decision.
This is awesome man. I've done a couple of stats classes and knew that it would be advantageous to apply to my campaign stats - but had no idea how lol.
Thanks a lot 
Sounds awesome will check it out soon!
This is awesome! Thanks for the helpful spreadsheet and training video (and brief stats lesson-- had to rummage through my brain from college stats class).
Dude sweet man. Luckily I'm a student so I can pick up a copy of Excel 2010 for next to nothing. Totally worth it.
working great on newest Excel on OS X 1.7 (Lion) anyway it can be adapted to use CPC or CPM more easily?
Is it working for directlinking? Cos views/clicks is same. It shows me 96% profitable on 2600 clicks with 3 conversions 
edit: yup working, i add one more zero in CPV by mistake. It's 100% NO.
not working for me - i get the #NAME? on scenarios.
wow. great stuff. downloading now!
ok watched the video. unfortunately some of the formulas dont work in OsX Numbers or Google Docs. Could you post the actual formula used in the cells and ill try and find the Osx/Google Docs equivalent.
I would love to see a couple of enhancements:
1) based on target profitability & probability calculate the number of views & budget required. Ie if I wanted to achieve 100% expected ROI with 75% probability of profit, how long would I need to run the test for
2) And then as an enhancement on this, if I was running 3 landing pages, and 4 offers - what would be the views per target, lander, offer before I can make the go / no go decision
awesome thanks..is it possible to replace the cpv with cpc..so that it can be used for cpc campaigns too?
Great share, thanks!
Awesome stuff......but it doesn't work at all on my abacus......may have to upgrade to slide rule.
More please
Anyone got this working with OpenOffice Calc?
I can say it works awesome on excel and can be used for anything banners, ppv, ppc.....i have tested it against winning and loosing camps and its rather accurate(past data).......def something I will be using daily
Could someone please post the formulas for cells B8, C8 and D8 if you have Excel and the downloaded spreadsheet is working for you. Thanks!
Okay . . . this should work with versions of Excel earlier than 2010. (I used the legacy stat formulas.)
ROI Calculator v2
Thanks for all the feedback thus far. I want to make this a usable as possible, and will start to incorporate the suggestions and extend the functionality.
One thing I'll definitely be adding is a comparison calculator. For instance, if I had two campaigns, one with 200 clicks and 1 conversion, and the other with 300 clicks and 2 conversions, how confident can I be that one is the winner?
I also plan to incorporate tijn suggestion of making a predictive table of stopping criteria (i.e., stop 200 impressions and 0 conversions, or 450 and 1 conversion, etc.)
Please don't be shy about asking what you would like to see!
The formulas are:
L2: =(H3+(((ABS(NORMSINV((1-$D$4)/2)))^2)/2))/(H2+((ABS(NORMSINV((1-$D$4)/2)))^2))
L3: =(H4+(((ABS(NORMSINV((1-$D$4)/2)))^2)/2))/(H3+((ABS(NORMSINV((1-$D$4)/2)))^2))
L4: =($L$2*$L$3*$D$2-$D$3)/$D$3
B8: =CRITBINOM($H$2,$L$2,RAND())/$H$2
C8: =CRITBINOM($H$3,$L$3,RAND())/$H$3
D8: =(B8*C8*$D$2-$D$3)/$D$3
Hope that helps!
Awesome stuff man! T
Thanks for sharing.
awesome share..will save me alot of time!!
Oooh .. nearly there for me with the new version. Column CR is OK now but column CTR and ROI give me #VALUE! Grrr......
yeaaah this is sweet dude! cheers inversion 
Awesome Stuff MAN!
Thank you.
Thanks Inversion...popped in some stats on my newest campaign & looks 81% profitable. this tool is awesome you're like the Jonas Salk of iM
cool.
now we can talk about what it takes to find a winner as far as # & $
Lucrumnet
thanks for sharing V2 - this one is working for me now.
Just did some back testing and discovered that I may have bailed on a few camps too soon since their % probability of profit was >75%.
However, these camps had no conversions. Therefore, I wonder about the reliability of this based only on views, clicks, cost and payout with no conversions.
If you're working with a really crappy offer that doesn't convert I don't how this is gonna help.
Any thoughts
Thanks Walter. This is the sum of all awsomeness. (WR x ROI = A
)
wow... thanks inversion.... this thing is GANGSTER so far... keeps getting better.
thanks man, this is sick!

THIS IS AWESOME!