false

73235

html_restricted

Your question is quite tricky: it nearly forces to think about how to measure the effectiveness of each channel, what is quite funky mainly from the point when we are talking about not specially direct sales campaigns or even better go "offline". And in fact I'd rather say that it's impossible to answer- at least without seeing some details. Let's create a very simple modell: Assumptions: -you are doing only email and ppc (not even visitors from other sources) -your only message is "buy cheap make-my-logo-bigger-cream here" -you reach exactly the same people with email a cpc -in the worst case after 2 clicks they buy (the ones who buy at all) So after leaving out smaller, not important factors that may cause confusion still impossible to answer whether the first or the last click is "better". Use th "last click": if your customers are "go-compare-then-buy-asap"- then the "last click" alone will say that the email was useless. Then as a responsible mm you stop spending on email. Then you realize that you've lost all revenues as you customers are not just "go-compare-then-buy-asap" but also "would-never-think-about-make-my-logo-bigger-cream-if-sy-wouldnt-have-spammed-me-with-a-mail" type customers. Then in your next job you try another method (another company selling make-my-logo-bigger-cream): attribute revenue based on the "first click". In this case as your customers are "would-never-think-about-make-my-logo-bigger-cream-if-sy-wouldnt-have-spammed-me-with-a-mail" type folks you'll realize that all "first click" is from the email. As a resposible mm you asap stop cpc. Then you realize that all revenue is lost as you customers are not only "would-never-think-about-make-my-logo-bigger-cream-if-sy-wouldnt-have-spammed-me-with-a-mail" and "go-compare-then-buy-asap" type folks but also "delete-spam-mails-in-a-minute" type people so without cpc they'll simply buy from some of thousands of competitors selling "make-my-logo-bigger-cream". After such a successful carrier in your third job you know at least that don't stop either email-marketing or cpc else you'll have to look for a new position. And fortunately this company (selling domain-farm-eliminator foam- strictly applied on server side) has done a little bit less perfect job as the previous ones so some of the target audience is reached only by mail, some only by cpc and some with both, and also not so easily categorizable like previously. And as the domain-farm-eliminator foam is much more well know then the previous product sometimes people are searching for it so sometimes there are buyers directly from cpc, sometimes directly from email and sometimes through both. But strangely enough according to the information from the marketing analytics department: if the folks buy, then they buy within 24 hours from the first meeting with your ads (cpc or mail). And of course as a responsible mm you want to ensure that you make the most of your budget- what is the best ratio to spend on email marketing and cpc. From the previous experiences you suspect that probably using the first or the last click might not be perfect for all situation and you may start wonderring using both. But how? When a customer buys through cpc, but previously clicked on the email then 20 or 80% of the revenue should be attributed to the mail? Then after minute of thinking you'll find the fairly intuitive solution: Re=R(ROIec-ROIc)/(2ROIec) Rc=R(ROIec+ROIc)/(2ROIec) Where R stands for Revenue Re intuitively for the revenue (share) attributed for email Rc the revenue attributed to cpc-advertising, ROI is ROI, ROIc the ROI of "pure cpc advertising" ROIec is the ROI of clicks where the first is email second is on cpc (so the trackable conversion ruote is email>>cpc>>sales as in the original scenario). Obviously the same can be played when the conversion route is swapped cpc>>email... The funny thing about this very primitive formula: we came to it through a not really realistic modell, but usually it works if we take it out to the real life and apply it to attribute revenue to advertising channels based on the last 2 click- and suppose a reasonable timeframe: for example only consider email the "first click" if it was within a week of purchase and for example for cpc we can set the same timeframe. (which means we don't care what was for example 2 weeks ago). Another advantage if you previously "buried" yourself with enough data- you can use this idea to test the reasonable limits: easily can be extended to use include the information of all (supposing that it's finite) previous tracking points. So when you see that incuding the previous tracking point doesn't really make difference (at least when the attributed value tends to be near to 0 or negative)- then most likely you've reached the resonable limit. The same applies for the timeframe: you can include a function of time as a damping factor- and from enough data you will be able to determine the resonable limits (in fact estimate the damping time-function)- that for example a click on a cpc ad 3 month ago might have not had such a huge effect on the buying decision triggered by the current weekly spam.

Your question is quite tricky: it nearly forces to think about how to measure the effectiveness of each channel, what is quite funky mainly from the point when we are talking about not specially direct sales campaigns or even better go "offline". And in fact I'd rather say that it's impossible to answer- at least without seeing some details. Let's create a very simple modell: Assumptions: -you are doing only email and ppc (not even visitors from other sources) -your only message is "buy cheap make-my-logo-bigger-cream here" -you reach exactly the same people with email a cpc -in the worst case after 2 clicks they buy (the ones who buy at all) So after leaving out smaller, not important factors that may cause confusion still impossible to answer whether the first or the last click is "better". Use th "last click": if your customers are "go-compare-then-buy-asap"- then the "last click" alone will say that the email was useless. Then as a responsible mm you stop spending on email. Then you realize that you've lost all revenues as you customers are not just "go-compare-then-buy-asap" but also "would-never-think-about-make-my-logo-bigger-cream-if-sy-wouldnt-have-spammed-me-with-a-mail" type customers. Then in your next job you try another method (another company selling make-my-logo-bigger-cream): attribute revenue based on the "first click". In this case as your customers are "would-never-think-about-make-my-logo-bigger-cream-if-sy-wouldnt-have-spammed-me-with-a-mail" type folks you'll realize that all "first click" is from the email.  As a resposible mm you asap stop cpc. Then you realize that all revenue is lost as you customers are not only "would-never-think-about-make-my-logo-bigger-cream-if-sy-wouldnt-have-spammed-me-with-a-mail" and "go-compare-then-buy-asap" type folks but also "delete-spam-mails-in-a-minute" type people so without cpc they'll simply buy from some of thousands of competitors selling "make-my-logo-bigger-cream". After such a successful carrier in your third job you know at least that don't stop either email-marketing or cpc else you'll have to look for a new position. And fortunately this company (selling domain-farm-eliminator foam- strictly applied on server side) has done a little bit less perfect job as the previous ones so some of the target audience is reached only by mail, some only by cpc and some with both, and also not so easily categorizable like previously. And as the domain-farm-eliminator foam is much more well know then the previous product sometimes people are searching for it so sometimes there are buyers directly from cpc, sometimes directly from email and sometimes through both. But strangely enough according to the information from the marketing analytics department: if the folks buy, then they buy within 24 hours from the first meeting with your ads (cpc or mail). And of course as a responsible mm you want to ensure that you make the most of your budget- what is the best ratio to spend on email marketing and cpc. From the previous experiences you suspect that probably using the first or the last click might not be perfect for all situation and you may start wonderring using both. But how? When a customer buys through cpc, but previously clicked on the email then 20 or 80% of the revenue should be attributed to the mail? Then after minute of thinking you'll find the fairly intuitive solution: Re=R(ROIec-ROIc)/(2ROIec) Rc=R(ROIec+ROIc)/(2ROIec) Where R stands for Revenue Re intuitively for the revenue (share) attributed for email Rc the revenue attributed to cpc-advertising, ROI is ROI, ROIc the ROI of "pure cpc advertising" ROIec is the ROI of clicks where the first is email second is on cpc (so the trackable conversion ruote is email>>cpc>>sales as in the original scenario). Obviously the same can be played when the conversion route is swapped cpc>>email... The funny thing about this very primitive formula: we came to it through a not really realistic modell, but usually it works if we take it out to the real life and apply it to attribute revenue to advertising channels based on the last 2 click- and suppose a reasonable timeframe: for example only consider email the "first click" if it was within a week of purchase and for example for cpc we can set the same timeframe. (which means we don't care what was for example 2 weeks ago). Another advantage if you previously "buried" yourself with enough data- you can use this idea to test the reasonable limits: easily can be extended to use include the information of all (supposing that it's finite) previous tracking points. So when you see that incuding the previous tracking point doesn't really make difference (at least when the attributed value tends to be near to 0 or negative)- then most likely you've reached the resonable limit. The same applies for the timeframe: you can include a function of time as a damping factor- and from enough data you will be able to determine the resonable limits (in fact estimate the damping time-function)- that for example a click on a cpc ad 3 month ago might have not had such a huge effect on the buying decision triggered by the current weekly spam.

2009-03-02T01:40:10+00:00

true

3250

11060

2009-03-02T01:40:10+00:00