Comments on: Marketing Attribution Models http://blog.jimnovo.com/2007/08/28/marketing-attribution-models/ Moving from a Low Accountability to a High Accountability Business Model Fri, 4 Jul 2008 05:46:26 +0000 http://wordpress.org/?v=2.5.1 By: Jim Novo http://blog.jimnovo.com/2007/08/28/marketing-attribution-models/#comment-25942 Jim Novo Mon, 02 Jun 2008 23:39:48 +0000 http://blog.jimnovo.com/2007/08/28/marketing-attribution-models/#comment-25942 Perhaps the CPG folks are so into it because that's all they can really do? When you have actual customer behavior data, and are acting on individuals and not "the masses", the need for mix modeling is substantially reduced, IMHO. Goes hand in hand with whether you are a big spender on untrackable media. If so, then have at it! Perhaps the CPG folks are so into it because that’s all they can really do?

When you have actual customer behavior data, and are acting on individuals and not “the masses”, the need for mix modeling is substantially reduced, IMHO. Goes hand in hand with whether you are a big spender on untrackable media. If so, then have at it!

]]> By: mjwolfe http://blog.jimnovo.com/2007/08/28/marketing-attribution-models/#comment-25761 mjwolfe Sun, 01 Jun 2008 20:56:10 +0000 http://blog.jimnovo.com/2007/08/28/marketing-attribution-models/#comment-25761 It seems to me that most of the discussion on this blog is focused on "silos"...web analytics, sales force automation, crm analytics, etc, whichi s all good.....but only marketing mix modeling addresses the "whole MROI enchillada". Sure, there are some limits to how many variables you can put in a model, but you would be surprised how detailed it can be. It can address offline marketing, online marketing, direct marketing, promotions, etc, all in one swhoop. These silo tools are great for drilling down into the detail that mix modeling doesn't do, but none of these tools address the whole marketing investment enchillada. The problem with mix modeling now is that most of it is in CPG, but many other businesses can't do it, and don't get it. When they do, look out! It seems to me that most of the discussion on this blog is focused on “silos”…web analytics, sales force automation, crm analytics, etc, whichi s all good…..but only marketing mix modeling addresses the “whole MROI enchillada”. Sure, there are some limits to how many variables you can put in a model, but you would be surprised how detailed it can be.
It can address offline marketing, online marketing, direct marketing, promotions, etc, all in one swhoop. These silo tools are great for drilling down into the detail that mix modeling doesn’t do, but none of these tools address the whole marketing investment enchillada.

The problem with mix modeling now is that most of it is in CPG, but many other businesses can’t do it, and don’t get it. When they do, look out!

]]> By: Marketing Mix Modeling - Analytical approach to identify right marketing channel http://blog.jimnovo.com/2007/08/28/marketing-attribution-models/#comment-25453 Marketing Mix Modeling - Analytical approach to identify right marketing channel Fri, 30 May 2008 08:36:24 +0000 http://blog.jimnovo.com/2007/08/28/marketing-attribution-models/#comment-25453 Marketing Mix Modeling involves breaking up of sales volume into various components, and analyzing spend on each of them to calculate ROI from each of these components. With so many marketing channels in hand, particularly on web, it becomes too complex to analyze each channel. On web, one can extract massive amount of unstructured data, but then to extract meaningful information, you need to deploy analytical data models. Using predictive analytics, analysts can forecast sales through each of these channels and optimize the marketing spends to gain maximum value. In the last 10 years many Fortune 500 companies such as P&G, Kraft, Coca-Cola and Pepsi have made Marketing Mix Modeling an integral part of their marketing planning. Marketing Mix Modeling involves breaking up of sales volume into various components, and analyzing spend on each of them to calculate ROI from each of these components. With so many marketing channels in hand, particularly on web, it becomes too complex to analyze each channel. On web, one can extract massive amount of unstructured data, but then to extract meaningful information, you need to deploy analytical data models. Using predictive analytics, analysts can forecast sales through each of these channels and optimize the marketing spends to gain maximum value.

In the last 10 years many Fortune 500 companies such as P&G, Kraft, Coca-Cola and Pepsi have made Marketing Mix Modeling an integral part of their marketing planning.

]]> By: Are PPC Reports Inaccurate? Let’s Count the Ways (Friday Quiz Winner) http://blog.jimnovo.com/2007/08/28/marketing-attribution-models/#comment-16698 Are PPC Reports Inaccurate? Let’s Count the Ways (Friday Quiz Winner) Mon, 21 Jan 2008 12:58:17 +0000 http://blog.jimnovo.com/2007/08/28/marketing-attribution-models/#comment-16698 [...] Attribution is another systemic issue – people visit multiple times via multiple sources and at some point you (or your reporting software) has to decide which visit source gets the revenue credit for an eventual sale, if cookies or offline didn’t make it impossible to track in the first place! (Some very thoughtful posts here, here, and here.) The industry default is ‘last click’ but some analytics packages (such as SiteCatalyst and Coremetrics) provide control over that. (related post here) [...] [...] Attribution is another systemic issue – people visit multiple times via multiple sources and at some point you (or your reporting software) has to decide which visit source gets the revenue credit for an eventual sale, if cookies or offline didn’t make it impossible to track in the first place! (Some very thoughtful posts here, here, and here.) The industry default is ‘last click’ but some analytics packages (such as SiteCatalyst and Coremetrics) provide control over that. (related post here) [...]

]]> By: John Dawson http://blog.jimnovo.com/2007/08/28/marketing-attribution-models/#comment-5675 John Dawson Thu, 30 Aug 2007 08:05:02 +0000 http://blog.jimnovo.com/2007/08/28/marketing-attribution-models/#comment-5675 Hi Jim, I'm a marketing mix analysis guy (without a PhD it has to be said) and I think what we're really looking at here is a case of risk reduction. Is it more risky to attribute sales using some arbitrary system or better to use what clean data there is to create a precise "model" from the data. I'd say the latter with a caveat which is that models need to be validated and the measurement error acknowledged. The good news is that this is possible whereas with attribution it will never be possible. However the bad news is that if "measurements" are unreliable then it takes a very logical mind to acknowledge the potential for forecast error. One final point on attribution - i've rarely seen it done where three teams get together and give each other a third of the sales for three activities. Usually, they each try to claim as big a chunk as possible with the result that more sales are claimed than exist on the books. Hi Jim,

I’m a marketing mix analysis guy (without a PhD it has to be said) and I think what we’re really looking at here is a case of risk reduction.

Is it more risky to attribute sales using some arbitrary system or better to use what clean data there is to create a precise “model” from the data. I’d say the latter with a caveat which is that models need to be validated and the measurement error acknowledged.

The good news is that this is possible whereas with attribution it will never be possible. However the bad news is that if “measurements” are unreliable then it takes a very logical mind to acknowledge the potential for forecast error.

One final point on attribution - i’ve rarely seen it done where three teams get together and give each other a third of the sales for three activities. Usually, they each try to claim as big a chunk as possible with the result that more sales are claimed than exist on the books.

]]> By: Kevin Hillstrom http://blog.jimnovo.com/2007/08/28/marketing-attribution-models/#comment-5623 Kevin Hillstrom Wed, 29 Aug 2007 05:23:09 +0000 http://blog.jimnovo.com/2007/08/28/marketing-attribution-models/#comment-5623 :) I had a CEO recently call me about this very topic. He understood what you are writing about, he was looking for a reliable and reproducible way to understand the final outcome of the marketing mix. It was odd to hear a CEO And you're right ... this topic has been floated around since the 70s, in retail. Vendors looking to maintain their share of business, and catalogers hoping to validate the effectiveness of paper are jumping on the fractional allocation bandwagon. :)

I had a CEO recently call me about this very topic. He understood what you are writing about, he was looking for a reliable and reproducible way to understand the final outcome of the marketing mix. It was odd to hear a CEO

And you’re right … this topic has been floated around since the 70s, in retail. Vendors looking to maintain their share of business, and catalogers hoping to validate the effectiveness of paper are jumping on the fractional allocation bandwagon.

]]>