tag:blogger.com,1999:blog-32202893.post7865570511410828020..comments2023-10-18T08:32:17.510-07:00Comments on Kevin Hillstrom: MineThatData: Square Inch Analysis (SQUINCH) and Monthly Item Profitability ReportingUnknownnoreply@blogger.comBlogger4125tag:blogger.com,1999:blog-32202893.post-10395314172067212102008-05-01T08:14:00.000-07:002008-05-01T08:14:00.000-07:00Let me ask the question of you a little differentl...Let me ask the question of you a little differently?<BR/><BR/>You have two items. <BR/><BR/>The first item generated $20,000 sales, but $5,000 was spent on catalog advertising of this item and $5,000 on paid search.<BR/><BR/>The second item generated $20,000 sales, but $0 was spent on catalog advertising, and $6,000 was spent on paid search.<BR/><BR/>Tell me how you change your promotional strategy for each item?MineThatDatahttps://www.blogger.com/profile/14014200122021988374noreply@blogger.comtag:blogger.com,1999:blog-32202893.post-38910963421482815432008-05-01T05:28:00.000-07:002008-05-01T05:28:00.000-07:00I love the concept and constantly struggle with th...I love the concept and constantly struggle with the most appropriate way to do this, but how does it translate to something which is actionable from a promotional planning perspective across channels? It is nearly impossible for most companies to manage at this item level detail across all ongoing promotions, when most companies are dealing with hundreds/thousands of products and hundreds of promotions per year. <BR/>Automation of this would be one solution but merchandise mix requires some level of the human element as well. Thoughts?Anonymousnoreply@blogger.comtag:blogger.com,1999:blog-32202893.post-27187062739316233272008-04-30T09:07:00.000-07:002008-04-30T09:07:00.000-07:00The problem you describe has been around for more ...The problem you describe has been around for more than a hundred years.<BR/><BR/>A hundred years ago, Sears shipped homes to rural customers who received Sears catalogs. They mailed maybe two or three big books a year to a customer.<BR/><BR/>If Sears elected to not mail one of those catalogs, customers purchased from the other catalog, and purchased different merchandise.<BR/><BR/>The reality is that, for the most part, we won't be able to answer your question.<BR/><BR/>However, we can take steps toward answering your question. Few marketers want to take these steps.<BR/><BR/>First, we should execute mail/holdout tests for e-mail campaigns. We should offer ten versions of an e-mail campaign, each offering a different merchandise assortment. We then measure item-level, department-level, and total sales across each e-mail version (including a holdout group that doesn't receive the e-mail campaign).<BR/><BR/>This answers your question. Almost nobody wants to do this.<BR/><BR/>Same thing with catalog marketing. Go ask the folks at L.L. Bean about "incrementality". They are probably best at measuring the influence of combinations of catalogs on items sold across channels. They set up sophisticated tests to accurately measure how catalog advertising impacts other forms of advertising across phone, web and store channels.<BR/><BR/>But again, almost nobody wants to do this kind of testing.<BR/><BR/>Paid search testing isn't terribly difficult, either. Turn on various keywords one week, turn them off the next week, and watch sales change.<BR/><BR/>Many folks go with "matchback analytics". They make up allocation rules (rules that may not be based on reality), and then perceive they're doing well. And they are probably doing better than folks who do nothing.<BR/><BR/>So those are a few ideas.MineThatDatahttps://www.blogger.com/profile/14014200122021988374noreply@blogger.comtag:blogger.com,1999:blog-32202893.post-45958855847090019472008-04-30T08:00:00.000-07:002008-04-30T08:00:00.000-07:00The million dollar question (or the $3115 question...The million dollar question (or the $3115 question if you use your example): How do you allocate marketing costs? If someone gets an email about a widget and ends up buying a doo-dad which product pays the bill?Anonymousnoreply@blogger.com