MLV: Where You Feature Items Matters

In a recent project, I attached customer repurchase activity next year to the items a customer purchased last year. Then, I build a regression equation to determine if various merchandise categories and/or marketing channels and/or other factors play a role in helping an item launch customers to loyal status.

Here's what I learned (after controlling for customer quality).

• Items featured in catalogs caused customers to become "more loyal" in the future.
• Items featured in email campaigns caused customers to become "more loyal" in the future.
• Items that skewed to search (paid or natural) caused customers to become "less loyal" in the future.
• Items that were more likely to be tied to discounts/promotions were items that caused customers to become "less loyal" in the future.

Remember - we're analyzing customers who bought an item last year and determining the repurchase rate of those customers in the future - appending future repurchase back to the items customers bought last year. There's some chicken/egg stuff going on, of course (putting best items in catalogs, for instance).

Pay attention to the search finding ... Google plays a key role in influencing the future trajectory of your business. If Google brings customers who are less likely to buy in the future, and the items those customers buy become winners, then Google has mucked-up your brand ... they drive low-rebuy customers to you who like specific items that cause your merchants to offer more of those items (thereby attracting more low-rebuy customers).

Think about that for a moment - my project work shows that this dynamic is happening.

P.S.:  Facebook + Google get 75% of all digital ad dollars (click here). Your future requires you to find new customers outside of Facebook / Google / (and for catalogers, catalog co-ops).