The Problem With U/S/W Analysis
Put simply; it’s about the nature of averages. Even today, the problem is always in chain store systems with huge, store-level variations in demand for premium brands.
A consumer brand’s sales in any chain will vary in U/S/W based simply on variations in-store traffic. But chain store ‘systems’ also spread across zip codes, varying widely in cultural nuances, class makeup, and relative interest in premium CPG items. The Safeway location that sells the lowest amount of Lay’s potato chips (due to low traffic) may sell no units of your premium offering. Often, this is what I find.
If you’ve looked at a distributor sell-in report by store, you’ll see exactly what I’m talking about. The highest traffic stores will pull up your system U/S/W average mathematically in a distorted fashion, creating the illusion of decent, system-wide performance. In reality, 10% of the chain’s stores might carry the total number you’re looking at, while 30-40% of the stores are about to get delisted for lack of sell-through. This is your initial warning sign if you have many stores on a new account that haven’t re-ordered.
If you have low awareness of the critical zip codes where a chain operates, the velocity ‘spread’ will be painfully large. Lack of awareness, not shelf visibility, is the number one reason brands get delisted from a % of stores they just sold into the year before. Sometimes faster. It depends on how sleepy the category is.
The one good thing about U/S/W’s tendency to overestimate your actual velocity is that it causes you to produce more product than you probably need. In other words, it’s NOT a lousy velocity metric to use for forecasting production needs. It should contain some built-in padding for your supply chain. So, use it there, not to understand growth performance.
For more in-depth tips on diagnosing your growth like a pro, check out my course – Scrappy POS Analytics for Founders.