We hope you aren’t tired of reading about velocity! Velocity and distribution (which go hand in hand, as you’ll see below) are the topics we are asked about most often. We’ve reviewed our reader correspondence and pulled out a few more key points we wanted to share:
Question 1: Is Velocity Independent of Distribution?
Q. Can you help me understand my brand trends? My volume is up 12%, my distribution is up 20%, but my velocity is down -10%. I’m new to analyzing syndicated CPG data and my boss says that velocity is not “affected” by distribution. So what could be causing my brand’s velocity decline?
A. This is a really common point of confusion! Velocity adjusts for distribution but it is not totally independent of distribution. Velocity is a product’s average sales rate in its current locations. Therefore, you can usefully compare one product’s velocity to another product’s velocity, even if the second product has wider or narrower distribution. But that doesn’t mean that where you are distributed is not impacting your velocity.
In the reader’s example, sales rate in the newer stores must be poorer, on average, than in the original stores. That logically leads, with distribution gains, to velocity declines even though you have overall sales growth. And that’s a common pattern – the last stores to take your product often have the weakest sales rate.
That doesn’t mean your velocity will always decline when your distribution goes up, of course. Products that are expanding their distribution without eroding their average sales rates are exhibiting strong performance. It suggests these brand owners are getting into the right new markets or have a very strong consumer proposition or are backing their expansion with excellent marketing. But the opposite pattern, of eeking out additional distribution in marginal or weaker markets, and thereby reducing average sales rates for your product, is one we have seen many times.
You’ll also see this pattern (distribution and velocity moving in oppositive directions) work in reverse too. As you lose distribution, your overall (average) velocity will go up because the stores with the poorest sales will be the first to discontinue your product.
Advanced P.S: If your distribution measure is TDP rather than % ACV, then this “declining velocity with rising distribution” can also be seen when you introduce new items that are weaker than existing items. TDP incorporates reach (the number of doors) and depth (the number of items) into the distribution metric. So weaker than average gains in either component of TDP can lower your average sales rate when TDP is the denominator in your velocity metric.
Question 2: What Is Base $ per TDP?
Q. There are so many different velocity measures in my database. It’s hard to sort them all out. For example, what does Base $ per TDP tell me?? And if there is a negative % change, what should I take from that?
A. Base dollars is an estimate of your sales without the incremental impacts of trade promotion. Base dollars are a function of everyday price, competitive activity, the varieties in your line, consumer promotion, and much more. Read more about base versus incremental volume here. TDP is the sum of %ACV for all products in your product group (brand, category, whatever level you are analyzing). So Base $/TDP is a velocity measure that gives you a base sales rate averaged over your current distribution.
What should you take from a negative % change in this measure? As we learned in Question 1, it’s very important not to look at this measure in isolation. You need to look at what’s happening with your distribution. Is distribution up? If so, it’s not shocking or even unusual to have a decline in your velocity. You may need to watch your step if the newer retailers are very weak but the trend is not necessarily disastrous. Is distribution down? If so, that is concerning because even where you maintained distribution, presumably your strongest retailers, your base business is in decline.
And if you want to dig deeper into the drivers, looking below the surface metrics, you’ll know to start with things like everyday price and competitors rather than trade promotion, since we’re talking about Base $ per TDP.
Question 3: Best Velocity Measure Over Time?
Q. . I have always used $/%ACV as my velocity metric. However, I read your article about how $ per MM$ACV is the better metric to use when comparing products across markets. It got me thinking that there could be a metric that’s best to use across time. It seems like a negative quality of the $ per MM$ACV metric is that inflation would get rolled into this metric and overstate your distribution. By that I mean that even if you stay in the same stores in both periods, your MM$ACV would increase just from higher prices.
A. You were concerned that inflation would impact $ per MM$ACV but actually that measure will better control for inflation than $/%ACV. $ per MM$ACV is item dollars divided by total store dollars, so inflation will impact both the numerator and denominator. By contrast, with $/%ACV, inflation is only impacting the numerator.
Question 4: Best Velocity Measure for Channel Growth?
Q. Could you help me with finding appropriate velocity metrics? I analyze data for for multiple brands and the most granular I get is at the channel level (not retailer). I typically default to $/TDP. However, I have been wondering if I should also look at Average Weekly Dollars Per Store Selling Per Item. Can you clarify the difference between these two measures and how to think about it at the channel level (both for brand-level data and SKU-level data)?
A. Both of those metrics can be used at any product group level: sku, brand, category, etc. Both metrics adjust for the number of items in distribution, essentially giving you an average of the skus in the product group. Anything that uses TDP in the denominator will accomplish that because you are dividing by the sum of ACV of all products in your product group. Your alternative metric accomplishes the adjustment dividing the basic velocity metric (Average Weekly Dollars Per Store Selling) by the number of items in the product group. So both are valid and comparable velocity measures.
At the SKU level, TDP is equivalent to % ACV in the $/TDP measure. And the “per item” element becomes irrelevant in your alternative metric when you are at the SKU level. So they are equivalent there as well. They work fine as sku level velocity measures as well because they mathematically simplify themselves at the SKU level.
The big difference between them, then, is that one is weighted by ACV ($/TDP) and one is not (Average Weekly Dollars Per Store Selling Per Item). Weighting by ACV gives a product “more credit” if it’s in stores with higher overall sales across all products.
Why are you considering the change? I personally stick with ACV weighted measures because they do a better job of telling me the relative strength of my brand/product. I usually think the main reason for switching to a “per store” measure is because it’s preferred by your audience (since it is a more intuitive and tangible meaure) and you will not have credibility unless you use it. If that’s the case, then I would compare the numbers between the two metrics and see if you would draw different conclusions if you did make the switch. If you don’t feel you are drawing incorrect conclusions or changing your conclusions and your audience is happier seeing “per store” numbers, go for it.
Finally, I don’t think it’s any different thinking about it at the channel vs. the retailer level. It’s a question of whether you want to use the weighted or unweighted measure.
If you enjoyed this article, subscribe to future posts via email. We won’t share your email address with anyone.
This is such a great overview of some important velocity questions. While I was reading it, another view came to mind in reading this part of what you wrote:
“In the reader’s example, sales rate in the newer stores must be poorer, on average, than in the original stores. That logically leads, with distribution gains, to velocity declines even though you have overall sales growth. And that’s a common pattern – the last stores to take your product often have the weakest sales rate.”
A thought is that the consumer base might not be very expandable beyond the original distribution — that is, it’s mostly the same consumers are buying across the legacy *and* new outlets and the consumer base didn’t grow that much. So with the added distribution but similar buying rate, overall velocity drops.
I think that sometimes happens with specialty brands. Let’s say they were in Whole Foods originally, then they pick up Stop & Shop. The velocity drops at Whole Foods due to buyer migration, and it’s so-so at S&S. Overall, it drops.
It’s hard to prove this (or anything else) is exactly what’s happening when you see unexplained velocity declines accompanying distribution gains, but it is a good theory and one that can be investigated further.
Thanks Scott! I hadn’t thought of how buyer migration might be contributing to this pattern. You have a lot of experience with specialty brands so I appreciate your expertise. And, as you say, it’s a theory that could actually be investigated.