Measures how fast a product is moving in the stores where it is in distribution. Use when comparing across markets, when comparing products with different distribution levels. Used most often by Sales (in item ranking reports). Calculated by dividing some measure of sales (e.g. Units or Dollars) by the market’s All Commodity Volume in millions.
Here’s a post all about velocity measures: Velocity: How Well Your Product REALLY Sells. And this post is All About ACV.
Can you compare $MM ACV across retailers in different channels (i.e. kroger $MM ACV vs. Costco)?
Yes! In fact, you should always use a Sales per Million velocity measure when comparing different retailers, regardless of channel. You should not use a Sales per Point velocity measure across retailers since a “point” of distribution is not the same in Kroger as it is Costco, or even in Kroger as in other Grocery retailers.
Does this mean that comparing 2 retailers using Sales per Million normalizes the actual sales dollar at each retailer to that retailer’s share of market?
Yes but remember that it’s also taking into account your distribution at a particular retailer. The “per Million” part is MM$ACV in stores where your product is scanning, not MM$ACV for the entire retailer (unless your distribution is 100% at that retailer).
If lookingat a product within one category would you not need to normalize for velocity for that banner within that category given ACV is at the all commodity level? Said differently if you look at the Sales per MM ACV in a banner that over indexes on the category of the product you’re looking at (lets say per MM ACV the banner as a whole sells twice as much chocolate as another banner) than comparing the velocity per MM of one banner to another within chocolate isnt perfectly apples to apples as you’d expect to have twice the velocity in the banner that has twice the category velocity. Does this make sense?
Interesting question! Just because a category over indexes in a certain retailer (or other geography) does not mean that all brands do there also. Depending on what question you’re trying to answer, you may want to look at BDI and CDI (brand development index and category development index) instead. Those control for a brand’s or category’s sales in a geography compared to the population in the geography. Please use the Contact form if you’d like to discuss further. And…thaks for the reminder that I’ve been meaning to write a post about BDI and CDI for a while now. 😉
Was there ever a post about BDI/CDI written? Would be interested to read it! TY!
No, we haven’t written a post on that topic. Thanks for the suggestion!
what is the difference between Total US Multi-Outlet and Total U.S. CRMA Multi-Outlet?
CRMA stands for Competitive Retail Marketing Area. It’s the market definition for a particular retailer. For example, the CRMA for Kroger would be a combination of all the competitive stores, including Kroger, in the geographic area served by Kroger. There would not be a CRMA at the total U.S. level, only for a specific retailer.
If you’re checking the velocity of an item in the entire category (Total US- XAOC), would you use Sales per $MM ACV or Sales per pt of dist?
As long as you are only looking within one geography (even if it’s Total US XAOC), either measure is fine and will give you the same ranking and relative values between items. It’s when you are looking at different geographies that the velocity measure matters. If you are comparing the velocity of the same product across markets, then you should use Sales per $MM ACV. That’s because a point of distribution is not the same in say, New York as it is in Phoenix since New York is a much bigger market. But $1MM ACV is always the same no mater which market (or retailer) you’re in.
If you are looking at a defined selling period (ex. 6 weeks vs 7 weeks) over multiple years (ex. 2013 and 2014) and %ACV is increased, would you expect the $MM/ACV to go up or down? I am looking at a particular product in 1 channel where the distribution doubled from year 1 to year down however the $/$MM ACV has gone down.
First thing to keep in mind is that when you are comparing velocity measures between two periods, make sure that the periods are the same length. So 6 weeks vs. 6 weeks, for example. If the second period is longer than the first, then the velocity will almost always be higher. It is best to compare a period vs. the same period in the previous year ago, otherwise the velocity can change due to seasonality (if you’re looking at a product that is at all seasonal).
To answer your real question…velocity can go up or down as distribution increases, there’s no typical expectation. It’s a matter of how the product is performing in the new stores vs. the old ones. If it’s a new product, then maybe you’d expect that it would have higher velocity in the stores that carried it first (retailers supporting the new product more, less competition from other retailers). Then velocity would go down as more stores carry the new product. I see that you’re looking at $ sales per $MM ACV. It’s possible for that to go down if pricing was lower in the second period but physical volume was flat or even increased a bit.
Hope this helps!
Is there a way to equate $ / MM ACV opportunity to a $ volume opportunity.
For example, Retailer X sells $5,000 less $ / MM ACV than retailer Y. Retailer X $ Volume is $200K and retailer Y $ volume is $1.5MM. Retailer x ACV is 80 and retailer y ACV is @ 95.
Thanks!
Thanks for your question, Lauren! I will address this more fully in a post since there is no way to include nice looking tables/data examples in the Comment section. But the short answer is…
You don’t need to use the $/$MM ACV or even look at Retailer Y at all to calculate the dollar opportunity for Retailer X if they increase distribution. if sales at Retailer X are $200,000 and %ACV distribution is 80, then $/point of distribution is $2,500 (= $200,000 / 80). So for every point of distribution that is gained, Retailer X would sell $2,500 more. if distribution increases by 5 points to 85, then sales increase by $12,500 (= 5 * $2,500).
Hope that helps!
Can you sum $ per Million from a number of products to give you the velocity of a custom set?
Thanks for the question, Joe! Unfortunately, velocity is not an additive measure, so no, you cannot just add together the velocity of individual items to get the velocity of the aggregate of those items. Any measure that has any form of ACV in it (like $ per $MM ACV) is not additive. In order to get that measure for an aggregate that does not exist in your database, you would have to know the ACV distribution of the group of items as a whole.
Hope this helps.
Hi Sally,
I am struggling to calculate the sales per million. I have a product which has a sales of $63,765,308, % ACV Distribution 98, Market ACV 40. Can you please show me how do I calculate this to get the sales per million.
To calculate Sales per Million: Sales / (% ACV * Market ACV expressed in millions)
Market ACV is often already expressed in millions but you might need to divided the market ACV by 1MM depending on your data source
BUT, something is off with the numbers in your case. You have product sales of $63 million. Let’s assume your Market ACV is already in millions, because it’s a low number, so that is $40 million. Not possible – Sales for one product can’t be higher than Market ACV.
Perhaps your sales number is annual and the market ACV number is weekly? That is one possibility but you will have to investigate further. You have to make sure you have a time period match between the product sales and the market ACV.
Given $ Sales per MM ACV of product X in a particular market is $493, $ Sales is $9,854,814 and %ACV is 91, can the Market ACV $ be calculated?
Thanks,
Ann
You can do that calculation but you need to use a product that has 100% distribution or the calculation will not be correct. Here is a link to a post on how to do this:
I don’t see the link, can you re-post it please? And thanks for answering my other question so fast!
Sorry about that – here is the link to my post on calculating retailer ACV:
https://www.cpgdatainsights.com/understand-your-database/estimate-retailer-acv/
I have retailer-level data for 2 retailers for a category I am looking at, at the UPC level, xAOC. When I use $/$MM ACV in Retailer A data, does that metric mean $/$MM ACV for Retailer A, or for all xAOC? In other words, is a $/MM ACV metric “diluted” by say Walmart’s very high share of overall market, compared to a smaller retailer with smaller share of ACV? I have assumed that selling $1 in a retailer with smaller share of ACV is better than a $1 in a retailer with a larger share of ACV. Thanks!
When you have data for an individual retailer, all measures are for that retailer only. The “per million” in Sales per Million is ACV for that retailer.
Sally
I would like to find out my points of distribution using my IRI data and ACV%.
I don’t get my Sales per million data point so what can I do?
If you don’t have Sales per Million, you need to use sales per point of distribution as your velocity measure. Unfortunately, this will not allow velocity comparison across markets but, within a market or retailers, sales per point of distribution is a solid measure. Please refer to this full article about velocity:
Velocity: How Well Your Product Really Sells
Hi Sally,
I have two, maybe silly questions:
1. If I have a retailer’s $/MM ACV, can I compare that to its respective market/geography’s $/MM ACV?
2. Retailer X has a higher $/MM ACV than retailer Y, yet retailer X has much lower total $ sales than retailer Y. What does this mean?
Thank you!
These are not silly questions!
1. Yes, you can compare $/MM ACV between a retailer and a market (or any other geography). That’s an advantage of using that velocity measure – you can compare across different geographies.
2. It is quite common for a product to have a higher velocity but lower sales in one retailer vs. another. In your example, you can interpret it as the product moves faster in Retailer X but Retailer Y is probably a bigger chain in general, with higher overall sales and more shoppers.
Thanks for the reply! On my second question, yes Retailer Y is a bigger chain so that may explain why velocity is lower than Retailer X. I had thought that $/MM ACV can be interpreted as “for every $1MM dollars, item/brand sold X ” Doesn’t $/MM ACV account for stores sizes?
You’ve misunderstood the answer to your previous question.
$/MM ACV *does* account for store size.
That’s why Retailer X can have higher $/MM than Retailer Y even though Retailer Y is a bigger chain. Robin wasn’t saying that Retailer Y is bigger so that’s why it’s velocity is lower. Robin was saying Retailer Y is bigger so that’s why its total sales is higher EVEN THOUGHT its velocity is lower than Retailer X.
Hope that helps to clarify!
Hello – I’m looking at $/MM ACV at the brand level & comparing brands to one another during a 52wk time period at a single customer. Would the # of SKUs in each brand effect the results? Also, if items were only in for a portion of the evaluated time would that skew the results (i.e. new items)?
Thank you,
Yes! The number of SKUs and how long they have been in the market (especially for a 52-week period) would impact the results. You could calculate $ per item per MM ACV to control for number of items. Looking at a shorter period (like 12 or 24/26 weeks) can help if there are new items that have been sold for the the full 52 weeks.
What would cause my ACV go down, but my ACV ($MM) go up significantly?
If you mean distribution (%ACV) is down but velocity (Sales per $MM ACV) is up that is pretty common! When distribution goes down it is often the weaker stores (i.e. where the product is not selling as fast) that have dropped the product. Therefore the overall velocity goes up, since the stores that do still sell the product are the better stores (i.e. velocity is higher). I hope that helps! If I misunderstood the question then please click on the Contact link to clarify and I will get back to you and also update this response.
I am currently trying to determine the opportunity for a retailer by adding a similar item that is selling in the market. I currently have NIQ (Nielsen Scan Data). Is my methodology using $ / $MM ACV accurate:
Item 1 – Rest of Market $ / MM ACV = $50.30
Retailer ACV = $10,000,000,000
Retailer $$ Opportunity of adding Item 1 (selling in the market but not the retailer) = $50.30 x ($10,000,000,000/1,000,000) or $503,370
Is this a correct way to calculate the opportunity In your opinion
If this is not a proper calculation, how better can I do it using NIQ data
Yes, that is right.