Sell-out datasets, provided by companies such as Nielsen and IRI or directly by retailers, are a precious source of information for marketers. These datasets cover different markets, such as FMCG, cigarettes and over-the-counter medicine sales.
We are in 2012 and you have just been appointed Sales Director of Sisma, an Italian SME. With brands such as Cotoneve, Samurai, Farmacotone and Logex, Sisma competes with the likes of P&G and J&J in the highly competitive Wipes and Cotton Swabs markets.
You still know little about the market dynamics. Fortunately, you acquired a sell-out IRI dataset (to run the dataset in the app, type "sell-out wipes" in the "choose a sample dataset" widget). You also know the benefits of using variance analysis to understand what is going on.
See how a metric, here revenue, interacts along a pair of dimensions, here category and company.
Different players across the different categories in which Sisma - in orange - plays. Significant Sisma presence in three categories, marginal in the others.
See how a dimension - here category - plays out in the two periods on a set of metrics - here amount and units.
Non-Disposable glove sales fell significantly in terms of units.
The market is fairly fragmented in terms of players, with Sisma not far behind P&G.
Most companies operate with a single brand in a give category...
...while Sisma plays with multiple brands in each category....
...with many SKUs per brand (example below=cotton swabs).
The format dimension is correlated to the category.
The average price is going up while volume is going down.
Price distribution is clustered around the 2 Euro per product mark. Very few items are priced above 6 Euros. The price distribution profile is similar in the two periods.
Sales variance analysis
See aggregated sales variance or split variance between its components (volume, price, mix, drivers,..).
The market has gone up by 2.4M from 203 to 206M...
FIRST RESULT COMBINATION
These five facts explain the market growth in terms of total variance.
1️⃣ The growth of Pampers Baby Fresh Baby Wipes alone is equivalent to the total market growth.
2️⃣ Diva, across its different offerings, grew by another 1.7M.
3️⃣ Excluding the already mentioned Pampers and Diva products, Baby Wipes sales fell by 2.3M.
4️⃣ Manetti&Roberts grew by 1.4M in its categories, mainly Makeup Remover Wipes.
5️⃣ Other players in Makeup Remover wipes, excluding Diva and Manetti, lost 1.3M in sales.
*️⃣ This leaves a positive 0.6M balance tied to other issues.
This set leaves us with an open question: how much of the 2M loss of result row 3 is represented by P&G products not included result row 1? In other words, excluding the successful Baby Fresh line, how important was the loss of P&G in other Baby Wipes products?
There are different ways of getting an answer. The quickest is probably simply to plot the data that makes up our result row 3. The loss is indeed about 2M.
We have our answer if we slice by Company. Excluding the result row 1 gain, P&G is loosing exactly 1M of sales (from 14.7 to 13.7M), tied to a significant drop in volumes.
SISMA - AGGREGATE VARIANCE
Sisma's sales have grown from 19.5 to 20M...
...gaining a bit of share.
SECOND RESULT COMBINATION
These three facts explain Sisma's growth in terms of total variance.
1️⃣ Disposable Gloves sales up by 700k (+27%).
2️⃣ Cotonet 200Ct Cotton Swabs up by 250k (+135%)
3️⃣ All the rest down by 500k (-3%).
*️⃣ This leaves no balance tied to other issues.
PRICE AND VOLUME VARIANCE
Variance can be split into two components: Price - the revenue change due to a change in price, everything else equal - and Volume & Mix - the revenue change due to a change in volumes or in mix, everything else equal.
MARKET - PRICE AND VOLUME VARIANCE
Market growth appears driven by price increases, partially offset by negative volume & mix.
THIRD RESULT COMBINATION
These five facts explain the market growth in terms of price and volume variance. They confirm the importance of Pampers Baby Fresh growth and give insight into the dynamics of the Glove categories.
1️⃣ Non Disposable Gloves sales fell in terms of volume & mix by 2.1M...
2️⃣ ...but recovered 1.9M in terms of price pointing to possible demand elasticity.
3️⃣ Pampers Baby Fresh Wipes alone grew 1.8M just in terms of volume & mix...
4️⃣ ...while the other Baby Wipes fell by 1.9M.
5️⃣ Disposable Gloves grew in price by 1.3M.
*️⃣ This leaves a positive 1.4M balance tied to other issues.
SISMA - PRICE AND VOLUME VARIANCE
Sisma's growth is also driven by price increases.
FORTH RESULT COMBINATION
This set of facts confirms that net revenue increase is due to price increase, with zero net impact of volume change.
1️⃣ All Cotonet products up by 510k in volume & mix.
2️⃣ Non Cotonet Cotton Swabs lost 620k in volume & mix.
3️⃣ Cotoneve products, excepting Cotton Swabs, gained 475k in volume & mix.
4️⃣ Makeup Remover of other brands lost of 477k.
5️⃣ Overall price increases had positive impact of 450k.
*️⃣ This leaves 136k positive balance tied to other issues.
Mix variance is the impact on sales of a change in the internal proportion of the products sold. It is the "average price" component, net of the price change of the single products.
MARKET - MIX VARIANCE
The increase in price has been accompanied by an impoverishment of the mix of products sold. The pure change in volumes has had practically no impact on sales.
FIFTH RESULT COMBINATION
This set of facts draws attention of the negative change of mix in Baby Wipes sales.
1️⃣ The pure volume growth of Pampers Baby Fresh alone is equivalent to the total market growth.
2️⃣ Diva, in terms of pure volume (excluding mix effects), grew 2M.
3️⃣ Volume of other P&G Baby Wipes fell by 2M.
4️⃣ Non-Disposable Gloves grew in price by 1.9M.
5️⃣ All Baby Wipes fell by 1.6M in terms of mix.
*️⃣ This leaves a negative 0.3M balance tied to other issues.
SISMA - MIX VARIANCE
Sisma has a different pattern. Growth is driven by price, with a small enrichment of product mix, offset by a drop in volumes.
SIXTH RESULT COMBINATION
This set of facts draws attention of the positive impact of Logex Professional and of Cotonet to drive sales.
1️⃣ Logex Professional 100Ct Disposable Gloves up 890k by pure volume.
2️⃣ Logex Professional 100Ct Disposable Gloves down 670k by mix.
3️⃣ Other Logex Disposable Gloves up 670k by mix.
4️⃣ Cotonet 200Ct Cotton Swabs up 540k by pure volume.
5️⃣ All other Cotton Swabs down 400k by pure volume.
*️⃣ This leaves a negative 400k negative balance tied to other issues.
NEW AND LOST VOLUME VARIANCE
Has the volume increase to do with new offerings? New offerings are combinations where at least one element has changed. A new product, or an old product offered in a new format.
MARKET - NEW AND LOST VOLUME VARIANCE
Market growth is driven by new offerings. There is no impact of product delisting. Most of the reduction in sales is due to products present in the shelves in both periods.
SISMA - NEW AND LOST VOLUME VARIANCE
Sisma tells a similar story, with volume growth driven by New and by Price.
SEVENTH RESULT COMBINATION
Separating new and lost volume variance returns a set of facts.
This set draws attention of the limited importance of New references to drive Sisma sales .
1️⃣ Logex Professional Disposable Gloves positive pure volume impact of 0.9M.
2️⃣ Pure volume impact of all other existing products negative 1M.
3️⃣ Logex Professional Disposable Gloves negative mix impact of 0.7M
4️⃣ All other Logex positive mix impact of 0.9M.
5️⃣ Cotonet 200Ct Cotton Swabs negative mix impact of 0.3M.
*️⃣ This leaves a positive 700k balance tied to other issues.
A keen trade marketer, you look into Weighted Distribution.
Distribution, the metric commonly used for tracking product availability, is measured in numeric and weighted terms. Numeric Distribution is the percentage of stores handling a product. Weighted Distribution is the percentage of stores handling a product weighted by the product's category store sales.
The power of the Weighted Distribution metric is that it shows how well a brand is distributed relative to where customers buy a particular category of products.
If Weighted Distribution for a given product doubles, one would expect sales to double as well, everything else being equal. If sales do not double, and prices have not changed, then sales per POS must have gone down. The underlying assumption is that if a customer does not find a given product in a shop, she will not change shop in order to find it.
MARKET - WEIGHTED DISTRIBUTION
Market weighted distribution improved. However, this had little impact on sales, since sales per shop fell in a similar proportion.
Considering also the effect of the change in distribution, returns a set of facts.
EIGHTH RESULT COMBINATION
This set draws attention of the increase of distribution of Baby Wipes partly compensated by lower sales by POS .
1️⃣ Pampers Baby Fresh up 2.2M in terms of sales by POS impact.
2️⃣ Other P&G Baby Wipes down -3.1 in terms of sales by POS impact.
3️⃣ All Baby Wipes (including P&G) up 2M due to better distribution.
4️⃣ Non Disposable Gloves down 2.1M due to higher sales by POS.
5️⃣ Non Disposable Gloves up 1.9M due to higher price.
*️⃣ This leaves a positive 1.5M balance tied to other issues.
SISMA - WEIGHTED DISTRIBUTION
Sisma had a similar dynamic, just more accentuated. A significant effort at improving presence in shops did not result in an increase in sales. The sales increase was driven by price and mix.
For comparison, P&G's growth was driven by better distribution. Price increases were offset by an impoverishment of mix.
Considering the effect of the change in distribution returns a set of facts.
NINTH RESULT COMBINATION
This set draws attention of the neutral combined effect of increase in distribution and loss of sales per POS.
1️⃣ 1.3M negative overall impact of sales per POS.
2️⃣ 1.1M positive impact of higher distribution.
3️⃣ 0.7 negative mix impact of Logex Professional Disposable Gloves.
4️⃣ 0.8 positive mix impact of all other Sisma products.
5️⃣ 0.4 positive price impact across all products.
*️⃣ This leaves a positive 0.5M balance tied to other issues.
ANNEX - SLICING BY DIMENSION
In this section the data is analyzed with a standard slicing by dimension approach.
The Format and Ean dimensions have been ignored: (i) hard to interpret a change in format; (ii) Ean too detailed a dimension to give real insights.
The dataset contains six different dimensions.
VARIANCE BY CATEGORY
The datasets covers the different categories in which Sisma operates. The main four categories by sales change are: Baby Wipes, Makeup Remover Wipes, Frost Bags, Disposable Bags. Other categories make up about one third of the market.
MARKET BY CATEGORY - AGGREGATED VARIANCE
The plots are ordered by decreasing amount of variance. In absolute terms, the largest growth comes from Baby Wipes. Disposable Gloves is the most dynamic category in terms of % growth.
Disposable Gloves lead the race in terms of growth, but do not make the "top four" categories list in terms of revenues. Significant volume drop of Non-Disposable Gloves, at stable revenue.
The two year price profiles are similar.
MARKET BY CATEGORY - PRICE AND VOLUME VARIANCE
Splitting variance into Price and Volume & Mix shows that growth of Disposable Gloves (and all other categories) is driven by price, not by volume/mix. Non-Disposable Gloves is interesting because the impact of the increase in price is offset by a decrease in volumes, possibly pointing to demand elasticity.
The capacity of "isolating" effects is a benefit of variance analysis. Standard Price - Volume correlation analysis gives little insight at the total "Non-Disposable Gloves" category level...
...even though this might in fact be due to the influence of the least priced items (0-0.5 Euro) on the regression.
If we filter the category on a specific company, for instance Spontex, try splitting the category, for example, by company, there is still no measurable correlation.
Comparing year over year, the price distribution profile moved clearly to the right.
MARKET BY CATEGORY - MIX VARIANCE
There is a notable mix impoverishment (customers buying a greater proportion of less pricey products) in both Wipes categories. Non-Disposable Gloves came down by pure volume, with unchanged mix.
MARKET BY CATEGORY - NEW AND LOST VARIANCE
New products drove sales in the Baby Wipes and in the Disposable Glove categories. Sales of existing product weakened in the Non-Disposable Gloves category.
MARKET BY CATEGORY - WEIGHTED DISTRIBUTION
Distribution improved for Baby Wipes. Volumes per POS decreased for Non-Disposable Gloves, with flat distribution.
SISMA BY CATEGORY - AGGREGATE VARIANCE
Sisma's growth was fueled by Disposable Gloves, with sales in the other categories essentially flat.
Cotton Swabs and Makeup Remover Cotton represent the majority of Sisma's sales.
Sisma operates with over 20 brands most tiny
Sisma's price profile did not change much between the two years...
...with the highest tickets in the Disposable Gloves category.
Price-volume correlation is very low for all Sisma categories.
DISPOSABLE GLOVES BY COMPANY - AGGREGATED VARIANCE
Among competitors in the Disposable Gloves category, Sisma did extremely well...
...growing both in terms of price and of units sold.
DISPOSABLE GLOVES BY COMPANY - MIX VARIANCE
Sisma's significant growth in Disposable Gloves volumes was only partially offset by an impoverishment of mix.
Logex was Sisma's winning brand in Disposable Gloves...
..with most gains driven by its Professional line.
Disposable Gloves products generally cost around either 1 or 5 Euros.
Sisma's Disposable Gloves are both in the 1 and in the 5 Euro sub-segment.
DISPOSABLE GLOVES BY COMPANY - NEW AND LOST
Sisma Disposable Gloves sales growth was not driven by new product launches. A significant number of new references by minor "other" producers ("Altri Produttori") were listed.
DISPOSABLE GLOVES BY COMPANY - WEIGHTED DISTRIBUTION
Sisma gained both in terms of distribution and of sales per POS. The sales losses of Spontex, the Disposable Gloves category leader, seem tied to some sort of delisting.
SISMA BY CATEGORY - DISTRIBUTION VARIANCE
The disappointing performance of Sisma's largest category, Cotton Swabs, was tied to both an impoverishment of the mix, and to a reduction of sales per store.
VARIANCE BY COMPANY
Next dimension up for analysis is company.
MARKET BY COMPANY - AGGREGATED VARIANCE
Diva, Manetti&Roberts and P&G gained the most sales.
DIVA BY CATEGORY - AGGREGATED VARIANCE
Diva grew substantially in Baby and Makeup Remover Wipes.
MANETTI&ROBERTS BY CATEGORY - AGGREGATED VARIANCE
Manetti grew substantially in Makeup Remover Wipes but did not make progress in Cotton Swabs, where Sisma is a large player.
P&G BY CATEGORY - AGGREGATED VARIANCE
In this market P&G only plays in the Baby Wipes category, where it gained substantial sales.
MARKET BY COMPANY - PRICE AND VOLUME VARIANCE
Spontex, an important competitor of Sisma in the Glove categories, lost significant sales due to volume & mix. P&G's growth was driven by price.
MARKET BY COMPANY - MIX VARIANCE
Significant impoverishment of P&G's product mix, offset by higher volumes.
P&G price profile in this market is centered at 3 and goes all the way to over 10 Euros and more. Widened between the two years.
MARKET BY COMPANY - NEW AND LOST VARIANCE
Growth of P&G is driven by new products, offset by the mentioned impoverishment of mix.
MARKET BY COMPANY - WEIGHTED DISTRIBUTION
Sisma: more distribution offset by lower sales by POS. P&G: more distribution offset by impoverished mix. Spontex: less distribution not offset by higher sales by POS.
VARIANCE BY BRAND
In this market context slicing by brand is not helpful because many companies (Diva=>International, P&G=>Pampers, Manetti&Roberts=>Manetti&Roberts,Vileda=>Vileda,...) operate with a single brand.
SISMA BY BRAND - AGGREGATED VARIANCE
Most of Sisma's growth driven by Logex (Gloves) and Cotonet. Smaller brands loosing sales.
SISMA BY BRAND - PRICE AND VOLUME VARIANCE
Logex growth tied to prices and volume & mix. Cotonet growing on volume & mix with flat prices. Smaller Sisma brands loosing volumes.
SISMA BY BRAND - MIX VARIANCE
Negative variance mix impact on Cotoneve. Other brands: mix improved or remained unchanged.
SISMA BY BRAND - NEW AND LOST VARIANCE
Little impact of new and lost products.
SISMA BY BRAND - WEIGHTED DISTRIBUTION
Cotonet, Cotoneve and Logex sales helped by better distribution. Minor brands are loosing significantly in both distribution and sales per POS.
VARIANCE BY PRODUCT
Four products represent about 40M (20%) of the 200M market.
MARKET BY PRODUCT - AGGREGATED VARIANCE
Significant growth of Pampers Baby Fresh (accompanied by fall of other Pampers products). A Spontex glove product essentially disappeared from market.
MARKET BY PRODUCT - PRICE AND VOLUME VARIANCE
Sales change of the products that have increased or dropped the most is tied to volume & mix variance, with a limited price impact.
MARKET BY PRODUCT - MIX VARIANCE
Generally negative mix variance impact for products with highest sales change.
MARKET BY PRODUCT - NEW AND LOST VARIANCE
Limited impact of "new", possibly "new format", entries at the product detail level.
MARKET BY PRODUCT - WEIGHTED DISTRIBUTION
Pampers Baby Fresh sales driven by more volumes per POS. Opposite true for other Pampers product Sole Luna. Spontex Seconda Pelle almost delisted.
SISMA BY PRODUCT - AGGREGATED VARIANCE
Significant growth of Logex Professional Medium and Large Disposable Gloves.
SISMA BY PRODUCT - MIX VARIANCE
Logex Professional growth tied to both price and volume. Mix not significant at the product level.
SISMA BY PRODUCT - WEIGHTED DISTRIBUTION
Positive distribution impact for Cotonet. For a few other Sisma products, more distribution did not have the hoped-for results.