SKU cannot negotiate with gravityThursday 11 April 2019
“The odds catch up. Probability is like gravity: you cannot negotiate with gravity.”
Det. James 'Sonny' Crockett - Miami Vice
I was in Miami last week, sat for a couple of days with some of the US’s top retail analytics leaders. As you all know, the real meeting begins in the bar afterwards, and it was fascinating to discuss some of the universal Rules that underpin how to optimise for SKU mix, Pricing, and indeed Space. These Rules are like gravity, and you can’t negotiate with gravity. In fact SKU cannot negotiate with gravity!
I was with our analytics tool development partner Ugam. Working with Ugam on the analytics tools has been great, because in the process of coding the tools, we’ve revealed the underlying relationships between SKU share and price elasticity, for example higher category-share SKUs have lower price elasticities.
Working in grocery retail, OC&C and Ugam have developed 10 universal rules that govern how grocery SKUs interact. It was great to road-test these in Miami, and have a smart audience to interact with and test the maths. Anyway, here are the 10 universal truths of grocery retail:
- Higher category share SKUs have lower price elasticity. This is a (near) universal immutable law
- Higher category share SKUs have higher SKU substitution, I.e.: although they have higher Sales, a lower proportion of their volume is incremental to the category. (You will note that this sets up an interesting optimisation opportunity, that you can easily increase the price of your best-selling SKUs, and the small lost volume flows into your other SKUs, not is lost from the category. I.e.: subject to some maths about the Margins, this is generally great news)
- Higher price elasticity is correlated with lower substitution. The more marginal SKUs at the edge of the range are more price sensitive but actually bring incremental volume to the category
- The exception to this rule being, high price SKUs display high price elasticity, which is correlated with high substitution. (Because more volume sits at the lower price point, the former rule is more common)
- Lower price SKUs tend to have higher category shares (not a giant surprise!)
- Lower price SKUs have lower price elasticity and higher substitution (and this is partly because of Rule #1 Higher category share SKUs have lower price elasticity)
- Categories with stretched price ladders have low category price elasticity, and low average SKU price elasticity (conversely, a compressed price ladder leads to high average SKU price elasticity). Often retailers get obsessed about their price relative to competitors... but typically 90% of the value to be realised from price optimisation comes the balancing relativity of SKU pricing and optimising the price ladder (and just 10% from looking at competitors) within the Retailer not worrying about competitors
- Categories with lots of SKUs (and so lots of low share SKUs) tend to have high price elasticity and low substitution. However this might be an artefact of our sample data sets being from doing a good job on Ranging, and that in adding a tail of SKUs, these have been well selected SKUs and bring incremental volume
- There is generally an opportunity to create a cash margin positive impact by de-ranging the worst 10% of SKUs in any category (worst meaning low margin and high cannibalisation). Often, taking out 20% of SKUs will still be cash margin positive. Time and time again we see Total category Margin being killed by low Margin SKUs cannibalising high Margin SKUs. Although we like to optimise SKU mix algebraically, a simple scorecard can spot the worst offenders, columns for example: Sales volume, Sales value, % Margin, % Incrementality (i.e.: 1 minus Substitution). You should kill the Margin killers... Skewer the scud SKUs!
- Categories with a wide spread of margin have the greater SKU rationalisation opportunities, as there tends to be more opportunity to replace demand for low margin SKUs with high margin SKUs
Pricing strategy and SKU selection are the most important growth levers for a category manager and apart these 10 rules are helpful while making decisions around SKU rationalisation, promotional strategy etc. Of course every Retailer and every Category is unique, but these types of data driven rules are useful to get your head in the game at the outset.
As it happens, as the conference night wore on, our attention turned to Space (allocation). If Price and SKU mix are really just 2 different expressions of the same problem (say like Energy and Mass), and Space optimisation and SKU mix are inextricably linked, then of course Price is not independent of Space. These were smart people and pretty soon we’d worked out how price elasticities change depending on SKU count and Space allocation to the category. Smart people get this kind of thing, and it’s fun to hang out with them.
But of course, sure we’re know-it-alls, so I have to acknowledge my favourite Crockett quote...
“Knowing the answers doesn't make it easier, does it?”