In Key performance metrics for grocery retail product availability, I outlined the challenges retailers face when choosing key performance indicators. Following that post, several people contacted me to say that metric creation is only the start. We must also consider the targets we set against those metrics.
I wholeheartedly agree.
At the start of every financial year, retailers set their business objectives, goals, and targets for the months ahead. They identify where they would like to be, from a revenue, margin, shrink, or waste perspective (to name but a few). However, they rarely base these expectations on robust science. All too often, they base these forecasts on equations as basic as last year plus or minus one or two per cent! For a sector that once led the way in advanced data analysis and its interpretation, this does not feel best in class.
When you compare it to the relatively immature practice of sports science, we have seen innovative concepts espoused in the likes of Billy Beane’s Moneyball model: accelerate and overtake. Association football and rugby, the latter a professional sport for only the last 25 years, now have clubs that adopt rigorous and robust analytical assessments. This covers everything from game plan strategy and player development to on-field tactics and individual contributions. The aim is to optimize and accurately predict results. Crucially, these assessments are grounded in realistic expectations of performance progression. If you barely escaped relegation last year, then you do not set yourself the target of champions this year!
So, what can retailers learn from this?
Setting the right performance targets is not easy at the category, store, region, and national level. You want your team to buy into the targets you set. You want them to feel they are fair, representative, and genuinely motivated to achieve. But to get this level of engagement, especially when there is a bonus for success, you must apply rigor and discipline.
Several commonly used approaches are summarised below. But if we want to set a new standard, then I propose we adopt a combinatorial modeling approach to the use of data in retail.
|
1 Cascaded Target (e.g. X%)
A deconstructed value allocated unscientifically across business functions/depts etc. Everybody contributes X% or your contribution is Y%. |
Pros · Simple approach (even if embellished as sales-weighted allocation) · Everybody participates to a degree · Theoretically maps up to the bigger picture Cons · Not a fair allocation or realistic expectation · Hard to achieve |
|
2 Last Year Plus
Expectation of a relative or fixed improvement on the previous period (i.e. year) |
Pros · Constant need to improve regardless of current performance/position · Simple approach · Everybody participates Cons
· Not a fair allocation – possibly unachievable · Lacks relevance to the achievable opportunity gap |
|
3 Simple Performance Offset
You underperform compared to format X, Region Y, etc., so must improve by Z |
Pros · Simple approach · Potentially accurate benchmark/comparison Cons
· Lacks fairness, especially if you do not derive your control sets scientifically · Target (and current offset) unrealistic |
|
4 Simple Explanatory Model
Construct a set of explanatory inputs and target predictions, e.g. vanilla Regression model |
Pros · Relatively simple approach · Takes into consideration some of your input values Cons
· It only creates an actual vs expected offset – what happens when Actual>expected? · Potentially arbitrary input scenario adjustments · Unmanaged multi-collinearity so you may find it a challenge to identify driver actions |
|
5 Sophisticated Explanatory Model
A model that attempts to isolate all driver and lever impacts |
Pros · Sophisticated approach – creates a ‘fair’ expectation · Isolates controllable inputs, i.e. how to drive it. Cons · More data and modeling intensive · Difficult to understand |
|
6 Combinatorial Model Approach
Control Set generation (i.e. similarity)
Explanatory model (i.e. isolated drivers and levers)
Lever adjustment scenarios (i.e. weights and fair benchmarking) |
Pros · Greater predictive power. Helps you set a ‘Fair’ value · Connects to the underlying drivers and levers to enable action and performance improvement · Robust benchmarking (i.e. fair comparison) for lever adjustments Cons · Most labor-intensive approach · Requires an ‘out-of-control-set’ scalar to enable universal comparison |
Ultimately, if you have a meaningful target, then everyone across the organization should stand to benefit from its attainment. To enable this, you must provide tools that define and isolate the value of each process and action, or lever, in relation to the target metric.
I have spent the past 10 years building and refining a proprietary model that manages complex and highly correlated systems. This sits alongside a visually representative scorecard of strength, direction, and certainty of each lever input.
Combining lever importance with a reliable store comparison benchmark, using robust and similarity-derived representative control sets, does two things.
At the end of the day, we are all interested in getting the biggest bang for our buck, especially when constrained by limited resources.
The numbers can tell us a lot. But you must also embed expert interpretation before you invest resources. For example, assessing for cause and effect.
Higher nil-pick occasions (items ordered online and then not found by the picker in-store) may correlate with, and be caused by, higher online sales and the number of orders for that store. However, higher nil-pick occasions, despite appearing at the same time, definitely will not cause higher online sales and order numbers.
If you want to know more about how you can set fairer, achievable targets in your organization, please get in touch.