October 18, 2017
Improve your bottom line through better distribution management and restocking.
In our previous article, we looked at measuring and optimizing the impact of prices and promotions. Today we are looking at another powerful tool in the Evo Pricing arsenal – optimizing replenishment.
Warehousing costs money.
Unsold items cost money.
Stocking out of popular items costs money.
So, when you implement a seamless system that puts the right products in the right place at the right time, you reduce costs and improve profitability.
The distribution management and restocking headache
Distribution management and restocking are major issues for all retailers. However, they take on particular importance in sectors like fashion, where products are mid to high value and seasonal.
You are trying to predict trends, and even weather, sometimes a whole year in advance!
With that level of advance investment and uncertainty, once you have your new season in stock, it’s vital to keep your finger firmly on the pulse of every store.
And that is the beauty of Evo Pricing. It suggests stock allocations at the most granular level, while at the same time giving local managers the freedom to work their own magic. Together, human and data system draw their conclusions from a wide variety of sources: we call it “the new man-machine alliance”.
One concrete example of this holistic approach? Think of geographical location and climate. A Turin store, based in the cooler north of Italy, would still be selling a lightweight down jacket in March. But further south, in the warmer climates of places like Bari, a store wouldn’t want the jacket replenished – they’d rather release it to other stores!
Right product, right place, right time.
Or, in other words, the question we continually ask ourselves, “What’s the right number of pieces of a particular size of a particular item to send to every store in a given week?”
The answer to this is a systematic weekly process.
The Evo Pricing Weekly Process
On Sunday night, our automated data feed looks at sales and inventory levels and makes an initial proposal. On Monday morning, individual store managers add their insights. Then, in the afternoon, the replenishment process is set in motion. Finally, on Tuesday, shipments from the warehouse and trans-shipments across stores actually happen.
The process itself is very simple and user-friendly.
But it’s based on extremely complex AI algorithms which are constantly being updated. With every passing week, the system learns from the continuous sales feedback, incrementally improving its forecasts. Like a survey of customers and stores, but automated.
Underlying the whole process is a two-part rationale: store budget and item requests.
The store budget
The budget for each store depends directly on the value of the current stock versus the potential sales of the next 4 weeks.
Stores can increase their budget by releasing inventory – thus solving the traditional unwillingness on part of stores to get rid of non-performing stock (it is still a potential sale for them, albeit with low probability). But now, every item released increases the budget of a value equal to its selling price!
Conversely, every item ordered spends the budget for an amount equal to its price. So the stores can request any items they want, modifying the system’s proposal, until they spend their entire available budget.
Prof. Sunil Gupta, of Harvard Business School, commented that our store budget process “is very clever” and that “he likes our innovation”!
The item requests
For each SKU (e.g. item, color and size), there are 3 types of requests:
- Order: require one or more pieces of the item
- Release: make the item in stock available for other stores; the items released will be taken only if required by other stores
- Mandatory release: impose the forced withdrawal of an item with low potential but needed by another store.
The value of the ordered pieces is subtracted from the budget. Those that are released or mandatorily released increase the budget.
An added bonus – finding new popular products
What all this adds up to is information about the popularity of items in real time – an internal survey, if you will, that can directly feed into merchandising and planning decisions.
Plus, because of the greater flexibility and the convenience provided by the deep granularity of the data, stores can now trial and discover new customer trends, and therefore demand which would otherwise not be served. One of our clients called this “the women’s touch in fashion”!
Smart replenishment, the modern way.
About the author
Martin Luxton is a writer and content strategist who specializes in explaining how technology affects business and everyday life.
Big Data and Predictive Analytics are here to stay and we have only just begun tapping into their enormous potential.