10 ways to reduce inventory and improve service – part 1
This was prompted by a question on the CILT’s eDiscussion forum. I thought the topic deserved a little more room for explanation, so here are my top ten tactics for simultaneous inventory reduction and service improvement. I have divided this into two posts – five tactics today, the next five coming up in part 2.
Perhaps the question should have been phrased in terms of reducing supply chain cost rather than just reducing inventory, but as typical total holding costs for inventory are about 30-35% of valued stock per annum, it’s not a bad start.
1. Reduce lead times
This will allow a downward adjustment of safety stocks, and an improvement in availability. If supplier lead times can be reduced to below the required lead time to the customer, this will remove the requirement to hold stock altogether. Halving lead time should reduce safety stocks by about 30% for the same availability.
For retail supply chains with long lead times and short season cycles, buying decisions are often limited to an initial purchase generated “in the dark” (al buio as they say at Gucci) – before any customers have bought the products. Reducing lead times so that there is time for a further top-up order early in the season when real demand has been measured can have a dramatic effect.
Remember that not all lead time is due to the supplier: it also includes the time to raise an order, the time to book stock in, and half the order review cycle. If stocks are checked on a weekly cycle for reorder, that’s 2½ working days of lead time – check daily and that comes down to ½ a day.
Other candidates for reduction are transport times and associated customs delay – resourcing to a less remote but higher unit cost supplier may lead to lower total supply chain costs, certainly lower inventories. Working closely with a supplier can also remove administrative delays at their end. Because much manufacturing lead time is accounted for by order queuing effects, choosing a production site that operates with more spare capacity will also help.
2. Improve reliability of supply
Unreliable supply is one of the reasons for holding safety stocks: if delivery is guaranteed on the due date then safety stock can be reduced to that needed to cover common-cause variability of demand.
A good start is sharing demand, forecast and stock positions with suppliers – transparency of information throughout the supply chain improves reliability and mitigates Bullwhip/Forrester effects (as anyone who has played the Beer Game can testify). This need not be terribly high tech – I have done this very successfully with a weekly dump from legacy mainframe into preformatted spreadsheets that were then emailed to suppliers. With regular telephone reviews and meetings this can have a powerful effect in reducing supply chain failures.
Another way of helping the supplier deliver more reliably is to place a smoother order schedule. A forecast method that doesn’t react wildly to short-term changes in demand helps, as does a sensible order quantity calculation.
3. Order more frequently
Order little and often: this reduces the cycle stock. Ordering twice as often will halve the cycle stock. There is a cost: each order comes with an administration overhead and a labour cost to receive the goods. The former can be mitigated with automation and management by exception.
The other benefit of this is to reduce the Bullwhip/Forrester effect – a phenomenon in which small variations in demand get amplified up the supply chain. This is a frequent cause of excess stock in supply chains. In this case it helps the supplier – his stock levels should be lower (and should be able to offer a more reliable and lower cost service as a result). This more than compensates for the smoothing effect that larger orders have where demand is stable.
4. Expedite more effectively
Solving supply chain failure quickly – or, better, resolving a potential failure before it happens – will improve service. I used to work with an automotive spares business that achieved about 96% availability from stock to its dealer network. It had clever forecasting and stock policies, but it also had a slick event management system that highlighted potential stock-outs.
Strong relationships with suppliers meant that the expediting process alone was able to contribute 1% point to availability. Without that, availability would have been at 95% – or put another way, good expediting reduced service failure by 20%.
Another lesson I learnt there was the power of passing good information to the customer. So where there was a shortage, the expediters produced good quality ETAs (Estimated Time of Arrival) that were not only used in planning, but also communicated directly to customers. Even if this doesn’t prevent a service failure, it can sometimes alter the customers’ perceptions of that failure.
5. Forecast more accurately
I have written about this before in my post Best forecasting method for your supply chain. Better forecasting means lower safety stocks and/or higher levels of availability. It also means a reduced exposure to excess and obsolete stock risk (a large contributor to the holding cost of stock in most businesses).
- Simple methods often work best (because tuning complicated forecast models to work well over large ranges is problematic).
- Use robust statistics by capping exceptional demands before using the demand history to forecast.
- Use demand rather than sales to forecast where possible.
- Manage new product introduction and end of life carefully.