RSKRisk, cost & sustainabilityStop 5 of 9
Resilience: time to recover & survive
How long the network can keep running through a disruption, how fast it recovers, and how to fix the places where those don't match.
Read atChange any time
You can’t predict which supplier will catch fire, which port will flood or which road will close. You can measure how exposed you are to each one. Two simple numbers do it:
- Time to recover (TTR): how long a site, such as a supplier plant, a warehouse or a route, would take to get back to normal after a disruption, or how long it would take to qualify a replacement.
- Time to survive (TTS): how long your supply chain could keep meeting demand without that site, using stock, alternative sources and spare capacity elsewhere.
If TTR is longer than TTS, you have a gap: the business would start losing sales before the site recovers. Say a specialist resin supplier would take 8 weeks to recover from a fire, but you hold 3 weeks of stock and have no qualified alternative. You have a 5-week exposure on every product using that resin.
The power of this approach, popularized by David Simchi-Levi and colleagues, is that it doesn’t need you to estimate the probability of rare events. It works from things you can measure: stock, capacity and recovery times. It often finds hidden risks in cheap, unglamorous parts rather than the big, obvious suppliers.
The fixes then trade cost for safety:
- Redundancy: more safety stock, spare capacity or a second qualified supplier.
- Flexibility: plants and lines that can make several products, multi-skilled teams, flexible contracts.
- Visibility and speed: early warnings and pre-agreed playbooks shorten recovery.
Words to use
- Time to recover (TTR)
- How long a disrupted node takes to return to full function.
- Time to survive (TTS)
- How long the network can meet demand without that node.
- Redundancy vs flexibility
- Spare stock or capacity held in reserve, versus resources that can switch to where they are needed.
Time to survive with a partial backup
The supplier needs 8 weeks to recover. You hold 3 weeks of stock.
- No alternative: stock runs out after 3 weeks, then nothing for 5 weeks. Lost output: 5 weeks of demand.
- A pre-qualified second supplier who can deliver 50% of your needs from week 2: for the first 2 weeks stock falls from 3 weeks’ cover to 1. From week 2, the backup covers half of demand, so the last week of stock lasts 2 more weeks. Time to survive becomes 4 weeks, and from week 4 to week 8 you run at 50%. Lost output: 2 weeks of demand instead of 5.
In practice, time to survive is calculated with a network model: remove a node, let the model reroute supply and use stock wherever it can, and find how long demand can still be met. Repeating this for every critical node gives the exposure map.
Rank by financial exposure
For each node, multiply the shortfall (weeks of lost output after time to survive runs out) by the profit those weeks would earn. This risk exposure figure ranks nodes without needing the probability of a fire or flood, and it often puts small, cheap suppliers at the top: a ₹5 lakh-a-year component can hold up ₹50 crore of sales.
Three families of fixes
| Fix | Examples | Best when |
|---|---|---|
| Redundancy | Safety stock, second supplier, spare capacity | Recovery is slow and the item is critical |
| Flexibility | Multi-product lines, flexible contracts, cross-trained staff, alternative routes | Many products or nodes share the risk |
| Speed of response | Early warning, pre-agreed playbooks, supplier recovery support | Disruptions are detectable early |
Each costs money in normal times, so the target is the biggest exposures first, with the cheapest fix that closes the gap.
Redundancy or flexibility?
Sheffi (2005) distinguishes two ways to build resilience. Redundancy (extra stock, spare capacity, more suppliers) is simple but costs money continuously and protects only against the disruptions it was sized for. Flexibility (interchangeable parts, multi-skilled people, plants that can make several products, contracts that allow rapid changes) helps against many kinds of disruption at once and often improves day-to-day performance too. Resilient companies combine them, with flexibility doing most of the work.
Disruptions cost more than lost sales
Hendricks and Singhal (2005) studied hundreds of companies that announced supply chain disruptions and found that they underperformed their peers substantially in shareholder returns over the following years, with higher volatility. Disruptions damage customers’ trust, raise costs and distract management, effects that last much longer than the disruption itself. That is the business case for paying for resilience before anything goes wrong.
Stress tests
Borrowing from banking, companies run their network models through severe but plausible shocks: a key port closed for a month, the largest supplier lost, a region locked down. A reverse stress test asks the opposite question: what would it take to stop us delivering to our top customers? The answers become the design brief for resilience investments.
Standards and continuity planning
Risk-management and business-continuity standards, such as ISO 31000 for risk management and ISO 22301 for business continuity management, give structure: identify critical activities, assess risks, set recovery objectives, write continuity plans, test them and improve. Customers increasingly ask suppliers for evidence of such plans.
The price of resilience
Lean, just-in-time supply chains worked well in stable conditions and were exposed by the pandemic and the chip shortage. Swinging to “just in case” everywhere is expensive too. The balanced answer is targeted: extra protection where time to recover is long and the stakes are high, and lean operation elsewhere.
Interview kit
Practise it out loud
Answer each question yourself before opening the model answer.
I’d assess resilience by comparing, for each critical supplier, plant, warehouse and route, the time to recover after a disruption with the time to survive on existing stock and alternatives. Where TTR exceeds TTS, revenue is exposed. It doesn’t require estimating rare-event probabilities and often reveals risks in low-spend parts. Mitigations combine redundancy, like stock, spare capacity and second sources, with flexibility and faster detection, prioritized by revenue at risk.
A supplier would take 8 weeks to recover from a fire. You hold 3 weeks of stock. What do you do?
That’s a 5-week gap where sales are at risk. Options: increase stock of that part, qualify a second supplier or a different material, secure backup capacity, or redesign the product. I’d choose based on the revenue at risk and the cost of each option.
Why can a cheap part be a bigger resilience risk than an expensive one?
Because risk depends on what stops when it’s missing, not on its price. A ₹5 sensor from a single source can halt a whole production line, while an expensive part with several suppliers and plenty of stock poses little risk. TTR and TTS analysis captures this; spend-based analysis doesn’t.
Connected stops
Where this idea feeds in, or gets its inputs from.
Sources and further reading
- Simchi-Levi, D., Schmidt, W. & Wei, Y. (2014). From superstorms to factory fires: Managing unpredictable supply-chain disruptions. Harvard Business Review, 92(1/2), 96–101.
- Sheffi, Y. (2005). The Resilient Enterprise: Overcoming Vulnerability for Competitive Advantage. MIT Press.
- Hendricks, K. B., & Singhal, V. R. (2005). An empirical analysis of the effect of supply chain disruptions on long-run stock price performance and equity risk of the firm. Production and Operations Management, 14(1), 35–52.