It has been an eventful few weeks. From the fridge hacks in France to the FAA flight groundings in the US and the global Microsoft outage, recent events have highlighted how disruption can travel through the suppliers we rely on. These relationships help our businesses operate and grow, but they also raise an important question. How well prepared are we to keep serving our customers when a supplier disruption affects us?

Because you trust your providers to keep things running. Most days they do. The difficulty comes on the day they do not, because your customers will not see a supplier problem. They will see a service that did not work.

You are not expected to control every supplier

Let us start with some reassurance. No business can prevent every supplier disruption. A familiar, widely used provider can still fail, and it can fail in ways that reach well beyond its own platform. What you can do is understand where a disruption would hurt, decide how much of it you can tolerate and prepare a response that has been tested. This is what we call third-party resilience. It is the ability to anticipate, withstand, respond to and recover from supplier disruption while protecting the services your customers rely on.

Many organisations already hold the pieces. They have a supplier register, a risk assessment and a business continuity plan. Each one is useful. The trouble is that they often sit apart from one another. A register tells you who your suppliers are. A dependency map shows where their failure would affect your business. A continuity plan sets out the response. Testing shows whether that response works.

When these four are connected, you move from resilience on paper to resilience in practice.

Consider a plan that says, “use an alternative communication channel”. It sounds sensible. It leaves important questions open. Is that channel available? Can staff log in to it? Does it rely on the same provider that has just failed? Can customers use it safely?

Here is a simple three step plan to answer those questions.

Step 1: Know what you are protecting and what it depends on

Begin with the services that matter most to your customers. These are the services where a disruption could cause real customer harm, operational damage or financial loss. Describe them in business terms, such as processing insurance claims, administering member benefits, fulfilling orders or coordinating deliveries.

This framing matters. Email is a supporting capability. The service it enables is what customers actually depend on. For each critical service, name an accountable owner and define the outcome that must be protected. A business impact assessment can then show how the consequences grow over time. Look especially at peak periods and at customers who are more vulnerable to delay. If you are new to this, start with a small number of services and expand as your approach matures.

Next, map what sits behind each service. Connect it to the processes, people, systems, data, integrations, locations and third parties it needs. Where you can, look beyond the immediate supplier. Two providers that look separate may rely on the same cloud platform, telecommunications network or subcontractor.

For each important dependency, ask three things. What would stop or slow down if it became unavailable? What alternatives exist? How long could those alternatives last? Record what you do not know as well. An unverified assumption about a supplier’s recovery capability should stay visible as a gap to investigate.

Step 2: Decide how much disruption is tolerable, then plan for it

An impact tolerance is the most disruption a critical service can withstand before the consequences become unacceptable. Time is central, but it is not the only measure. You may also look at the number of customers affected, outstanding cases, missed deadlines or the volume of unprocessed work.

A tolerance is different from a recovery time objective, which is a recovery target. Your target should leave enough room to activate workarounds, restore operations and work through accumulated backlog before the service breaches its tolerance. The UK Financial Conduct Authority draws this same distinction in its operational resilience guidance.

Here is an illustrative example. A claims service might need to receive and triage urgent claims. Management could set a tolerance that urgent cases receive an initial assessment within four hours of receipt, even during a disruption. They could then set a response target of activating an approved alternative intake channel within 30 minutes.

These figures are only examples and would need to be set based on your own evidence about customer needs, service obligations and operational consequences.

With tolerances in place, build continuity plans around each dependency. A good plan names the activation triggers, the person with decision authority, the responsibilities, the supplier escalation contacts and the customer communication arrangements. It also describes the minimum service level you intend to maintain, and the resources needed to do it.

Once done, check that each workaround is genuinely usable. A manual process needs enough people, accessible information, sound controls and a way to reconcile records afterwards. An alternative provider needs real capacity and a realistic activation time. Make sure authorised staff can reach the plan and the contact details during the very disruption you are planning for. Finally, decide how you will clear backlogs and return safely to normal.

Step 3: Test it, fix it and prove it

A plan is a promise until it has been tested. Choose a severe but plausible scenario and test it against the tolerance you agreed.

A Microsoft-style communication outage makes a useful starting point. Imagine email is unavailable for a full working day, recovery timing is uncertain and customer requests keep arriving. Work out which other capabilities remain available instead of assuming everything fails. Bring business owners, operations, IT, risk and the relevant suppliers into a tabletop exercise. Then validate your key assumptions with controlled practical tests, such as activating the alternative communication channel or running sample cases through a manual workflow.

Measure what happens. Look at activation time, processing capacity, customer delays, backlog growth and recovery accuracy. Note which results were observed and which were only estimated. A discussion exercise can reveal gaps, but it cannot prove that a workaround will cope with live demand.

Every finding should become an action with an owner. Link it to the affected service, dependency, plan and risk. Set a due date and agree what evidence will show the work is done. If an alternative process lacks capacity, update the arrangements and test again.

Sometimes a service cannot stay within tolerance. When that happens, tell management plainly. Record the interim measures, the funding or prioritisation decisions and the date of the next validation. It can be tempting to loosen the tolerance to match a weak recovery capability. That makes the report look better, but the harm to customers remains.

Give your board and regulators evidence they can assess

Strong resilience work earns its keep when leaders ask, “How do we know?” A useful report explains which critical services were tested, which scenarios were used, whether tolerances were met and where confidence is still limited.

Back it up with current dependency maps, approved tolerances and their rationale, versioned plans, exercise records, measured results, findings, action owners and retest outcomes. Show the decisions the board needs to make, including unresolved exposures and investment needs.

The evidence and reporting should align with the requirements that apply to your sector and jurisdiction. The FCA’s published observations is a helpful international reference.

Where to start this week

The next supplier outage is not a question of if. You do not have to fix everything at once, and you do not need to begin with a large program. You need one good starting point.

Choose one critical service. Map what it depends on. Agree how much disruption is tolerable. Connect the continuity plan, test it and close the gaps you find. Each service you complete makes the next one easier. When you do this, you can answer the question that customers, boards and regulators ultimately care about. Can the business keep delivering the services people rely on when a third party fails?

With a tested plan, the answer is a confident yes.

If you would like a hand choosing that first service, we would be glad to talk it through with you.