Security testing

I would not begin security testing by buying another platform or launching a large programme. I would find the boundary, the accountable owner, and one scenario worth testing.

1Where I'd start

The first pass should map data flows, authorization and release controls, establish a baseline, assign ownership, and run a bounded test. If that test reveals a problem, fix the most consequential part and repeat it before scaling.

The question I'd ask: What can we verify in the next thirty days?

2Where plans fail

I would challenge the assumptions around data flows, authorization and release controls. Is the information current? Does the control work under pressure? Can the team demonstrate the result without relying on a vendor slide?

The question I'd ask: What assumption would break the plan if it proved false?

3Beyond the jargon

For me, the point is to protect the trust boundaries in software. That requires understanding data flows, authorization and release controls. I would ask what happens when the expected control fails, what evidence survives, and who is empowered to act.

The question I'd ask: What would you need to see before trusting the result?

4What good looks like

I would start with verified exploit-path closure and make sure the definition is reproducible. If the number improves while the underlying exposure stays the same, the measurement is misleading.

The question I'd ask: What outcome would change a budget or operational decision?

5The call leaders can't dodge

I would put three things on the table: the choice, the evidence that supports it, and the consequences of waiting. In security testing, I would examine data flows, authorization and release controls before recommending investment or operational change.

The question I'd ask: Who owns the decision when the evidence is incomplete?

From theory to something useful

With security testing, the work has to connect to a real result: protect the trust boundaries in software. I would agree on scope, identify an owner, test one meaningful scenario, and record what actually happened. The measure I would watch is verified exploit-path closure, supported by enough context to explain why it changed.

I don't assume one answer fits everyone. Your environment, threat exposure, operating constraints, and tolerance for disruption should shape the work. If a recommendation cannot be explained to the people who have to operate it, it probably isn't finished.

Let's talk about your situation

If you're dealing with security testing, tell me what's at stake and what decision you need to make. I'll tell you whether I can help and what a practical next step might look like.

Explore: More on application security · Work with me