There are two ways to put numbers on security risk, and arguments between them are older than most of our tools. One says: score each risk as low, medium or high and colour it on a grid. The other says: estimate the expected loss in rupees per year and compare it with the cost of fixing it. I will weigh both, then say which I would pick for most Indian mid-sized companies, and then connect it to what ISO 27001 actually requires.
The vocabulary, kept short
A asset is something of value: data, a system, a process, a reputation. A threat is something that could harm it, such as ransomware or a careless administrator. A vulnerability is a weakness the threat can use, such as an unpatched server or no backup. Risk is the chance that a threat uses a vulnerability, combined with the damage that would follow. The usual shorthand is likelihood times impact.
Approach one: qualitative scoring
You rate likelihood from 1 to 5 and impact from 1 to 5, multiply, and place each risk on a grid. A risk scoring 20 outranks one scoring 6. It is fast, workshops can do it in an afternoon, and it needs no financial data, which is its great attraction.
Its weaknesses are real. Different people mean different things by “likely”. Multiplying ordinal scores creates false precision: a 5 by 1 and a 1 by 5 both score 5 but are very different situations. Everything tends to drift to “medium”. And you cannot easily answer the question a finance head will ask: is a ₹6 lakh control worth buying?
Approach two: quantitative estimates
Here you estimate, for each scenario, the single loss expectancy (SLE), the cost of one occurrence, and the annualised rate of occurrence (ARO), how many times a year you expect it. Multiply and you get the annualised loss expectancy (ALE).
An invented example. A logistics company estimates one ransomware event would cost ₹40 lakh in downtime, recovery and lost orders. Based on its exposure and industry experience, it thinks such an event could happen about once every four years, so ARO is 0.25. ALE = 40 lakh x 0.25 = ₹10 lakh a year.
Now compare a control. Immutable backups and endpoint detection cost ₹3 lakh a year and, the team judges, cut the likelihood to once in twenty years (ARO 0.05), giving an ALE of 40 lakh x 0.05 = ₹2 lakh. The yearly reduction is ₹8 lakh against a cost of ₹3 lakh, a net benefit of about ₹5 lakh. That is a language the board understands.
The weakness is equally clear: every input is a guess. We have little reliable frequency data for most attacks, and impacts vary wildly. The risk is that a neat spreadsheet makes soft guesses look hard. The sensible practice is to use ranges instead of single values, state assumptions, and treat the output as a way to compare options, not a forecast.
Where I land
Start qualitative to find the shortlist, then go quantitative for the top five or ten risks where a spending decision hangs on it. Pure grids never tell you what to buy; pure spreadsheets across hundreds of risks never get finished. Mixing them takes more discipline, but it gives you speed where speed is needed and rigour where money is on the table.
After scoring: four things you can do with a risk
- Mitigate it: apply controls to reduce likelihood or impact.
- Transfer it: buy cyber insurance or contract the risk to a supplier. Responsibility for your customers’ trust stays with you, though, and insurers expect baseline controls.
- Avoid it: stop the activity, for example not collecting data you do not need.
- Accept it: knowingly live with it, with sign-off from someone who has the authority to accept it, and a review date.
What remains after treatment is residual risk. It is never zero, and a risk register that pretends otherwise has not been taken seriously.
How ISO 27001 uses all this
ISO/IEC 27001 specifies requirements for an information security management system (ISMS), a managed, continually improving set of policies, processes and responsibilities. It does not prescribe one risk method. It requires you to define a risk assessment process with criteria for acceptable risk, apply it consistently, identify risks to confidentiality, integrity and availability, assign risk owners, choose treatment options, and keep records.
Two documents show how risk drives controls. The risk treatment plan records what you will do about each risk. The Statement of Applicability lists every control in Annex A, says whether it applies, and justifies inclusion or exclusion. The 2022 edition of the standard’s Annex A groups 93 controls into four themes: organisational, people, physical and technological. You are not required to implement all of them; you are required to justify your choices by reference to your risks.
The cycle is the point: assess, treat, monitor, review. Management reviews and internal audits feed the next round, and an external certification body checks that your system actually runs, not just that documents exist. A certificate shows the process is in place; it does not guarantee you will never be breached.
One small writing tip for risk registers: phrase each entry as a scenario (“attacker uses stolen vendor credentials to access the billing system”) rather than a vague label (“third-party risk”). Scenarios can be scored, owned and tested. Labels just sit in a spreadsheet.
Figures and names in the examples are illustrative.
Leave a Reply