How to Do a Business Impact Analysis (BIA): Questionnaires, Interviews and Outputs

How a business impact analysis is run in practice: scope, interview questions, a time-based impact scale, a dependency table, BIA form fields and the most common mistakes.
Technical summary
Technology
ISO 22301, ISO/TS 22317, BIA
Topic
Running a business impact analysis
Level
Intermediate
Prerequisites
  • An approved business continuity scope
  • Current process list and process owners
In this article8 sections

In a company where a disaster recovery plan has been written but it isn’t clear which system comes up first, the first hour of an outage is spent arguing over the order. The cause is almost always a business impact analysis (BIA) that’s incomplete or was never done. A BIA measures which process an outage hits, how quickly and how hard; the recovery order, RTOs and backup investments come from this data. Below we explain the BIA in the order you work in practice, from drawing the scope to management approval.

The purpose of the BIA and how it differs from risk analysis

Risk analysis asks, “What could happen, and how likely is it?” The BIA takes likelihood off the table and looks at a single question: “If this process stops, what do we lose as the hours go by?” Clause 8.2.2 of ISO 22301:2019 expects the following from the BIA: defining impact types and criteria, identifying the activities that support products and services, determining the time frame within which the impact of not performing an activity would become unacceptable (also known as MTPD), setting resumption times for prioritized activities (RTO), and identifying the resources and dependencies needed.

The two pieces of work feed each other; one doesn’t replace the other. Information security risk analysis within an ISMS runs along an asset and threat axis; for the method and sample records, see the ISMS risk analysis guide. The BIA runs along a process and time axis. A mistake you may come across is adding a “likelihood” column to the BIA form. The moment that column goes in, low-likelihood but devastating outages, such as losing the building or the ERP database being completely corrupted, drop to the bottom of the scoring table. And the plan skips them.

Scope and process list

The BIA is done within the same boundaries as the business continuity management system’s scope. Even if the scope will be kept narrow, start the first round with the list of products and services the organization delivers externally; prioritization works backwards from the service. The order is:

  1. Products and services: Outputs delivered to customers, regulators or group companies (for example order fulfillment, shipping, payroll, technical support).
  2. Processes: The processes that produce each service from start to finish. A department name on the org chart isn’t a process; instead of “Accounting”, write “supplier payments”, “e-invoice issuing”.
  3. Activities: Steps within a process that have separate resource and time needs. In most organizations, the BIA is done at this level; going down to task level makes data collection unmanageable.

Confirm the process list with process owners before the BIA starts. Each row gets an owner, a backup person and the locations where the process runs. If location information is left blank, you’ll have to ask the same question again later in the site loss scenario and when deciding on an alternative workplace.

Data collection methods

There are three methods, and most projects use them together. The choice depends on the size of the organization, how familiar process owners are with continuity concepts and how much time can be set aside.

Method Strength Limitation When it fits
Questionnaire (form) Fast, consistent data across many processes Concepts are misunderstood, answers are optimistic Multi-site, mature organizations; preliminary data
One-to-one interview Depth, hidden dependencies come to light Time-consuming, depends on the interviewer Critical processes, first BIA round
Workshop Agreement on dependencies between processes Dominant participant effect, hard to schedule Chains of interdependent processes

The order that works in practice: first, a short pre-form goes to process owners (process description, systems used, headcount, seasonal peaks). Then, in a 60–90 minute interview, the form is completed together and the impact assessment is done. Finally, the owners of interdependent processes are sat side by side in a workshop. If sales says “orders must be processed within four hours” while the warehouse can only ship the same order the next day, that contradiction only surfaces at that table.

The format of the questions determines the result. The answer to “How critical is this process?” is always “very critical”. Instead, ask concrete, time-based questions: “The system stopped at 09:00 in the morning; what happens by noon? By the next morning? After three days?” “Could you run this work by hand for a week, and with how many people?” “Is it worse if it stops at the end of the month or the middle?” If the answer doesn’t mention a concrete loss, ask the question again.

Time-based impact scale

Have management approve the impact scale before the BIA starts; otherwise each process owner interprets “high impact” in their own way and the results can’t be compared. The scale has two axes: impact categories (financial, customer and contractual, legal and regulatory, reputation, health and safety, operational) and severity levels. Each level is defined with measurable thresholds specific to the organization. The table below is a sample template; amounts and thresholds are set according to the organization’s scale.

Level Financial (sample threshold) Customer / contract Legal / regulatory
1 – Negligible A very small share of monthly revenue The customer doesn’t notice No obligation
2 – Limited Can be absorbed within budget Delay for a single customer Internal record required
3 – Serious Affects the quarterly target Contractual penalty, complaint A notification obligation may arise
4 – Critical Affects the annual target Risk of losing a key customer Risk of administrative sanctions
5 – Catastrophic Threatens the company’s survival Permanent loss of the customer base Risk to the operating license

Then the impact level is marked for each activity across time bands. Time bands are chosen to suit the nature of the work; a typical set is 4 hours, 1 day, 3 days, 1 week, 2 weeks and 1 month. Each cell gets the highest category level in that band. The first time band in which the impact crosses the “unacceptable” line (for example, level 4) gives that activity’s MTPD. The RTO is set below this, leaving room for recovery and verification. How targets are reconciled between the business and IT is a separate topic, covered in the guide setting and approving MTPD, RTO and RPO targets. Turn the time and data loss targets accepted by the business units into technical requirements for choosing technology based on RPO and RTO targets.

Identifying dependencies: people, systems, suppliers, facilities

The real value of a BIA is often not in the impact score but in the dependency list. What stops you during recovery is usually a forgotten dependency: a license server, a banking screen only one person knows, a printer configured on only one machine. For each prioritized activity, question four types of resources separately:

  • People and skills: How many people, with what skills and authority, are needed to run the activity at a minimum level? Knowledge that depends on one person (a banking screen only one person knows, a signature only one person is authorized to give) is flagged separately.
  • Information systems and data: Applications such as the ERP, email, file server and bank integration, and acceptable data loss. Ask the process owner, “What’s the oldest data you could work with?”; this answer is the source of the RPO request.
  • Suppliers and external services: Cloud services, logistics company, internet service provider, maintenance company. How to question critical suppliers’ continuity capacity is explained separately in the guide business continuity at critical suppliers.
  • Facilities and equipment: Resources that can’t be moved or are hard to replace, such as the building, production line, warehouse, special printer, safe or signing device.

When putting dependencies into a table, don’t lose the direction: “Process A provides input to B” isn’t the same as “Process A receives input from B”. When direction is recorded correctly, inconsistencies reveal themselves; if an upstream process has a longer RTO than a downstream one, the downstream target stays on paper. This is the first place to look when validating the BIA. At manufacturing sites, OT systems go on this list too; backups of PLC programs and engineering workstations are often skipped. For this special case, see the article IT/OT business continuity in manufacturing.

Sample BIA form fields

The form should be simple enough to fill in together during the interview. A forty-field Excel file sent to a process owner comes back either blank or filled with copy-and-paste answers. The fields below are a sufficient start for most organizations:

Section Fields
Definition Process/activity name, owner, backup owner, location, related product/service
Working pattern Working hours, seasonal peaks, normal headcount
Impact Level by category for each time band, with a justification note
Target request Proposed MTPD, RTO, acceptable data loss, minimum service level
Resources Minimum staff, applications, data, equipment, facilities, suppliers
Workarounds Ability to work manually, for how long, time to clear the backlog
Dependencies Processes it receives input from and provides output to
Approval Process owner approval, date, next review date

The “Workarounds” field is the one most often skipped, but it’s decisive in choosing a strategy. If a process can run by hand for three days, its system RTO can be longer. On the other hand, the work that builds up during manual operation must also be recorded: the pile of paper to be entered when the system comes back effectively lengthens recovery time. An RTO that doesn’t take this into account is misleading.

Validating and approving the results

The data collected doesn’t go to management in raw form. A consistency check comes first:

  • Are the RTOs of processes that feed each other in logical order?
  • Are the acceptable data loss levels requested for processes depending on the same application very different from each other? (As the application will be protected with a single RPO, the strictest request decides.)
  • Did every process say “4 hours”? This usually shows the scale was misunderstood.
  • Have single-person, single-supplier and single-location dependencies been flagged?

Then the results go to management as a one-page summary: prioritized activities and proposed MTPD/RTO values, each activity’s critical dependencies, the first obvious gaps against current technical capacity and decisions awaiting approval. Management’s approval is recorded; this record is both the basis for choosing a strategy and the evidence of the BIA process in audits. ISO 22301 expects the BIA to be reviewed at planned intervals and when significant changes occur. In practice, this means a full round once a year plus interim updates for new products, systems or organizational changes. A BIA that isn’t updated is as misleading as an old network diagram.

In teams running a BIA for the first time, an external facilitator can save time on interview technique and scale design; the BIA and business continuity plan service explains the scope of this work. For the BIA’s place within the management system, see the ISO 22301 consulting page. For the standard itself, the ISO 22301:2019 page is the reference; BIA-specific guidance is in the ISO/TS 22317 technical specification.

Frequently asked questions

Who should do the BIA: IT or the business units?

The owner of the BIA is the business continuity manager, while the source of the data is the process owners. IT is at the table to confirm system dependencies and show current recovery capacity. A BIA done by IT alone measures system criticality, not business impact, and ranks priorities wrongly.

How often should the BIA be updated?

A full review once a year is common practice. When there’s a new product or service, a major system change, a merger, a change of location or a significant outage, the BIA records of the affected processes are updated in between.

Does the BIA need to calculate financial impact to an exact figure?

No. The aim is the right ranking, not accounting precision. Threshold ranges approved in advance with management are enough to compare processes. Exact figures are calculated separately only for gaps that require an investment decision.

How can a BIA be simplified in a small company?

If there are few processes, the form can be filled in during a single workshop with all process owners. Time bands can be reduced, and impact categories cut to three or four. What mustn’t be skipped: an approved scale, a list of prioritized activities and recorded dependencies.

Support

Contact us for support or a discovery session on this topic.

Contact form
[email protected]

Eretek Compliance

We work with you on compliance with ISO 27001, IATF 16949, TISAX and similar standards, covering risk assessment, control design, documentation and audit preparation.

Knowledge Center

Other articles