How to Calculate NPS Across Locations, B2B Accounts and Customer Segments

Calculating NPS for one customer group is simple. The challenge begins when an organisation has hundreds of locations, several B2B stakeholders inside the same account and different customer segments.

The formula does not change. The reporting logic does.

This guide is for CX teams managing complex NPS programmes. It covers how to calculate NPS briefly, then focuses on the enterprise decisions that can distort reporting: response eligibility, rollups, small samples, account structures and inconsistent programme definitions.

How Is NPS Calculated?

Net Promoter Score is calculated by subtracting the percentage of Detractors from the percentage of Promoters. The formula remains exactly the same whether you have 50 responses or feedback from hundreds of locations.

NPS = % Promoters − % Detractors

Responses are grouped as:

• Promoters: 9–10
• Passives: 7–8
• Detractors: 0–6

Passives are not added or subtracted directly, but they remain part of the valid response base.

For example, imagine 250 valid responses include:

• 125 Promoters
• 75 Passives
• 50 Detractors

Promoters represent 50% of responses and Detractors represent 20%.

NPS = 50% − 20% = +30

That is the calculation.

Resonate CX already has a dedicated guide that goes deeper into the standard calculation, sample size and broader NPS methodology. Guide to calculating an NPS score

The enterprise challenge begins after the arithmetic: which responses should be combined, and what should the resulting score actually represent?

Define the Valid Response Base First

In enterprise reporting, the denominator can create more problems than the formula. Before comparing locations, accounts or segments, decide exactly what counts as a valid response.

Set consistent rules for:

• Incomplete, test or employee responses
• Duplicate submissions
• Customers responding more than once in a period
• Former, trial or inactive customers
• Survey triggers included in the programme
• Customers who change location, segment or account

If eligibility rules change, the score can move even when customer sentiment has not.

That is why methodology changes should be documented. When an executive sees NPS move six points, the team should be able to establish whether customers changed, the response mix changed or the calculation population changed.

How to Calculate NPS Across Multiple Locations

For customer-level enterprise NPS, combine the underlying Promoter, Passive and Detractor counts from every location in scope, then calculate the score from that combined response base.

Do not simply average location scores.

Imagine Location A has 20 responses and NPS +70, while Location B has 800 responses and NPS +10.

Averaging the two location scores produces:

(+70 + +10) ÷ 2 = +40

But that gives 20 responses exactly the same influence as 800.

For NPS across all responding customers, aggregate the underlying responses first.

Enterprise programmes therefore benefit from keeping three views distinct:

Enterprise NPS: calculated from all valid customer responses within the reporting scope.

Location NPS: calculated independently for each individual site.

Location comparison: shows how sites perform relative to one another, alongside response volumes and trends.

All three are useful. They answer different questions.

How Should You Handle Small Locations and Segments?

A score based on six responses should not be treated with the same confidence as one based on 600. Small samples can swing sharply because only a few customer responses need to change.

Set reporting rules before comparisons are published.

Depending on the programme, teams may:

• Suppress scores below a defined response threshold.
• Combine several periods.
• Display response counts prominently beside the score.
• Group smaller entities where appropriate.
• Focus on trends rather than one-period rankings.

The goal is not to hide poor results. It is to avoid false precision.

A regional manager should be able to see the difference between “NPS is -20 from five responses” and “NPS is -20 from 500 responses” before deciding how much confidence to place in the comparison.

How to Calculate NPS Across B2B Accounts

B2B programmes introduce another calculation decision: who represents the customer?

A single account may contain end users, administrators, buyers, technical contacts and executive sponsors. If ten people from one account respond while another account contributes one response, a contact-level rollup gives the first account far more influence.

That may be valid when measuring individual contacts. It is different from assessing account health.

Keep the views explicit:

Contact-level NPS: shows how responding individuals answered.

Account-level reporting: shows how customer accounts are performing or how feedback is distributed within each account.

Define the methodology before seeing the results rather than applying a weighting rule afterwards.

Also preserve different stakeholder perspectives. A satisfied executive sponsor and a frustrated daily user may be describing two genuine parts of the same B2B relationship.

How Should Customer Segments Be Reported?

Segment NPS should answer a business question, not create dozens of extra scores.

Useful dimensions may include:

• Customer tenure
• Service tier
• Product
• Region
• Lifecycle stage
• Account type

Keep the segment definition, response volume and reporting period visible.

A stable enterprise NPS can hide deterioration in an important customer group. At the same time, a change in the response mix can move the enterprise score even when sentiment within individual segments has barely changed.

For example, if a historically lower-scoring customer segment suddenly represents a much larger percentage of respondents, the organisation-wide NPS may fall without an equivalent decline inside every segment.

Use segment reporting to explain the enterprise picture, not replace it.

Keep Different NPS Programmes Separate

Relationship NPS and feedback collected after a specific interaction do not represent the same customer context. Combining them can create movement caused by the survey mix rather than a genuine experience change.

Keep separate calculations for:

• Relationship NPS
• Transactional NPS
• Journey or episodic NPS
• Benchmark programmes, where applicable

Leaders can see several measures together, but they should not be rolled into one unexplained score.

Resonate CX already covers relationship, transactional and episodic NPS in a dedicated guide. Explore the different types of NPS

For enterprise reporting, the rule is simple: compare like with like.

Keep Reporting Windows Consistent

Rolling reporting can stabilize low-volume programmes, but the reporting window must stay clearly labelled.

For example, a May-to-July score and a June-to-August score share two months of the same responses. A small change between the two does not mean the entire customer population changed.

The same principle applies when survey frequency, reporting periods or invitation rules change.

Document the change. Otherwise, teams can spend time investigating what appears to be a customer experience issue when the real change happens in the measurement process.

How Should Enterprise NPS Be Interpreted?

Calculation answers “What is the score?”

Interpretation asks “What should we investigate?”

They are different jobs.

Review NPS alongside additional evidence:

Evidence	What It Adds
Customer comments	Reasons customers describe in their own words
Location or account	Where the change is concentrated
Customer segment	Which groups are moving differently
Journey context	Which experience surrounded the feedback
Response volume	How much evidence supports the score
Behavioural or operational information	What else changed during the same period

A five-point fall in NPS is a signal, not proof of root cause.

Customer comments may reveal recurring concerns. Segment reporting may show where those concerns are concentrated. Operational information may show what else happened during the same period.

AI can help organise comments, surface themes and identify patterns worth investigating. It should support investigation and decision-making, not replace customer and operational context or claim to prove why the score changed.

Build an NPS Reporting Audit

Before publishing enterprise NPS, make sure the headline number can be traced back to its responses.

Calculation Check	What to Verify
Response base	Only eligible responses are included
Categories	Promoters, Passives and Detractors reconcile
Filters	No location, account or segment is unintentionally excluded
Duplicates	The agreed rule is applied consistently
Period	Start and end dates are correct
Hierarchy	Location and account structures are current
Programme type	Different NPS contexts remain separate
Integrations	Expected response feeds are complete
Method change	Any change is documented

This is particularly important when NPS reaches leadership reports.

If somebody asks why the enterprise number changed, the answer should not be, “That is what the dashboard shows.”

The calculation should be traceable.

What to Look for in an Enterprise NPS Platform

An enterprise NPS platform should make complex reporting clearer without hiding how the result was produced.

Look for capabilities including:

• Multi-location and account hierarchies
• Segment reporting
• Response-volume visibility
• Role-relevant permissions
• Separate programme views
• Access to underlying customer feedback
• Flexible reporting periods
• Integration monitoring
• Clear auditability

Ask one practical question during evaluation:

Can a CX lead explain which responses produced the enterprise number, then move into the comments and context behind it?

AI should make that investigation faster. It should not make the calculation logic less visible.

How Resonate CX Supports Complex NPS Reporting

Resonate CX helps organisations connect NPS scores with customer comments, location and customer context, and reporting views that support investigation across complex programmes. It does not change the standard NPS calculation methodology.

AI-Powered Surveys can capture additional context following customer responses. AI-Powered Text Analytics helps organise unstructured feedback into themes and sentiment. Robyn AI gives authorised users a natural-language way to explore available CX information and patterns. These capabilities support investigation rather than treating an AI-generated pattern as proven root cause.

For multi-location organisations, that makes it easier to move from a headline score into the customer evidence and context that may show where attention is required. CX Benchmarking can add industry, location and time-based comparison context where supported.

The value is not a different formula. It is keeping the number connected to evidence teams can investigate and act on.

Frequently Asked Questions

How do you calculate NPS across several locations?

Combine valid Promoter, Passive and Detractor responses from all locations in scope, then calculate NPS from the combined response base. Do not average individual location NPS scores when creating a customer-level enterprise result.

Should Passives be included when calculating NPS?

Yes. Passives remain part of the valid response denominator even though only Promoter and Detractor percentages are used in the final subtraction.

Should B2B NPS be calculated by contact or account?

Both can be useful. Contact-level NPS represents responding individuals, while account-level reporting represents customer accounts. Define and label the methodology rather than blending the two.

Can NPS scores from different customer segments be compared?

Yes, when survey design, timing and response rules are comparable. Always review response volume and check whether changes in segment mix are affecting the enterprise result.

Should relationship and transactional NPS be combined?

No. They represent different customer contexts. They can appear together in reporting, but their calculations and labels should remain separate.

Make the Enterprise Number Explainable

NPS does not become more complicated because the formula changes. It becomes more complicated because the organisation does.

Define the response base. Roll locations up from underlying responses. Decide how B2B accounts will be represented. Treat small samples carefully. Keep programme types and reporting periods consistent.

Then interpret the score alongside customer comments, journey context, segments and other relevant evidence.

That creates an NPS number leaders can question and teams can trace without guessing how it was assembled.

See how Resonate CX can help your organisation connect enterprise NPS reporting with customer context and clearer investigation. Request a demo.