Does your brand walk its talk?

BIS compares the values your brand declares with the values its communication carries, value by value and channel by channel. Every point of the score traces back to a sentence.

Request a demo See how it works

71brand values in the classification
5moral motives as the comparison layer
3types of inconsistency, each with an owner
0constants fitted per brand
EN · PLlanguages measured

Telecom operator, anonymised. Measured on the brand’s full materials, 2026-07-16.

Every score decomposes down to a single sentence.

The panel shows the score, the values that pull it down and the sentences behind each drop.

app.brandintegrityscore.com

Brand Integrity Score

66.2/ 100

Telecom operator

Value-bearing sentences641

Settingssame for all three brands

Largest drops

Drop in the score of each value

Service reliabilityafter the nationwide outage

−42%

Frame shift

Fairnessafter the regulator fines

−34%

Customer orientation

−31%

Evidence at sentence level

“Consumers have a right to expect an offer to be fair and free of catches”

Fairness− contradicts the value

Coverage of the brand · reasoning behind a regulator decision

+ supports the value○ neutral− contradicts the value

material sentence by sentence tone of the whole material material score effect on the channel result

Consistency between what a brand says and what it does is very important to 56% of Gen Z.

65%

of Gen Z find a moral message from a brand appealing

53%

of Gen X and Baby Boomers find it appealing

62% · 61%

honesty and trustworthiness, the only qualities Gen Z ranks above consistency

YouGov Profiles, US adults, continuous fieldwork Nov 2024 – Nov 2025, weighted sample · stated attitudes, not measured behaviour

How many brands can put a number on that consistency?

Every metric can read green in a quarter where communication stopped matching the declaration.

Reach, impressions On target

Sentiment On target

Share of voice On target

Survey research On target

Declaration vs. communication No number Not measured

MetricWhat it answersWhat it cannot see

Reach, impressionsHow many people were exposedWhether the message carried the values the brand declared

SentimentWhether the tone about the brand is positiveWhich value that tone is positive about, and whether it is the value that matters

Share of voiceHow much is said about the brand versus rivalsWhat exactly is said, and whether it matches the core of the identity

Survey researchWhat audiences say they thinkWhat the brand actually published, sentence by sentence, across channels

The shared gap

None of these compares declaration with communication. They describe reception or volume. The question of whether a brand says what it promised to say has no number attached, so it gets settled by whoever argues best in the room.

Counting mentions confuses what is discussed with what is said about it.

“Consumers have a right to expect an offer to be fair and free of catches”

Where the sentence comes from From the reasoning behind a regulator decision fining the operator 25 million zloty for misleading consumers. Among the same materials sit a second fine of 15.3 million and a nationwide outage. Coverage of the brand, telecom operator

Tool that counts mentions onlynot a BIS result

Assigned value: Fairness
Sentence reading: + supports the value
Article score: 83.9 out of 100
Verdict: consistent with the value

BISsame sentence

Assigned value: Fairness
Sentence reading: − contradicts the value
Effect: lowers the score of the value
Evidence: this sentence and its source

Effect at brand level · same operator, same materials

Counting mentions

97.3

BIS

66.2

Across all the materials this way of counting puts the operator at 97.3 out of 100, above a confectionery brand with no penalty at all. The tool sees that the article is about fairness. It does not see that the article is about its absence. This is how keyword and sentiment tools count.

BIS reads the same sentence as contradicting the value, and the same operator comes out at 66.2.

Source: coverage of the operator scored by counting mentions only, 2026-07-16 · sentence from the reasoning behind a regulator decision

BIS measures the distance between the declaration and actual communication, value by value.

Weight in the declaration

Illustration

Declaration sources

Brand book and brand platform
Code of conduct, policies
Strategy and communication goals

Declared values · weight

Trust34%

Reliability27%

Fairness22%

Transparency17%

Confirmed by the brand team

The declaration comes from the brand’s own documents.

We extract the declared values and their relative weight from the brand book and brand platform, the code of conduct and policies, the strategy and communication goals. The system proposes, the brand team confirms.

Presence in communication

Illustration

Declaration and communication, per value

Trustaligned

Declaration

34%

Communication

31%

Reliabilitydrift

Declaration

27%

Communication

12%

Fairnessaligned

Declaration

22%

Communication

20%

Transparencyoverreach

Declaration

17%

Communication

29%

Owned media and coveragePosts, video, podcastsPrint, brand exposure, creative

Communication is everything the brand puts in front of people.

For every declared value we measure how strongly it is present in what the brand published: owned media and coverage, posts, video and podcasts, print, brand exposure and creative.

Tone: supports or contradicts

Illustration

One material, sentence by sentence

“Every fee is listed on one page, before you sign”
+ supports the valueFairness
“The new plan is available in every store from Monday”
○ neutralNo value carried
“Consumers have a right to expect an offer to be fair and free of catches”
− contradicts the valueFairness · quoted in coverage of a regulator fine

tone of the whole material+ 1○ 1− 1

Presence is not support.

Every sentence is read for whether it supports the value or contradicts it. A report on a regulator fine mentions fairness and still counts against it.

Per channel and per period

Illustration

Value score per channel and quarter

Q1Q2Q3Q4 Owned media82807881 Coverage74705255 Social71584441 Video and podcasts7679below threshold77

75–10060–7545–600–45

Every drop is tied to a channel and a period.

A value that holds up in press releases and fades on social is a channel problem, not a brand average. Below the measurement threshold we report nothing: a channel with two materials in a quarter still produces a number, but an unstable one.

What BIS does not measure

Sales or market share
Reputation
Creative quality
Board intent
Whether the values were well chosen

The score is a brand’s consistency with its own declaration, not a moral verdict and not a revenue forecast.

The score is set by measurement, not by an AI judgement.

A sentence from communication

Not the name of a value but the frame carrying it.

“Obesity is a disease, not a matter of willpower”

One of 71 brand values

The names a brand already uses for itself. We ground them in the established Schwartz classification of values, tested across 49 cultural groups and 32 languages.

71 values in five groups, one group per motive

One of five moral motives

The five motives that moral judgements come down to. This layer is what makes one brand comparable with another, and one market with another.

CareFairnessLoyaltyAuthorityPurity

01 · Measurement

Every sentence is compared with the description of a value, and we measure how close they are in meaning. That is arithmetic, not an opinion.

Measure first, describe second

02 · AI

The AI enters only after the measurement. It decides whether a sentence supports the value or contradicts it, and puts the finished result into words. Never the other way round.

What that gives you: the same materials return the same score on a rerun, and every number traces back to the sentence it came from.

Source: BIS measurement architecture · value classification: Schwartz and Cieciuch (2022) · moral motives: Haidt and Graham

The same numeric gap means three different problems and three different decisions.

Drift

the value fades out

What the measurement shows: A value that matters in the declaration shows up far more weakly in communication
Example from the materials: The value disappears from social while holding up in press releases
Decision: Agency brief and content plan. The problem sits in production, not strategy

Overreach

promise beyond the declaration

What the measurement shows: A value shows up more strongly than the declaration accounts for
Example from the materials: Communication leads with a value absent from the brand book and the policies
Decision: Either write it into the declaration or withdraw a promise the company does not cover

Frame shift

the value in a bad light

What the measurement shows: The value is present but in a bad light: mentions that contradict it cut its score by more than 30%
Example from the materials: Service reliability: a nationwide outage cut the score of this value by 42%
Decision: This is not a content plan problem. Communication alone will not fix it

Why the distinction matters

The three types have three different owners. Drift goes back to the content team, overreach to whoever owns brand strategy, frame shift to operations and crisis communication. A falling score on its own names no addressee.

DriftContent team

OverreachBrand strategy owner

Frame shiftOperations and crisis communication

Source: inconsistency types recognised by the measurement engine, calibration 2026-07-16

Inconsistency costs more than consistency pays.

Mechanism: hypocrisy

A gap between claim and action makes audiences read the brand as hypocritical. Specific claims carry more risk here than abstract ones.

Wagner, Lutz & Weitz (2009), Journal of Marketing 73

02n = 548

Path: trust and switching

A study of 548 people: inconsistency between mission and action first reads as hypocrisy, then lowers trust, and finally raises willingness to switch to a competitor.

Jung, Bhaduri & Ha-Brookshire (2020), Journal of Product & Brand Management 30(3)

Asymmetry

Inconsistency hurts disproportionately more than consistency helps. Detecting a gap is therefore worth more than confirming alignment.

Šerić & Mikulić (2020)

23%→ 33% in the next edition

What is deliberately missing

The popular “consistent branding means 23% more revenue”. That is a survey of 200 organisations in which respondents estimated the effect themselves, published by a brand-template vendor. In the next edition the same figure grew to 33%. We do not use it, and we would not hand it to anyone defending a budget.

Sources: the peer-reviewed studies cited on the cards. The rejected figure comes from a vendor survey (Lucidpress / Demand Metric, 2016 and 2019)

One formula, no per-brand constants, separates three brands by 14 points.

FMCG brand (A)

80.2

FMCG brand (B)

77.3

Telecom operator

66.2

2,180 + 1,665 + 641 value-bearing sentences across the three brands
every sentence checked separately
the same settings for all three brands
the spread comes purely from the content of the materials

The drops land exactly on the values the operator strained: service reliability after the outage (−42%), fairness after the regulator fines (−34%), customer orientation (−31%).

What we do not know

01
No study measures the chain from value consistency to sales. Ours does not either. The test against an independent criterion is planned and unpassed.
02
How much a place of publication counts is our own call, not a law. The split itself (how often, in what position, in what tone) rests on research. The specific weights we set ourselves.
03
Below the measurement threshold we report nothing. A channel with two materials in a quarter still produces a number, but an unstable one, so we withhold it.

Until that test is passed, none of our materials claims that this score predicts business results.

Source: measurement on the full materials of three brands, 2026-07-16 · review of the research: no study links value consistency to sales outcomes

The first measurement is one quarter as a reference reading.

Calibration

We extract values from the brand’s own documents and map the sources in each channel. The system proposes, the brand team confirms.

One quarter measured

Results per value, per channel and per topic, split into the three types of inconsistency, with evidence down to the sentence.

Readout with the team

An hour walking through the result and deciding what goes back to the brief, what to strategy, and what is not a communication problem at all.

Why a quarter and not a month

A shorter window does not gather enough material in most channels, so the reading would show publication noise instead of a change in consistency. A quarter sold on its own sets the reference point for the ones that follow, and can cover a past period where materials are still retrievable.