Skip to main content

Metrics

A metric is a single dimension of brand perception that respondents are surveyed on. Every analysis or ranking query asks for one or more metrics, and each one becomes a series in the result.

Choosing a metric​

You supply metrics as keys in metrics_score_types (see Score types & scoring populations). Use the key (left column) in requests; the label is the display name.

The 16 metrics fall into three groups:

Brand perception — what people think and feel about a brand:

KeyLabel
indexIndex (Overall Brand Health)
impressionGeneral Impression
qualityQuality
valueValue
reputationCorporate Reputation
satisfactionCustomer Satisfaction
recommendRecommendation

Media & communication — what people are hearing and saying:

KeyLabel
buzzBuzz
attentionAttention
adawareAdvertising Awareness
womWord of Mouth Exposure

Purchase funnel — what brands people plan to use or use now:

KeyLabel
aidedAided Brand Awareness
considerConsideration
likelybuyPurchase Intent
current_ownCurrent Customer
former_ownFormer Customer

The exact survey wording for a metric in a given market and category can be retrieved from the question-texts endpoint. The metrics glossary lists the standard question for each.

Two kinds of scale​

  • Net metrics (−100 to +100): Buzz and the six Index components (Impression, Quality, Value, Reputation, Satisfaction, Recommendation), plus Index itself. These are three-point measures (positive / neutral / negative), scored by subtracting the negative percentage from the positive.
  • Percentage metrics (0 to 100%): the other eight — these are two-point (yes/no) measures.

This distinction is why the score type behaves differently for some metrics — see Score types & scoring populations.

The Index metric​

index is not asked directly — it is an average across the six brand-perception metrics (Impression, Quality, Value, Reputation, Satisfaction, Recommendation). Because it is built from six underlying metrics, it counts as six when the service sizes your request against the data limits. See Request size & limits.

How a metric is scored​

Requesting a metric is only half the picture. Two further choices determine the actual number returned:

  • the score type — which responses count toward the score, and
  • the scoring population — who is included in the base.

Both are covered in Score types & scoring populations.