Skip to main content

Guide: benchmark a brand against its sector

Goal: compare one brand against its category, to see whether it is over- or under-performing the norm.

Approach​

Put two queries in one analysis: the brand, and the sector aggregate. Using the sector median (is_median: true) gives a "typical brand" benchmark that is not skewed by outliers. Keep the period and any resampling identical so the two series line up.

Request​

{
  "meta": { "version": "v1" },
  "data": {
    "queries": [
      {
        "id": "brand",
        "entity": { "region": "us", "sector_id": 1, "brand_id": 1007 },
        "period": { "type": "year", "offset": 1, "amount": 1 },
        "metrics_score_types": { "impression": "net_score" },
        "resample": { "size": 1, "type": "month_from_day" },
        "filters": []
      },
      {
        "id": "sector-median",
        "entity": { "region": "us", "sector_id": 1, "is_median": true },
        "period": { "type": "year", "offset": 1, "amount": 1 },
        "metrics_score_types": { "impression": "net_score" },
        "resample": { "size": 1, "type": "month_from_day" },
        "filters": []
      }
    ]
  }
}

Reading the result​

Two result blocks come back, identified by their id (brand and sector-median). Overlay the two score series; the gap between them is your benchmark.

Variations​

  • Use the sector mean (omit is_median) for a volume-weighted average instead.
  • Add more brands as extra queries to compare several at once.