November edition now live · fresh rankings for 903 brands View rankings →
method v1.3

Methodology

How we choose which brands to include, what we measure, and how the measurements become ranks.

Who's included

Brands enter as candidates through public site rankings, ecommerce platform detection and a set of secondary lists. Every candidate is classified and then checked against the rules below. When the classification isn't clear, a person reviews it.

included
own_brandSells its own-brand physical products, made or sourced
own_checkoutSells direct through a checkout on its own site
online_presenceA real online presence: sold mainly online or through specialist shops, not supermarkets
size_floor1,000+ branded searches a month, plus a top-1M site or 20k+ social followers
excluded
retailerRetailers, marketplaces and multi-brand stores
mass_retailSupermarket and drugstore brands, sold mainly through grocery, drugstore and big-box chains
brand_groupGroups and holding companies of several brands (their brands can be listed one by one)
d2c_native

Nearly all sales happen on the brand's own site.

d2c_led

The brand's own site comes first, with some sales through retail or Amazon.

wholesale_led

Mostly sold through retailers and stockists, with its own online store alongside.

We assign tiers from public signals: a store locator, a stockists page, retail partner logos and Amazon listings.

What we measure

branded_search

Branded search

How many people search for the brand by name each month, worldwide. The clearest sign of demand a brand has created.

scale weight 0.35
site_traffic

Site traffic

Estimated visits to the brand's own store, blended with public website popularity rankings.

scale weight 0.35
social_followers

Social audience

Followers across Instagram, TikTok, Pinterest and YouTube, captured as monthly snapshots.

scale weight 0.30

Ad activity feeds Momentum only.

Revenue direction (estimated)

Every brand page shows whether we estimate the brand's revenue is growing, flat or shrinking compared with a year earlier. It answers one question: which way is this brand's revenue heading? It is a direction of travel, never a figure.

  1. Signals. Year-on-year growth in branded search and site traffic, equally weighted: the online signals most closely tied to sales. Social audience is left out because follower counts track revenue poorly.
  2. Periods. The last three months against the same months a year earlier (60%) and the last twelve months against the twelve before (40%), so seasonal brands are compared like with like.
  3. Direction. Roughly +10% or more year on year reads as growing, −10% or worse as shrinking, anything between as flat. When confidence is low, we need a bigger move (about ±25%) before calling it either way: with weak or mixed evidence, flat is the honest answer. Calls within a couple of points of a cut-off are marked "close to flat".
  4. Confidence. High when both signals point the same way for a brand that sells mainly through its own site. Lower when the signals disagree, when only one is available, when search volume is small, or when the brand sells mostly through retailers and stockists, whose sales we can't see.

Each brand page charts the two signals over 24 months against their average in the first of those years (the dashed line): above it means higher than a year earlier. Seeing them separately shows when they disagree, which is what makes confidence low.

What it can't see: price changes, conversion rates, and sales through retailers, marketplaces and Amazon. A brand can grow revenue on flat traffic by raising prices, or lose it to discounting. Treat it as an informed indicator, and the confidence level as part of the answer. Settings are in scoring.yaml under revenue_direction.

Scale and Momentum

scale

how big the brand is today. A 0–100 score built from branded search (35%), site traffic (35%) and social following (30%). It uses a log scale, so a brand 10 times bigger scores roughly 30 points higher. Higher means bigger.

momentum

how fast it is growing. Year-on-year growth in branded search, site traffic and social, over the last 12 months and the last 3 months. 50 means flat. Above 50 is growing (a brand that doubled scores about 88); below 50 is shrinking (halved scores about 12).

Scale is size, Momentum is direction: a small brand can grow fast, and a big one can be shrinking.
BrandScaleMomentum
Mejuri 79 · #9 37 big, but shrinking
Weezie 51 · #320 82 smaller, but growing fast

How scores work

  1. Normalise. Each metric is scored 0 to 100 on a log scale between fixed anchors (for example, branded search: 1,000 a month scores 0, one million scores 100). Every 10x in size adds the same points, and a brand's score doesn't change just because other brands join the index.
  2. Scale is a weighted blend of search, traffic and social audience.
  3. Momentum blends 12-month growth (60%) and 3-month growth (40%), where the 3-month figure compares the last three months with the same months a year earlier, so seasonal brands are treated fairly. Each growth figure combines search, traffic, social and ads. Momentum is absolute: 50 means flat, higher means growing, lower means shrinking.
  4. Smoothing. Inputs use 3-month averages. A brand must appear two months in a row to enter the Top 50 risers.
  5. Editions. Each edition is named for the month it's published and uses data up to the end of the month before: the September edition covers data to the end of August.
  6. Emerging brands. Brands under 5,000 monthly branded searches get a Scale score but no Momentum score.

We publish ranks, scores and an estimated revenue direction only, never raw vendor figures or revenue amounts. Each change to the method gets a new version number and a changelog entry.

scoring.yaml
version: 1.3

normalise:
  transform: log10          # every 10x adds the same number of points
  anchors:                  # [value scoring 0, value scoring 100], clamped
    search: [1000, 1000000]         # monthly branded searches
    visits: [1000, 10000000]        # monthly site visits
    rank: [1000000, 1000]           # public website popularity rank
    social: [1000, 10000000]        # followers, all platforms

scale:                      # how big a brand is today
  search: 0.35
  traffic: 0.35             # visits 0.5 + popularity rankings 0.5
  social: 0.30

momentum:                   # how fast it is growing; 50 = flat
  growth_12m: 0.60
  growth_3m: 0.40           # year-on-year: last 3 months vs same 3 a year ago
  growth_blend:
    search: 0.40
    traffic: 0.35
    social: 0.20            # counts after 3 monthly snapshots
    ads: 0.05
  doubling_scores: 88       # growth of 2x scores 88, halving scores 12
  min_baseline_searches: 5000   # below this: "emerging"

smoothing:
  window_months: 3
  risers_consecutive_months: 2

revenue_direction:          # estimated direction of revenue, not an amount; 50 = flat
  signals:                  # the online signals most tied to sales (social is left out)
    search: 0.5
    traffic: 0.5
  periods:
    growth_3m: 0.6          # year-on-year: last 3 months vs the same 3 a year ago (current direction)
    growth_12m: 0.4
  growing_above: 57         # ~ +10% year on year
  shrinking_below: 43       # ~ -10% year on year
  low_confidence:           # weak or mixed evidence needs a bigger move before we call it (v1.3)
    growing_above: 64       # ~ +25% year on year
    shrinking_below: 36     # ~ -25% year on year
  borderline: 2             # within this many points of a cut-off: "close to flat"
  agree_band: 0.05          # signals "agree" unless they point opposite ways by more than ~5%