Why HyperFans / The research
The right users
do much more.
Some customers spend more. Some keep coming back. Some bring new people in, or create things the whole community uses.
Across games, music, retail and online communities, the same practical question appears: who creates the most value, and what makes them choose you?
Our thesis: understand those people, build content around what matters to them, and test whether it attracts more people who buy, return and participate.
Read the 50 findingsWhere the idea starts
True fans.
Better ways to reach them.
Kevin Kelly’s 1,000 True Fans describes how a small, committed audience can support a creator through a direct relationship. Li Jin’s 100 True Fans extends the idea: serve a narrower group with something valuable enough to pay more for.
HIRO’s bet is to make this more practical for brands, apps, games and IP: combine deep audience knowledge, creative judgment and AI production to test more stories, interests and languages in parallel.
These essays frame the idea. The 50 findings below are the research collection, with each study’s audience, period and measure preserved.
The evidence
50 findings.
Five ways users add value.
Read each finding on its own terms. Spending, repeat use, referrals and contribution are different outcomes. These studies show differences between customer groups; they do not show that every high spender is a fan or that content caused the difference.
50 of 50 findings
Spending
Purchases, basket size and customer value.
SpotifySuper listeners: 2% of monthly listeners, 50% of ticket sales through Spotify.
Audience, period & context
- Audience
- Spotify super listeners compared with other monthly listeners; platform-defined engagement segments.
- Period
- Spotify publication, July 2025
- Method
- Spotify analysis of listening, ticket sales and sharing; averages across artists.
- What this covers
- Ticket sales through Spotify only. Artist totals, observation dates and sample sizes are undisclosed.
- Source
- Spotify for Artists: Super Listeners
- Checked
- 2026-10-19
US music consumersMusic fans were 40% of consumers and accounted for 75% of music spending.
Audience, period & context
- Audience
- Nielsen's aficionado, digital and big-box music fans versus other consumer segments.
- Period
- 2013 Music 360 / April 2013 consumer report
- Method
- Consumer survey segmentation, including a 1,000-person US ePanel and separate PledgeMusic and SXSW samples.
- What this covers
- Historical category spending, not individual artists' revenue. Both concentration measures come from one study.
- Source
- Nielsen: Turn It Up
- Checked
- 2026-10-19
US music superfansUS music superfans reported 80% more monthly music spending than the average listener.
Audience, period & context
- Audience
- US listeners aged 13+ reporting engagement with favorite artists in at least five ways.
- Period
- 2023 Midyear Music Report; August 8, 2023 article
- Method
- Luminate consumer survey; 19% of US music listeners met its superfan definition.
- What this covers
- Self-reported spending and attitudes. Purchasing partly defines the cohort; the comparison does not establish a content effect.
- Source
- Luminate: Why Are Super Fans So Valuable?
- Checked
- 2026-10-19
UK music festivals | Eventbrite28% of UK festival attendees accounted for 56% of annual festival-ticket spending.
Audience, period & context
- Audience
- 504 UK adults aged 18-49 who had attended a UK festival in the previous year.
- Period
- June 17–28, 2016 research; UK release August 8, 2016
- Method
- Survey clustering based on spending, attendance frequency, influence and repeat attendance.
- What this covers
- Self-reported clusters partly defined by purchasing. The linked North American sample supports the same study family, not an independent case.
- Source
- Eventbrite/MusicWatch; issuer release reproduced by National Outdoor Events Association; primary Eventbrite North American analysis
- Checked
- 2026-10-19
KongregateBuyers spending $500+ lifetime were 4% of Kongregate buyers and generated 72% of virtual-goods revenue.
Audience, period & context
- Audience
- Kongregate virtual-goods buyers; 2.1% of active accounts bought over six years. Virtual goods represented 75% of platform revenue.
- Period
- Six years before GDC 2015; some game comparisons use 2014.
- Method
- First-party platform purchase analysis and an anonymized game's session comparison, presented at GDC 2015.
- What this covers
- Revenue shares concern virtual goods only. The session finding comes from one unnamed game and groups users by eventual spending.
- Source
- kongregate-2015
- Checked
- 2026-10-19
Gala Technology | sports gamesGala Technology’s top 1% of paying users generated 71.5% of top-ups in 2021.
Audience, period & context
- Audience
- Paying users ranked by top-ups; the top 1% numbered 7,228 in 2021 and 4,311 in H1 2022.
- Period
- 2021 and first six months of 2022; December 2022 IPO prospectus.
- Method
- Company payment records and percentile tables disclosed in its December 2022 IPO prospectus.
- What this covers
- Top-ups are not recognized revenue; paying users are not all users. The company also identifies this concentration as a risk.
- Source
- gala-2021
- Checked
- 2026-10-19
Android games | AppsFlyerTop 5% of Android game payers generated 45.95% of in-app purchase revenue, averaged across apps.
Audience, period & context
- Audience
- Paying users ranked within each app; the resulting app-level shares were averaged.
- Period
- Q4 chart in January 2023-June 2024 study; Q4 2023 inferred from window.
- Method
- Anonymized provider analytics; the overall report covers 15,000 gaming apps meeting its monthly acquisition threshold.
- What this covers
- Paying users, not all players. The chart subset is unspecified; the average is not dollar-weighted and excludes advertising revenue.
- Source
- appsflyer-2024
- Checked
- 2026-10-19
Mobile games | Swrve sampleTop 10% of monthly payers generated 48.4% of in-app purchase revenue in Swrve’s mobile-game sample.
Audience, period & context
- Audience
- February 2016 active players in sampled mobile games, with payers ranked into deciles.
- Period
- February 2016.
- Method
- Aggregated transaction analytics from Swrve clients; the report describes tens of millions of users and fraud filtering.
- What this covers
- One month of historical purchases, excluding advertising. Exact game selection and weighting are incompletely disclosed.
- Source
- swrve-2016
- Checked
- 2026-10-19
Mobile gamers | ApplifierPlayers gaming 10+ hours weekly reported $15.15 monthly spending, versus $2.55 among those playing 1–4 hours.
Audience, period & context
- Audience
- 3,000 mobile gamers recruited via Mechanical Turk; heavy payers spent $10+ monthly.
- Period
- February 2014; GDC 2014.
- Method
- Survey cross-tabulation of self-reported spending, gaming hours and social behavior, presented at GDC 2014.
- What this covers
- Historical convenience sample. Sharing differences lack exact chart labels; associations do not establish causality or guild leadership.
- Source
- applifier-2014
- Checked
- 2026-10-19
Gaming fans | Fandom/comScoreHeavy gaming fans were 3.5 times as likely as light fans to buy during release week.
Audience, period & context
- Audience
- Heavy gamers played 15+ hours weekly; subgroup sizes were undisclosed.
- Period
- 2017 publication
- Method
- Shared survey of 5,000+ people and separate 250,000-person behavioral panel.
- What this covers
- Observational comparison; subgroup sizes undisclosed.
- Source
- Fandom/comScore, The Power of the Fan, issuer press release
- Checked
- 2026-10-19
eBayEnthusiast buyers were roughly 13% of eBay buyers and generated 71% of gross merchandise volume.
Audience, period & context
- Audience
- 19 million enthusiast buyers within a base of 147 million buyers.
- Period
- Year-end 2021 buyer data, presented at Investor Day in March 2022
- Method
- Company buyer segmentation and purchase records presented at its 2022 Investor Day; buyer share calculated as 19 million divided by 147 million.
- What this covers
- Gross merchandise volume is not eBay revenue. Historical, company-defined enthusiasts are not necessarily advocates.
- Source
- eBay
- Checked
- 2026-10-19
EtsyHabitual buyers were approximately 7% of Etsy buyers and generated about 40% of marketplace gross merchandise sales.
Audience, period & context
- Audience
- Buyers spending $200+ and purchasing on at least six days in the preceding 12 months.
- Period
- 2025
- Method
- Company transaction records segmented by spending and purchase frequency, disclosed in its 2025 annual report.
- What this covers
- Gross merchandise sales are not Etsy revenue. Spending and frequency define the cohort; advocacy is unmeasured.
- Source
- Etsy
- Checked
- 2026-10-19
MytheresaMytheresa’s top 3.8% of customers generated approximately 42.6% of gross merchandise value in FY2025.
Audience, period & context
- Audience
- Mytheresa customers and merchandise value, excluding the other LuxExperience segments.
- Period
- Fiscal year ended June 30, 2025
- Method
- Company customer-program and merchandise-value analysis disclosed in its FY2025 annual report.
- What this covers
- Gross merchandise value is not revenue or profit. Spending concentration does not establish advocacy or fandom.
- Source
- https://s206.q4cdn.com/479417833/files/doc_downloads/2025/Annual-Report-LuxExperience-30-6-25.pdf
- Checked
- 2026-10-19
Neiman Marcus GroupNeiman Marcus Group’s top 2% of customers generated approximately 40% of sales in FY2022.
Audience, period & context
- Audience
- Top customers as a share of all group customers; spending, purchases and retention describe that top group.
- Period
- FY2022 ended July 31, 2022
- Method
- Company customer-segment analysis reported with its FY2022 results.
- What this covers
- Observational sales concentration. Wealth and purchasing power can contribute; emotional fandom and marketing effects are unmeasured.
- Source
- https://www.prnewswire.com/news-releases/neiman-marcus-group-delivers-5-billion-of-gmv-in-fy22-demonstrating-strength-of-integrated-luxury-retail-model-301646944.html
- Checked
- 2026-10-19
FarfetchFarfetch’s top 1% of customers represented 27% of revenue in FY2019.
Audience, period & context
- Audience
- Farfetch's highest-value customer group, discussed alongside its Private Client program.
- Period
- FY2019
- Method
- Company customer concentration disclosure in its FY2019 annual report.
- What this covers
- Historical result; the source calls the measure revenue. It does not establish current performance, causality or fandom.
- Source
- https://s22.q4cdn.com/426100162/files/doc_financials/2019/ar/2019-Annual-Report.pdf
- Checked
- 2026-10-19
Luxury customers | Bain-AltagammaBain estimates just over 2% of luxury customers accounted for 45% of global personal-luxury purchases in 2024.
Audience, period & context
- Audience
- Very important clients within Bain's personal-luxury customer analysis.
- Period
- 2024 estimates; published 2025-01-17
- Method
- Market estimates combining company results, macroeconomic data and over 100 expert interviews; observatory covers 280+ luxury companies and brands.
- What this covers
- Estimated spending concentration, not measured advocacy. Bain's customer and market definitions differ from BCG's.
- Source
- Luxury industry / Bain-Altagamma
- Checked
- 2026-10-19
Luxury customers | BCG-AltagammaBCG estimates top-tier customers were 0.1% of luxury buyers and generated over 23% of spending, excluding mobility and wellness.
Audience, period & context
- Audience
- Top-tier luxury consumers spending over EUR50,000 annually, using BCG's market definition.
- Period
- 2024 model; report July 2025
- Method
- Market analysis using a 7,000-person survey, 1,000 in-depth top-client surveys, interviews and the Altrata-Wealth-X database.
- What this covers
- Spending concentration does not establish fandom or advocacy. Customer tiers and market scope differ from Bain's.
- Source
- Luxury industry / BCG-Altagamma
- Checked
- 2026-10-19
CostcoExecutive memberships were just over half of Costco’s paid memberships and accounted for 75.6% of sales.
Audience, period & context
- Audience
- Executive memberships versus all paid memberships, not total cardholders.
- Period
- FY2026 Q4 highlights
- Method
- Company membership counts and sales penetration disclosed in its FY2026 Q4 investor supplement.
- What this covers
- Executive memberships are just over half of paid memberships. Counts are a snapshot; the self-selected tier does not establish emotional fandom.
- Source
- https://www.sec.gov/Archives/edgar/data/909832/000090983226000084/costex9928-k92426.htm
- Checked
- 2026-10-19
TargetTarget Circle 360 subscribers spent eight times as much on average as nonmembers.
Audience, period & context
- Audience
- Active Circle members and Circle 360 subscribers, each compared with nonmembers.
- Period
- FY2024 report
- Method
- Company comparison of customer spending reported in its FY2024 annual-report CEO letter.
- What this covers
- Customer shares, sample sizes and observation window are undisclosed. Self-selected membership does not establish a causal spending lift.
- Source
- https://corporate.target.com/investors/annual/2024-annual-report
- Checked
- 2026-10-19
NikeNike digital members spent nearly twice as much per transaction as other Nike consumers.
Audience, period & context
- Audience
- Digital members versus other Nike consumers; comparison is per transaction.
- Period
- FY2017 Q3 earnings call
- Method
- Company transaction-value comparison reported on its FY2017 Q3 earnings call.
- What this covers
- Historical basket value, not annual spending or lifetime value. Membership's causal effect and cohort sizes are undisclosed.
- Source
- https://www.sec.gov/Archives/edgar/data/320187/000032018717000022/a8-kq3fy17erccexhibit991.htm
- Checked
- 2026-10-19
AdidasAdidas reported members’ consumer lifetime value was more than twice that of nonmembers.
Audience, period & context
- Audience
- Membership-program members versus nonmembers, using company-calculated lifetime value.
- Period
- FY2021 reporting; LTV model horizon undisclosed
- Method
- Company lifetime-value comparison disclosed in its FY2021 annual report.
- What this covers
- The lifetime-value formula and model horizon are undisclosed. Association does not establish membership's causal impact.
- Source
- https://report.adidas-group.com/2021/en/group-management-report-our-company/global-sales.html
- Checked
- 2026-10-19
Gap Inc.Gap’s active loyalty members spent twice as much annually; the presentation implies a nonmember comparison.
Audience, period & context
- Audience
- 40+ million loyalty members purchasing within 12 months, including credit-card holders.
- Period
- Rolling 12 months through October 2022
- Method
- Company loyalty-cohort comparisons in its FY2022 Q3 presentation, covering the 12 months through October 2022.
- What this covers
- The slide implies a nonmember comparator without separately defining it. Observational differences do not establish program causality.
- Source
- https://s204.q4cdn.com/320226404/files/doc_financials/2022/q3/3Q22-Slides-FINAL.pdf
- Checked
- 2026-10-19
Ulta BeautyUlta customers shopping in stores and digitally historically spent over three times as much as store-only customers.
Audience, period & context
- Audience
- Channel use among US loyalty members; spending comparison uses multichannel versus store-only customers.
- Period
- FY2025 ended January 31, 2026 for cohort share; historic spending comparison
- Method
- Company channel-use and historical spending comparisons disclosed in its FY2025 annual report.
- What this covers
- The share and spending comparison do not explicitly cover the same period. They cannot be combined into a revenue share.
- Source
- https://www.sec.gov/Archives/edgar/data/1403568/000110465926035243/ulta-20260131x10k.htm
- Checked
- 2026-10-19
Bath & Body WorksBath & Body Works loyalty members spent approximately 30% more than nonmembers in four pilot markets.
Audience, period & context
- Audience
- Loyalty members versus nonmembers in four pilot markets.
- Period
- Four-market loyalty pilot reported July 2021; exact dates unspecified
- Method
- Company pilot-market customer comparison reported at its July 2021 investor meeting.
- What this covers
- Historical pilot without a disclosed randomized design, sample size or observation window. The result does not establish current nationwide performance.
- Source
- https://investors.bbwinc.com/static-files/fc187eb8-aa5f-42c2-9f15-50873dc47a05
- Checked
- 2026-10-19
Consumer goods categoriesNielsen reports the highest-buying 10% of category households can account for at least 30% of sales.
Audience, period & context
- Audience
- The highest-buying 10% of households in a product category.
- Period
- March 2015
- Method
- Nielsen summary of category research, discussing 15 studied categories.
- What this covers
- A potential pattern, not a universal ratio. Detailed category tables and profit calculations are not published in the summary.
- Source
- Nielsen: The Making of a Multicultural Super Consumer
- Checked
- 2026-10-19
Japanese character fans7.9% of surveyed Japanese female character fans aged 20–29 reported lifetime character spending of ¥100,000+.
Audience, period & context
- Audience
- 1,000 Japanese adults aged 20+ who liked famous characters; women's age-20-29 subgroup size undisclosed.
- Period
- Survey June 4–6, 2025; approximate cumulative spending to date on respondents' favorite characters
- Method
- Neo Marketing survey comparing self-reported cumulative spending bands by age and gender.
- What this covers
- Cumulative periods vary by respondent. Spending bands cannot establish average spending, lifetime value or company revenue shares.
- Source
- Neo Marketing
- Checked
- 2026-10-19
Marvel comics fans69% of avid Marvel-comics fans reported buying comic memorabilia or collectibles, compared with 39% of casual fans.
Audience, period & context
- Audience
- US adults; Marvel-comics subgroups included 440 avid and 855 casual fans within a 2,200-adult survey.
- Period
- Survey November 19-22, 2021; results published December 2021
- Method
- Morning Consult survey comparing self-reported purchase frequency across fandom-intensity groups.
- What this covers
- Purchase frequency, not transaction records, spending amounts, lifetime value or share of Marvel revenue.
- Source
- Morning Consult National Tracking Poll 2111133, table MCEN1_5, printed page 32 (PDF page index 31)
- Checked
- 2026-10-19
Film fans | Fandom/comScoreHeavy movie fans were 2.2 times as likely as light fans to watch a film on opening weekend.
Audience, period & context
- Audience
- Heavy movie fans watched 9+ hours weekly; subgroup sizes were undisclosed.
- Period
- 2017 publication
- Method
- Shared survey of 5,000+ people and separate 250,000-person behavioral panel.
- What this covers
- Observational comparison; subgroup sizes undisclosed.
- Source
- Fandom/comScore, The Power of the Fan, issuer press release
- Checked
- 2026-10-19
TV fans | Fandom/comScoreHeavy TV fans spent 50% more on television than light TV fans.
Audience, period & context
- Audience
- Heavy TV fans watched 20+ hours weekly; subgroup sizes were undisclosed.
- Period
- 2017 publication
- Method
- Shared survey of 5,000+ people and separate 250,000-person behavioral panel.
- What this covers
- Observational comparison; subgroup sizes undisclosed.
- Source
- Fandom/comScore, The Power of the Fan, issuer release
- Checked
- 2026-10-19
US sports fans | DeloitteSelf-described US sports fanatics reported six times the annual spending of casual fans.
Audience, period & context
- Audience
- Approximately 4,000 US adult sports fans, generating 9,159 league-level responses across five major leagues.
- Period
- Survey April 2015; report copyright 2015
- Method
- April 2015 survey; each participant answered for up to three leagues using self-identified fan affinity.
- What this covers
- Historical self-report. League responses are not unique people, and the cohort's share of total revenue is unknown.
- Source
- Deloitte, Loyalty scoreboard: exploring fan engagement, pages 1–2
- Checked
- 2026-10-19
US consumer AI customersTop 1% of observed US consumer-card AI payers generated 19.5% of spending.
Audience, period & context
- Audience
- US consumer-card AI spenders observed in YipitData panels, across vendors.
- Period
- August 2026; a16z publication October 5, 2026.
- Method
- a16z analysis of YipitData consumer-card and e-receipt panels for August 2026.
- What this covers
- Panel size is undisclosed and purchases may include work tools. The result does not describe every AI app or establish fandom.
- Source
- consumer-ai-yipit-2026
- Checked
- 2026-10-19
OnlyFans | OnlyTraffic sample100 of one million OnlyTraffic-acquired fan subscriptions generated 20.2% of sampled spending.
Audience, period & context
- Audience
- 100 of one million fan subscriptions acquired through OnlyTraffic; unique-person deduplication undisclosed.
- Period
- Provider analysis · updated October 2026
- Method
- Vendor analysis of its own acquired-fan subscriptions; sample window and unique-person deduplication not disclosed.
- What this covers
- A vendor sample, not OnlyFans-wide results. Repeated subscriptions may not be unique people; no observation window or independent audit.
- Source
- OnlyFans / OnlyTraffic sample
- Checked
- 2026-10-19
Starbucks ChinaStarbucks China members using stores, delivery and ecommerce spent ten times as much as store-only members.
Audience, period & context
- Audience
- China members using all three channels versus retail-store-only members.
- Period
- Investor Day 2022; exact spend measurement window undisclosed
- Method
- Company customer-channel spending comparison presented at its 2022 Investor Day.
- What this covers
- China-specific observational comparison. Cohort sizes and spending window are undisclosed; adding channels is not proven to cause the difference.
- Source
- https://s203.q4cdn.com/326826266/files/doc_events/2022/09/1/CORRECTED-TRANSCRIPT_-SBUX_Morning-Session_Reviewed.pdf
- Checked
- 2026-10-19
German bank customersReferred bank customers were at least 16% more valuable than comparable nonreferred customers.
Audience, period & context
- Audience
- Approximately 10,000 bank customers, compared by referral status, demographics and acquisition timing.
- Period
- Almost three years of customer records; Journal of Marketing, January 2011
- Method
- Longitudinal analysis of customer profitability and retention over almost three years.
- What this covers
- Referral-acquisition comparison, not a superfan segment. The margin advantage declined with tenure and varied by customer group.
- Source
- Schmitt, Skiera & Van den Bulte: Referral Programs and Customer Value
- Checked
- 2026-10-19
Returning
Repeat purchases, visits and retention.
Broadway audiences5% of Broadway theatregoers accounted for 28% of tickets in the 2018–19 season.
Audience, period & context
- Audience
- 8,972 questionnaires returned across 49 Broadway productions and 98 performance times.
- Period
- 2018–19 season; published January 13, 2020
- Method
- Audience survey weighted to paid attendance; 17,400 questionnaires distributed with a 52% return rate.
- What this covers
- Tickets and admissions, not revenue. Attendance frequency defines the high-frequency cohort.
- Source
- The Broadway League, The Demographics of the Broadway Audience 2018–2019
- Checked
- 2026-10-19
NordstromNordstrom’s active loyalty members shopped three times as often as nonmembers.
Audience, period & context
- Audience
- Active loyalty-program members versus nonmembers.
- Period
- Reported at Nordy Club launch in September 2018; exact measurement window unspecified
- Method
- Company customer-behavior comparison reported in its September 2018 Nordy Club launch announcement.
- What this covers
- Historical comparison; observation window and total customer count are undisclosed. Membership's causal effect is unmeasured.
- Source
- https://investor.nordstrom.com/news-releases/news-release-details/welcome-nordy-club
- Checked
- 2026-10-19
Williams-SonomaWilliams-Sonoma’s The Key members purchased twice as frequently as nonmembers.
Audience, period & context
- Audience
- Cross-brand versus single-brand shoppers; separately, The Key members versus nonmembers. The Key had 10.2 million members.
- Period
- Investor presentation June 2020; exact measurement window undisclosed
- Method
- Company customer-segment spending and purchase-frequency comparisons presented to investors in June 2020.
- What this covers
- Matching measurement windows are unconfirmed, so no revenue share should be derived. Both comparisons describe one company.
- Source
- https://s24.q4cdn.com/161876561/files/doc_presentations/IRDeck20_Jun15.pdf
- Checked
- 2026-10-19
MLB fans | YouGov40% of very-interested female MLB fans bought merchandise every season, versus 20% of all MLB fans.
Audience, period & context
- Audience
- 774 MLB fans within a survey of 2,000 US adults.
- Period
- Survey February 29–March 1, 2024; published April 11, 2024
- Method
- Online survey weighted by demographics, politics and region; superfans defined as very interested in MLB.
- What this covers
- The all-fan comparison includes superfans. Gender-specific rates need weights to combine; no spending-share result is reported.
- Source
- YouGov, Female MLB superfans buy merch more frequently than their male counterparts
- Checked
- 2026-10-19
UK children's-book buyersHeavy buyers were 5.6% of UK children’s-book buyers and purchased 24.1% of units.
Audience, period & context
- Audience
- Estimated 14.6 million UK children's-book buyers in 2023; heavy buyers bought 16+ books annually.
- Period
- 2023 purchasing; report April 2024
- Method
- Nielsen buyer segmentation: light 1-5 books, medium 6-15, heavy 16+; combined shares calculated from published categories.
- What this covers
- Units, not revenue. Purchase volume defines the groups; buying children's books does not itself establish character fandom.
- Source
- Farshore/HarperCollins Children's Books Annual Review of Children's Reading for Pleasure, page 8
- Checked
- 2026-10-19
US and Canadian cinemagoersFrequent moviegoers were 11% of the US/Canada population aged 2+ and purchased 47% of cinema tickets.
Audience, period & context
- Audience
- US/Canada population aged 2+; frequent moviegoers attended at least monthly.
- Period
- 2019 attendance; survey January 3–29, 2020; published March 2020
- Method
- ENGINE online survey of 8,032 US adults; adults also reported children's attendance, supporting MPA's US/Canada estimates.
- What this covers
- The population denominator includes noncustomers. Ticket share is not revenue share; the underlying survey was US-based.
- Source
- Motion Picture Association, THEME Report 2019, executive summary and page 27
- Checked
- 2026-10-19
UberUber customers using both Mobility and Delivery had 35% higher retention than single-business customers.
Audience, period & context
- Audience
- Customers using both ride-hailing and delivery versus single-business customers.
- Period
- Q2 2025 earnings call
- Method
- Company comparison of customer cohorts, reported by its CEO on the Q2 2025 earnings call.
- What this covers
- Observation window and cohort sizes are undisclosed. Gross bookings are not revenue; service use does not establish fandom or causality.
- Source
- Uber
- Checked
- 2026-10-19
Restaurant and convenience-store loyaltyTop 10% of loyalty members accounted for over 40% of loyalty visits in Paytronix’s analysis.
Audience, period & context
- Audience
- Loyalty members in Paytronix's restaurant and convenience-store customer database.
- Period
- 2023 report; database window January 2020-June 2023; summary April 2024
- Method
- Provider analysis of loyalty transactions, summarized from its 2023 report.
- What this covers
- Loyalty-program spending, not all restaurant revenue. Detailed percentile aggregation and spending denominator are undisclosed.
- Source
- Paytronix: Loyalty Gets Personal
- Checked
- 2026-10-19
SweetgreenSweetgreen loyalty members buying digitally and in stores visited nearly twice as frequently as digital-only customers.
Audience, period & context
- Audience
- Members transacting digitally and in stores versus digital-only customers; cohort sizes undisclosed.
- Period
- Q4 2025 earnings call, February 26, 2026; measurement window undisclosed
- Method
- CEO Jonathan Neman's company-results statement; full earnings-call transcript hosted by The Motley Fool.
- What this covers
- Comparison is with digital-only customers, not all nonmembers. Customer selection may explain the difference; adding channels is not shown to cause higher frequency.
- Source
- Sweetgreen Q4 2025 earnings-call transcript
- Checked
- 2026-10-19
DoorDashAlmost 1/8 of DashPass members ordered 2+ times from each of two businesses monthly, versus almost 1/16 of nonmembers.
Audience, period & context
- Audience
- US DashPass members versus nonmembers; sample counts undisclosed.
- Period
- June 2023; company analysis published October 3, 2023
- Method
- DoorDash analysis of orders placed on its platform.
- What this covers
- Paid membership is self-selected. This comparison does not establish emotional fandom, causality or a share of total revenue.
- Source
- DoorDash: How DashPass members make the most of their local neighborhood
- Checked
- 2026-10-19
InstacartPaying Instacart+ active customers averaged four monthly orders, versus two for nonmembers, in June 2023.
Audience, period & context
- Audience
- Paying Instacart+ customers active in the month versus nonmembers.
- Period
- June 2023; September 15, 2023 IPO filing amendment
- Method
- Company order and membership records disclosed in its IPO filing; original text verified through Fast-EDGAR.
- What this covers
- Self-selected membership; no causal effect established. Spending is not Instacart revenue. The separate five-year comparison concerns different cohorts.
- Source
- Maplebear Inc. S-1/A, Amendment No. 2, September 15, 2023, original filing text mirrored by Fast-EDGAR
- Checked
- 2026-10-19
Referring
Referrals and self-described advocacy.
Consumer-brand fans | NRGBrand fans were 3.5 times as likely as nonfans to describe themselves as advocates.
Audience, period & context
- Audience
- 12,509 US adults aged 18-64; 247 brands, with an average 1,477 evaluations per brand.
- Period
- Survey April-May 2023; report July 2023
- Method
- Nationally representative online survey of stated fandom and brand attitudes, conducted in April-May 2023.
- What this covers
- Self-described advocacy and willingness, not observed referrals or purchases. The comparison does not identify the highest spenders.
- Source
- National Research Group, The Fandom Formula, pages 4 and 13
- Checked
- 2026-10-19
Customer referralsReferred customers made 31–57% more referrals than nonreferred customers, conditional on purchase activity.
Audience, period & context
- Audience
- Referred versus nonreferred customers in the study's field dataset.
- Period
- 2025 journal issue
- Method
- Field-data analysis with purchase-activity controls; preregistered lab experiments replicate the referral effect and a field experiment tests a reminder.
- What this covers
- This concerns referral origin and subsequent referrals. It does not identify the highest spenders or establish a universal effect.
- Source
- Journal of Marketing Research
- Checked
- 2026-10-19
Creating
Content and contributions other people use.
TwitterMost-active 25% of sampled US adult Twitter users produced 97% of their tweets, including retweets and replies.
Audience, period & context
- Audience
- Sampled US adult accounts ranked by tweet volume, including retweets, replies and quote tweets.
- Period
- Tweets June 12-September 12, 2021; published November 2021
- Method
- Pew survey of 2,548 US adult Twitter users, linked to activity from consenting participants' accounts.
- What this covers
- Historical content production, not revenue or positive advocacy. The result is specific to the sampled US adult population.
- Source
- Pew: Behaviors and Attitudes of US Adults on Twitter
- Checked
- 2026-10-19
English WikipediaTop 10% of English Wikipedia editors contributed 86% of measured persistent word views.
Audience, period & context
- Audience
- English Wikipedia editors; surviving words weighted by estimated readership.
- Period
- Historical data through 2006; GROUP 2007 paper
- Method
- Researchers reconstructed text authorship from edit histories and estimated readers' exposure using page-view data.
- What this covers
- Historical content-exposure measure, not article count, revenue or a quality rating. Bot contributions require the paper's methodological context.
- Source
- Priedhorsky et al.: Creating, Destroying, and Restoring Value in Wikipedia
- Checked
- 2026-10-19
Engaging
Time spent with a product.
Steam playersTop 20% of Steam accounts by playtime accounted for 82.4% of recorded cumulative playtime.
Audience, period & context
- Audience
- Registered Steam accounts ranked by recorded cumulative playtime.
- Period
- Ownership/playtime May-November 2013; prices April 2014; IMC 2016 publication.
- Method
- Researchers enumerated accounts through the public Steam API and analyzed recorded cumulative playtime.
- What this covers
- Playtime may include idle games and is not spending. A separate ranking held 73% of estimated library value; that estimate is not actual purchases.
- Source
- steam-2013
- Checked
- 2026-10-19
50 findings from 48 source URLs, across company reports, surveys, platform analyses and academic research. Reviewed October 19, 2026. Some findings share a source or report, and are not independent replications.
The question for your business
Who are your HyperFans?
Start with your own customers. Find who spends, returns, refers or contributes most. Learn what they care about. Build for those interests, then measure whether the next audience acts.
See how HIRO approaches it