Research
NexSpark Research: Original Data and Tests on Dating (2026)
NexSpark Research is where we publish the counts and tests we run ourselves on matchmaking, dating apps and dating trends. Each one is a dated snapshot with the method shown, and a CSV of the raw data where we can share it. It's all first-party: we read the sources directly and count what's there. You're free to cite any of it with a link.
How to cite
Source: NexSpark Research, nexspark.com/research, checked [date]
Use the date shown on the finding you cite. Data is free to use with a link (CC BY 4.0). The license lets you copy, share and adapt it, including commercially, as long as you credit NexSpark. Notes on every file are in the data README.
Findings
The Dating App Review Gap (2026)
In NexSpark’s November 2026 read of the 100 most recent US App Store reviews per app, Hinge drew 34 one or two star reviews, against 81 for Tinder, 83 for Bumble, 85 for Match and 88 for eharmony; 47 of eharmony’s mentioned billing words, against 4 of Hinge’s.
Method. The 100 most recent written reviews per app from Apple’s public US customer-review RSS feed. One page with the summary, key numbers, method and a citation line.
Matchmaker Price Transparency (2026)
When NexSpark checked the websites of 18 US matchmaking firms on November 19, 2026, 9 published a client price, and only 1 of those 9, VIDA Select, also stated how many introductions its price buys. In Columbus, 6 of 7 firms publish a price, from a $350 session to $50,000.
Method. We read each firm’s own website. A firm counts as publishing a price only when a dollar figure for the client fee appears on its own website; third-party reports don’t count.
NexSpark Matchmaker Price Transparency Count
When NexSpark checked the websites of 18 US matchmaking firms on November 19, 2026, 9 published a client price, and only 1 of those 9, VIDA Select, also stated how many introductions its price buys, which works out to about $1,200 per approved match if the $6,000 program is the one that carries the 5-match guarantee, which VIDA doesn't state.
Method. We read each firm's own website. A firm counts as publishing a price only when a dollar figure for the client fee appears on its own website; third-party reports don't count.
Tawkify BBB Complaint Read
In the 36 BBB complaints against Tawkify published between November 2025 and November 2026, 33 mentioned a refund or money back; Tawkify answered every one, and the BBB lists 1 as resolved.
Method. We read every published complaint on Tawkify's BBB profile and counted the ones that mention a refund or money back, match criteria, preferences or radius, and pause or forfeited matches.
NexSpark 15-Photo Reverse Image Search Test
In NexSpark’s November 19, 2026 test of 15 stock portraits, Google Lens traced 15 of 15 originals to their source but only 9 of 15 after a mirror, crop and slight brighten edit; TinEye found 15 of 15 originals but only 2 of 15 edited copies, Bing found 3 and 1, and Yandex blocked the test with a bot check.
On the AI-generated photo page: In our November 19, 2026 test, a mirror, crop and slight brighten hid 13 of 15 photos from TinEye, even though TinEye had found every one of the unedited originals. Read it here.
Method. 15 portraits licensed under the Unsplash License, each searched as an original and as an edited copy (mirrored horizontally, cropped to the central 75 percent and brightened 8 percent) in a US Chrome session. Two results were rescored the same day to apply the scoring rule strictly, and the page says which.
NexSpark Written-Review Gap (App Store)
In NexSpark’s November 2026 read of the 100 most recent US App Store reviews per app, Hinge drew 34 one or two star reviews; Tinder, Bumble, Match and eharmony drew 81 to 88.
Method. The 100 most recent written reviews per app from Apple's public US customer-review RSS feed, pulled November 18, 2026, counting the ones rated 1 or 2 stars. Written reviews skew negative for every app, so the comparison between apps is the finding.
NexSpark Billing-Complaint Count
In NexSpark’s November 2026 count of the 100 most recent US App Store reviews per app, 47 of eharmony’s mentioned billing words (charge, refund, subscription, bill, cancel or renew), against 4 of Hinge’s, 12 of Tinder’s and 22 of Bumble’s.
Method. The 100 most recent US App Store reviews per app, pulled November 18, 2026. A review counts once if it contains a billing word (charge, refund, subscription, bill, cancel or renew).
NexSpark Written-Review Gap (AI Dating Apps)
In NexSpark’s November 2026 count of Apple’s US App Store reviews, 82 of the 100 most recent written reviews of Sitch were 1 or 2 stars, against 66 for Amata, 57 for iris and 50 for Known.
Method. The 100 most recent written reviews per app in the US App Store, pulled November 18, 2026, counting the ones rated 1 or 2 stars. US store and iOS only, one pull per app. NexSpark publishes counts, not reviewer names or text.
NexSpark Match Group 10-K Series
From 2022 to 2025, Match Group’s paying users fell 13% while revenue per payer rose 26%, in NexSpark’s November 2026 read of the company’s annual reports.
Method. Total payers, revenue per payer and direct revenue copied as reported from the Results of Operations tables in Match Group's Form 10-K for fiscal 2021 through 2025. The percentage changes are NexSpark's arithmetic.
NexSpark Month-Theory Tracker
August theory has now had two search seasons, and the second was bigger: in the NexSpark Month-Theory Tracker (Google Trends, US, weekly, checked Nov. 19, 2026) it peaked at 59 on the 0 to 100 index in the week of Aug. 9, 2026, up from 46 in the week of Aug. 3, 2025, after never passing 2 in any week of 2022, 2023 or 2024.
On the November theory page: November theory has had one search season so far: in the NexSpark Month-Theory Tracker (Google Trends, US, weekly, checked Nov. 19, 2026) it never passed 2 on the 0 to 100 index in any week of 2022, 2023 or 2024, then sat between 46 and 53 for every week from Nov. 2 to Nov. 29, 2025, and fell to 19 the week after Thanksgiving as December theory jumped from 11 to 69. Read it here.
On the December theory page: December theory has had one search season so far: on Google Trends’ 0 to 100 index, search interest rose from 11 to 71 in the week of Nov. 30, 2025, right after Thanksgiving, and peaked at 88 the week of Dec. 7, while it never topped 2 in the same October to December weeks of 2022, 2023 or 2024. Read it here.
On the October theory page (Trends reading, no CSV): In NexSpark’s Google Trends reading of November 19, 2026, US searches for october theory peaked in the week of September 29 to October 5 in 2024 (63 on Google’s 0 to 100 scale) and the week of October 12 to 18 in 2025 (100), while the 2026 autumn peak was 42, reached in both the weeks of October 4 to 10 and October 11 to 17. Read it here.
Method. Google Trends, United States, weekly (Sunday to Saturday), as 0 to 100 index values rather than search counts. The latest week is partial.
NexSpark Breakup Calendar
In the NexSpark Breakup Calendar (Google Trends, US, weekly, checked Nov. 19, 2026), “breakup” searches in the four full years 2022 to 2025 peaked in April, February and December, with September ninth of twelve at an index average of 45.9 against 47.1 for the whole year, while “divorce lawyer” searches peaked in August, July and June, matching the 2016 University of Washington study’s August filing peak but not its March one.
Method. Google Trends, United States, weekly interest, averaged by calendar month over the full years 2022 to 2025. Index values, not search counts.
NexSpark Free Confidence-Guide Check
Of the free dating-confidence guides we checked on November 19, 2026, none pairs a full-length workbook with a multi-week practice plan and asks for no email: Psychology Today’s 11 tips run about 1,200 words with no citations, Utah State University Extension’s 7 tips run about 700 words with five references, MentalHappy’s worksheet needs a first name and email, and the Magnetism ebook is 27 pages with a 30-day plan and no email.
Method. Our reading of each guide's own page. Word counts are approximate. The check measures format, not how well any guide works.
NexSpark Atlanta Unmarried Ratio (ACS B12002)
By NexSpark’s computation from the 2024 American Community Survey (table B12002, read from Census Reporter on November 19, 2026), metro Atlanta has 86 unmarried men for every 100 unmarried women aged 35 to 44, against 109 nationally.
On the Atlanta city page: Metro Atlanta has 86 unmarried men for every 100 unmarried women aged 35 to 44 (2024 American Community Survey, table B12002, against 109 nationally). Read it here.
Method. NexSpark's computation from the 2024 American Community Survey, table B12002, read from Census Reporter. Unmarried means never married, widowed or divorced. Each figure is a one-year survey estimate with a margin of error.
Methods and standards
A result only goes on this page if it meets the same standard every time.
- Dated. Every finding carries the date we checked it. Sources change, so each one is a snapshot, not a permanent ranking.
- Named. Each finding has a name, so you can cite it and we can't quietly change what it means.
- Method shown. The linked page says what we counted, where we read it and what doesn't count.
- Raw data where possible. When we can publish the rows without personal data or reviewer text, there's a CSV at nexspark.com/data with a README. For the app review counts we publish counts, not reviewer names or text.
- Corrections noted on the page. When a result changes, the page says what changed and when. For example, on November 19, 2026 we rescored two results in the reverse image search test to apply the scoring rule strictly, and the test page says which ones.
- Limits stated. Review counts measure the review stream, not satisfaction. Search trend values are index numbers, not search counts. Where a comparison measures format and not outcomes, the page says so.
Spotted a mistake? Email hello@nexspark.com and we'll check it and note the correction on the page.
Matchmaking without the five-figure retainer
NexSpark pairs AI with a real matchmaker, and it's free to join and get matched during our launch.
By