Case studies
Pikkit leads its category. Still climbing.
On September 29 we asked 190 questions across four AI engines - some in the category's words, some in the bettor's. Pikkit was already the most-visible brand we tracked, at 27.6% of answers. Pikkit shipped changes to pikkit.com, and on October 8 we reran the exact same 190 questions.
The rerun: 29% visibility, #1 of all 80 tracked brands, up 1.4 points in nine days. 209 mentions across 720 answers. Named in 155 of 240 category answers, with pikkit.com cited in 101 of them. Action Network held flat over the same window; Betstamp slid 3.3 points.
Source: Pikkit white-glove audit, September 29 baseline and October 8 rerun. 720 answers per run; 0 errors, 0 truncated.
Pikkit29.0%+1.4
Action Network21.4%-0.4
Betstamp18.1%-3.3
Juice Reel13.6%-0.2
OddsJam12.9%+0.5
BettingPros6.5%+1.2
Deltas measured against the September 29 baseline: identical prompt set, identical engines.
Faithful Kids is the most-cited domain in its category.
September 20 baseline, before any work shipped: 190 questions, 720 answers, four engines. Faithful Kids was named in 60 of 359 category answers - 16.7%. faithfulkids.app was cited in 101 of those 359, the most-cited domain in the category, ahead of youversion.com, yippee.tv, bible.com and gominno.com.
On "choosing a Bible app for kids" it was named in 24 of 60 answers. Asked in a parent's own words, with app vocabulary banned, it showed up once in 360. That gap is the roadmap.
Source: Faithful Kids white-glove audit, September 20 baseline. This is a starting point; the post-optimization rerun belongs here once it exists.
Faithful Kids106.0%
Focus on the Family56.0%
Yippee TV47.0%
Minno37.0%
YouTube37.0%
Citation counts across the 720 tracked answers. Faithful Kids leads its entire tracked field.
VoiceOS: named early when named at all. The buyer's questions are wide open.
September 20 baseline: 190 questions, 720 answers, four engines. VoiceOS was named in 8.9% of category answers and voiceos.com was cited in 11.4%. When the engines named it, they named it early: average first-mention position 2.8, better than Wispr Flow, Dragon and Superwhisper. All four engines identified VoiceOS correctly in all 40 brand-seeded checks.
But asked what a buyer would actually type - "I lose the thought", "the keyboard is the bottleneck" - VoiceOS appeared in 0 of 360 answers. 228 of those 360 named no vendor at all. Those questions are still open.
Source: VoiceOS white-glove audit, September 18-19 snapshot, 760 recorded responses across 190 questions and four engines.
Home Game had no visibility. Jam showed them where to focus.
October 5 baseline: 720 answers to 180 unbranded questions about clipping APIs, video search and short-form distribution. Home Game was named in 0 of 720.
The audit turned "why doesn't AI know us" into an exact list of questions, sources and competitors to win. And zero is a baseline that only goes up.
Source: Home Game white-glove audit, October 5 baseline.
OpusClip11.5%
Apify7.8%
Descript6.8%
Reap6.1%
Adobe Premiere5.1%
Home Game0.0%
Who owns the questions Home Game is going after.
How Jam measures.
Two cohorts, kept apart. 90 questions in the category's own words, 90 in the buyer's words with solution vocabulary banned, plus 10 brand-seeded diagnostics that never touch the metrics. The gap between the two is usually the finding.
Competitors are tracked next to you - 15 to 350+ brands per customer, each with share of voice, citation count and average position. Reruns reuse the exact same prompt set, so movement is measured against your own baseline, not vibes.
Recently joined.
About Jam.
Raised from the co-founders of Hugging Face and Lovable, and some VC firms.
Hugging Face
Lovable