Contributor

Alex Rostovtsev

SEO & Automation · Elfsight · Beamtrace

Alex leads SEO and content operations at Elfsight, a no-code widget platform with 2M+ users. He also contributes to Beamtrace, an AI search visibility tracker that monitors how generative engines surface and cite content.

A decade as a geodetic engineer left an instinct for measurement and a low tolerance for claims that can't be verified. He writes about AI research for practitioners with decisions to make.

SEO and content pipelines at a no-code widget platform with 2M+ users.
AI search visibility tracker that monitors how generative engines surface and cite content.

Sep 1, 2026 · 9 min read
Query language decides which market a generative search tool can recommend
arXiv:2608.30052
Sep 1, 2026 · 9 min read
Rank-only incentives can make content look better to a ranker while drifting away from human quality
arXiv:2608.30466
Aug 25, 2026 · 9 min read
RAG systems can collapse when they start citing their own output
arXiv:2608.22118
Aug 20, 2026 · 8 min read
AI search mode cuts publisher referrals without making search feel better
arXiv:2608.18352
Aug 18, 2026 · 9 min read
Generative search can erode the web it depends on
arXiv:2608.15896
Aug 10, 2026 · 9 min read
AI assistants miss 85.6% of venues in a complete Bali market census
arXiv:2608.07069
Jul 28, 2026 · 9 min read
Citation type predicts who grounded LLMs name — and roster-based visibility misses almost all of it
arXiv:2607.23893
Jul 21, 2026 · 10 min read
DRNoise shows how one plausible false document can knock deep research agents off course
arXiv:2607.17291
Jul 20, 2026 · 9 min read
Chinese generative search cites only a small slice of available brand sources, and external quality scores do not predict what surfaces
arXiv:2607.15771
Jul 16, 2026 · 8 min read
GEO can change citations inside a fixed context, but it doesn’t show durable organic visibility
arXiv:2607.14035
Jul 16, 2026 · 9 min read
LLM brand answers are mostly unstable because language changes the signal
arXiv:2607.13304
Jul 7, 2026 · 9 min read
Open-web search answers more questions, but it makes source trust much harder to control
arXiv:2607.05217
Jun 25, 2026 · 9 min read
LLMs mostly source “brand reputation” from other people’s pages
arXiv:2606.25787
Jun 23, 2026 · 10 min read
AI brand “ownership” is moderately concentrated, but the winner changes by model
arXiv:2606.23057
Jun 23, 2026 · 10 min read
English-only AI reputation monitoring misses local champions in multilingual markets
arXiv:2606.23165
Jun 23, 2026 · 9 min read
AI visibility breaks down by entity, not just by mention count
arXiv:2606.21595
Jun 19, 2026 · 9 min read
AI search visibility starts with brand stature, not prompt tweaks
arXiv:2606.20065
Jun 17, 2026 · 9 min read
Incumbent brands get a built-in advantage in LLM recommendations — but only until a competitor clears a narrow threshold
arXiv:2606.17443
Jun 16, 2026 · 10 min read
LLM search agents can be pushed to endorse manipulated web content
arXiv:2606.16821
Jun 12, 2026 · 9 min read
One Polluted Page Is Enough to Hijack LLM Recommendations
arXiv:2606.13610
Jun 10, 2026 · 10 min read
AI brand recommendations move people onto the open web through search, not just mention counts
arXiv:2606.10907
Jun 9, 2026 · 9 min read
Safety-trained RAG models can turn a prompt injection into brand suppression
arXiv:2606.09204
Jun 8, 2026 · 8 min read
FullCite gets better quote grounding by separating the document from the evidence span
arXiv:2606.07130
Jun 4, 2026 · 10 min read
ChatGPT referral spikes can overstate AEO unless you control for platform growth
arXiv:2606.04362
Jun 2, 2026 · 7 min read
LLMs Score 94% on Cultural Knowledge Tests and 40% When the Answer Choices Are Removed
arXiv:2606.01879
Jun 2, 2026 · 7 min read
English Prompts Suppress Bengali Cultural Knowledge Even When Local Evidence Is Provided
arXiv:2605.30481
Jun 2, 2026 · 7 min read
LLM Fact-Checkers Score Well But Retrieve the Wrong Sources
arXiv:2605.30241
Jun 2, 2026 · 7 min read
An LLM Agent That Learns From Its Own Mistakes Beats Human Fact-Checkers on Health Misinformation 89% of the Time
arXiv:2606.02215
Jun 1, 2026 · 7 min read
AI Overviews Sent Users to Reddit. AI Mode Is Taking Them Back.
arXiv:2605.16428
Jun 1, 2026 · 7 min read
Ecosystem GEO Beats Page-Level Optimization by Up to 31 Points for Agent Search
arXiv:2605.12887
Jun 1, 2026 · 7 min read
When AI Cites AI: The Synthetic Source Problem in Generative Search
arXiv:2605.23684
Jun 1, 2026 · 6 min read
Frontier LLMs Hallucinate Up to 38% of Scientific Citations
arXiv:2605.14306
Jun 1, 2026 · 6 min read
Rewording a Buying Question Changes the Brands AI Recommends More Than Switching Models Does
arXiv:2605.27440
Jun 1, 2026 · 6 min read
Query-Specific Expiry: Why 'Recent' Isn't the Same as 'Fresh'
arXiv:2605.13052
Jun 1, 2026 · 8 min read
RAG Doesn't Flatten the Brand Hierarchy — It Just Moves Where You Lose
arXiv:2605.27439
May 31, 2026 · 7 min read
Semantic Metadata Makes Agents More Reliable, Not Smarter
arXiv:2605.28787
May 26, 2026 · 6 min read
One Bad Search Result Breaks Frontier AI Agents — Completely
arXiv:2603.00801
May 22, 2026 · 8 min read
AI Research Agents Fail in Ways Their Own Tests Can't See
arXiv:2603.07379
May 22, 2026 · 7 min read
Most Pages Get Zero AI Citations. Editing 5% Won 40% More.
arXiv:2603.09296
May 22, 2026 · 7 min read
AI Cites Sources It Never Checked — and Half Don't Hold Up
arXiv:2603.07287
May 22, 2026 · 8 min read
Generative Search Citation Share Is a Noisy Estimator, Not a Score
arXiv:2603.08924
May 22, 2026 · 7 min read
Schema Markup Barely Helps AI Find You. Restructuring Does.
arXiv:2603.10700