Author · Co-Founder and CEO
Prof. Aleks Farseev
Co-Founder and CEO, SOMIN · 32 articles on the SOMIN blog
About Aleks
Prof. Aleks Farseev is the co-founder and CEO of SOMIN, a research professor in digital-marketing AI, a Forbes Technology Council columnist, and co-author of 30+ peer-reviewed papers at ACM venues including Multimedia, WSDM and The Web Conference.
He leads SOMIN's research on audience profiling, personality-driven content and explainable AI for advertising; the papers and talks are on the research page.
Articles by Prof. Aleks Farseev
32 articles
How do you connect Meta's AI connectors (MCP) to Claude, ChatGPT or Manus?
Connect Meta's ads MCP server (mcp.facebook.com/ads) to Claude Code, ChatGPT developer mode or Manus: URL, permissions, OAuth steps and safety checks.
How do you know if AI is actually improving your campaigns?
Prove AI is improving campaigns with a holdout group, an incrementality test, a creative A/B test and guardrail metrics, not platform-reported results.
How much does enterprise social listening cost, and which tools give the best value?
Enterprise social listening is usually priced by custom quote. See the pricing models, what vendors publish, and how to compare value instead of headline fees.
What is SOMIN Paid Media, and how does it plan and score ads before spend?
SOMIN Paid Media reads what wins in your category, builds a fresh on-brand funnel and scores each ad for click-through before spend, with sign-off.
How much does SOMIN cost, and what is in each plan?
SOMIN plans start at US$529 a month for Light and reach US$5,290 for Agency Pro, with Enterprise on custom quote. See what each plan includes and how seats work
Why do ads in one category all look the same? Herd behaviour in advertising explained
Ads in one category look alike because brands copy each other when unsure what works. What herd behaviour is, how to measure it and how to break from it safely.
Hyper-personalisation and privacy: personalise without being creepy
Hyper-personalisation and privacy can coexist: use minimal data, explain why people see a message, let them steer it and keep sensitive traits out of targeting.
MCP for marketing agencies: what changes in 2026 when AI reads live data
MCP for marketing agencies lets Claude and other assistants read live category data. What the protocol is, what changes in 2026 and what agencies should build.
Will AI replace agencies? An honest answer for agency leaders
Will AI replace agencies? Not as a category, but it will replace tasks, some roles and hourly fees. What the evidence says and how agency leaders can respond.
What should you do when a buyer says your price is too high?
When a buyer says your price is too high, check your positioning first: in crowded categories, deals are lost on clarity and price is where confusion shows.
Has open-source software made marketing mix modelling easy?
Open-source tools and AI agents make building a marketing mix model easy, but knowing whether it is right still takes expertise, above all in Southeast Asia.
What is agentic commerce, and how do brands get chosen by AI shopping agents?
Agentic commerce is when AI agents search, compare and buy for people. Brands get chosen with machine-readable data, fast APIs and a focus on Share of Model.
Why are proprietary agency AI platforms a risk for brands?
Proprietary agency AI platforms lock brands into tools Gartner says half of will be obsolete by 2029. Ask for modular, vendor-neutral intelligence instead.
What can Meta's Manus agent do for performance marketers, and where does it fail?
Manus speeds up reporting, funnel analysis and creative audits in Meta, but misjudges awareness spend and sees little beyond Meta. A bakery audit shows both.
How is AI changing the role of the marketing scientist?
AI now handles most observational analytics, so the marketing scientist's job moves from building dashboards to explaining why, what it means and what to do.
Will AI agents like Manus replace dashboard SaaS in marketing?
For execution, largely yes: agents like Manus monitor, create, optimise and report on their own, reading data through MCP. What stays valuable is clean data.
How do Meta's AI connectors let outside AI agents work in your ad account?
Meta's AI connectors link ad accounts to outside AI tools through an ads MCP server, so agents can manage campaigns, pull diagnostics and organise catalogues.
What is the AI shadow ledger, and why does AI governance protect brand trust?
The AI shadow ledger is the hidden record of commitments ungoverned AI agents make. A decision architecture with gates and audit trails removes it.
How can AI localise a global brand's message? Lessons from Land Rover Defender
AI localises a brand's message by reading each market's audiences, rivals and creators. For Land Rover Defender, the UK, Japan and US each needed its own story.
What happens to billable hours when AI agents run marketing execution?
When AI agents run execution, billable delivery hours lose value and the premium moves to strategy. What Salesforce's agentic tools mean for agencies.
What did the Fable 5 shutdown reveal about Asia's AI platform risk?
The Fable 5 shutdown showed frontier AI access can vanish overnight. What it means for Asian businesses built on rented models, and how marketers can hedge it.
What is agentic AI in marketing, and how is it different from automation?
Marketing automation follows fixed if-this-then-that rules; agentic AI reasons, adapts and optimises on live data. What the shift means for teams and agencies.
How do advertisers adapt when Meta's AI replaces the media buyer?
Make creative your targeting: with Meta's AI choosing who sees ads, win with genuinely different creative for each persona, grounded in market data.
Why should an agency pitch sell strategic intelligence instead of content?
Agency pitches should sell strategic intelligence because AI has made content execution a commodity and CMOs now buy revenue and pipeline, not volume.
Why is your go-to-market strategy failing, and how do you fix it?
Most go-to-market failures are structural: systems never built to work together and nobody owning them. Fix it by treating marketing operations as strategy.
How do you choose an AI marketing agency that actually delivers growth?
Choose an AI marketing agency on platform ownership, transparency, human oversight, data integration and test velocity. Ask seven questions before signing.
Agency business model in the AI era: from retainers to agents
How AI changes the agency business model: why FTE retainers break when agents do production, which pricing models replace them and what clients still pay for.
SOMIN named among Echelon 2025's Top 10 Startups
SOMIN, the AI creative planning and competitive intelligence platform, was named one of Echelon 2025's Top 10 Startups after e27's TOP100 Program evaluation.
How is AI changing competitor and audience research? Lessons from Marketing Mondays
AI joins competitor ad analysis with audience data, so marketers see what rivals say and what buyers value. A Marketing Mondays case study: Miraflora Wagyu.
What is the AIDA model, and how do you use it to attract and convert customers?
AIDA stands for Attention, Interest, Desire and Action, the four stages a buyer goes through before acting. How to apply each one, with beginner examples.
How can businesses use large language models beyond chatbots?
Beyond chatbots, LLMs condense information and let people run complex tools in plain language, as the OpenAI Startup Fund's bets on Descript and Harvey show.
What are social listening tools, and which ones should marketers know?
Social listening tools monitor and analyse online conversations about brands, keywords and hashtags. The 11 tools we reviewed in 2024, plus SOMONITOR.