Ads, Data & AI Marketing Agents
Want to back the stream? The «داعم» membership on YouTube is EGP 60 a month: badges, emoji, and I answer your questions first on stream. Become a supporter
The gist
The stream follows the connected work of building and marketing a Shopify app: testing it in a real store environment, seeking additional perspectives and user feedback, monitoring speed and bugs, and making the product’s value quickly understandable. Much of the discussion concerns advertising readiness and measurement—reviewing creatives, correcting social-page branding, consolidating platform metrics into a data pipeline, and choosing a conversion event that balances signal volume with proximity to the real goal, such as installation or completed onboarding. Luqman argues that creative intuition alone is unreliable and that testing, market response, and data matter more, while stressing that trials need time to mature and that the campaigns had not yet demonstrated success. The group also discusses Meta’s shift toward broad targeting, “learning limited” campaigns, lead-capture funnels, and attribution, leaving questions such as the best optimization event, the value of lookalike audiences and affiliate marketing, and parts of tracking and implementation unresolved. The stream also presents experiments and proposals involving Claude, Codex, MCP, and specialized agents for campaign setup, ad review, cross-session strategy handoffs, content production, SEO, and social publishing. Luqman emphasizes human review because AI recommendations can become inconsistent, and describes designing an agent that challenges assumptions instead of automatically agreeing. Audience questions cover possible Arabic–English differences in token use and coding quality, the limits of using AI for advertising without marketing fundamentals, and how to choose a SaaS idea or learning specialization. Broader discussions address market and competitor research, unit economics, the difficulty of customer acquisition, and the value of starting with a focused market or specialty. They also include explicitly uncertain forecasts about autonomous agents, robotics, optional work, and large one-person companies, followed by personal—not universal—views on partnerships, accountability, disagreement, and the social effects of economic pressure.
87 moments from the stream
- 2:09:29
Will AI replace marketing? Today versus a 2031–2032 outlook
Luqman says the current planning experience does not show that AI has eliminated marketing, while clarifying that his broader claims concern a future outlook around 2031–2032 rather than present-day reality. He cautions against judging long-range predictions as though they describe today’s capabilities.
- 2:10:37
Could AI and Robots Make Work Optional?
A speculative discussion of a prediction attributed to Elon Musk that work could become optional, followed by a partly unclear prediction about large-scale robotics reshaping agriculture, manufacturing, and manual labor. The speaker stresses that the future remains uncertain and difficult to imagine.
- 2:15:19
An Opinion on the Motives of AI Industry Leaders
The speaker argues that some AI leaders’ statements reflect long-term missions more than an immediate pursuit of money or publicity, while acknowledging competition between companies. This is presented as personal opinion, not proof of their actual motives.
- 2:17:30
Avoid Inventing Reasons When Assessing Ad Attribution
A brief critique warns against overcorrecting and fabricating reasons to reject an option, noting that attribution may still be possible even if it is not configured today. The specific option being evaluated is not fully clear from this page.
- 2:20:18
An Initial Shopify vs. Meta Ads Scalability Comparison
Directing ads to Shopify is described as a good idea, but Facebook or Meta ads are tentatively favored as the more scalable option if feasible. A point about a new funnel begins here but is not completed within the page.
- 2:21:37
Language Model Limits and the Future of AI
The discussion describes a language model as a statistical next-token predictor that can still fail on seemingly simple tasks. It then speculates—without presenting the outcome as certain—about AGI or superintelligence, robotics, and AI’s future integration with the economy.
- 2:25:18
Choosing the Right Conversion Event for Meta Ads
The speakers discuss Meta’s commonly cited need for enough weekly conversions to stabilize learning and support scaling, while differing on how strictly that threshold should govern a campaign. They compare optimizing for a high-value event such as a direct install with an easier upper-funnel event such as submitting contact details: the cheaper event can produce more signals, but it carries less value and requires follow-up. They also explain how broad targeting and conversion signals help the platform identify likely responders, emphasizing that the key first decision is which conversion event to optimize for.
- 2:30:43
Optimizing for an Install Click or Completed Onboarding
A discussion of whether to optimize ads for an easier, cheaper event such as an install click or for a deeper, higher-intent event such as completing onboarding and selecting a Burnish subscription tier. The group notes that clicks may attract store owners who never finish the flow, references conversion tracking through Shopify and Meta, and keeps in-app tracking as an unconfirmed possibility.
- 2:36:44
Splitting Campaigns and Adapting to Event Costs
A proposal to split effort between a direct campaign and a lead-capture campaign, then reconsider the mix when their expected event costs appear similar. The qualified recommendation is to watch early results and adjust allocation if assumed event costs prove inaccurate rather than treating forecasts as fixed facts.
- 2:37:41
A Context-Aware Marketing Skill That Challenges Plans and Records Disagreements
A description of a custom AI marketing skill supplied with context and goals and tuned to challenge a plan realistically instead of agreeing automatically. The workflow also records points of disagreement so earlier disagreements and reasoning patterns can be revisited in future discussions.
- 2:41:00
Marketing as Reading and Strategy Work
Luqman describes marketing as different from rapid interactive coding: it involves more reading, discussion, and strategizing, with room for differences of opinion.
- 2:45:19
Designing a Specialized Marketing-Agent Workflow
He outlines a future plan that avoids relying on ads alone by combining them with organic traffic, SEO, and social publishing. One specialized agent would research industry topics and pass them to other agents that draft LinkedIn and X posts in a recurring workflow. He also distinguishes a skill from a background agent assigned to a specific task; this was still a proposed implementation, not a completed result.
- 2:48:39
How Meta Ads Targeting Has Changed
Answering an audience question, he contrasts the granular “laser targeting” common roughly ten years earlier with today’s broader-audience approach. He attributes to Meta the view that its algorithms can learn from the ad and conversion data, making the quality of conversion signals important to targeting. This reflects his explanation of the platform’s guidance, not a demonstrated test result on this page.
- 2:50:23
A Separate Strategy for Shopify Ads
He tentatively suggests that Shopify Ads may need its own strategy and different budget assumptions before the broader plan is finalized. The passage is brief and provisional, with no final outcome presented.
- 2:51:03
What Should AI Review in a Search Ads Audit?
Practical instructions for reviewing campaign history, strategy, cost per click, spend, and results, including search terms, keywords, and negative keywords. The page does not show the audit’s outcome.
- 2:51:59
Evaluating a SaaS Idea with Research and Criteria
Answering how the idea was chosen, the speaker says he spent several weeks researching and used AI tools to generate and compare ideas against considered criteria such as healthy unit economics, margin, reachable demographics, and competitive intensity. Prior experience from the merchant side helped him understand Shopify users’ needs; this is presented as his process and opinion, not a guarantee of success.
- 2:55:40
Building Is Not Enough: Reaching Customers
The speaker describes a pattern he has observed among builders: repeatedly creating products without entering the market or selling. In his view, customer acquisition and turning an unknown project into something people buy are among the hardest stages, and more building does not solve distribution.
- 2:56:54
Limits of Relying on AI for Advertising
Responding to a question about using AI for ads without experience, the speaker says useful assistance still depends on knowing how to ask and understanding marketing and measurement basics such as conversions. Planning what the tool should inspect and how it should present findings also remains difficult.
- 2:57:48
Matching a Business Idea to the Market, Goals, and Resources
The speaker argues that a market’s potential for virality is not enough to justify an idea if its unit economics do not fit the founder’s goals, resources, and time. He suggests defining those constraints in a structured conversation with an AI tool, then testing generated ideas against the chosen criteria. His assessment of the Egyptian market is presented as an opinion from that process, not a guaranteed general conclusion.
- 2:59:49
How Do You Validate Merchants’ Willingness to Pay?
A participant asks how to establish that merchants will pay for an app when no clear comparison is available. The speaker initially says thorough research should establish that a market exists and begins a practical example, but neither the example nor the answer is completed on this page.
Got a question about something in this stream?
Ask in the group