The First Numbers Are Coming In
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 documents preparations for a Shopify app intended to improve merchants’ product visibility in AI answers, with practical work on help-center-grounded support, onboarding emails, scan-flow clarity, progress states, and pre-merge review. Much of the session focuses on launch readiness: developing static and video ads with tools such as Codex, Remotion, and Higgsfield, then using audience feedback to prevent the creative from being mistaken for a shoe ad and to explain AEO within seconds. Luqman discusses testing multiple angles and formats across Meta and Shopify rather than predicting CTR, CPC, or CPI without data; comparing a direct-install link with a scan-and-email funnel; and using predefined success criteria and a unified analytics dashboard. Ad underdelivery, attribution, incomplete data integrations, target-market selection, pricing, and campaign structure remain unresolved or subject to testing. Other discussions cover reaching Gmail’s Primary tab, the ethics of reviews and UGC, avoiding unsupported marketing promises, and personal—not benchmarked—comparisons of AI tools. Audience questions also explore learning programming and marketing through practice, repository-level specialist agents, changing software roles, and building skills or projects with limited resources. No launch success is claimed: the session is presented as an evolving work record whose market results will require weeks of measurement.
171 moments from the stream
- 3:25:57
Self-Teaching Programming Before the Rise of AI Tools
A participant says they taught themselves programming without formally studying computer science, and the speaker views acquiring those skills before the rise of AI tools as a noteworthy achievement.
- 3:29:02
CLI Tools and Agent-Based Workflows
A positive view of running coding agents from the terminal, including a previous workflow built around Warp. The speaker is considering returning to CLI tools but does not want to disrupt a currently productive workflow midstream.
- 3:31:03
Reading Software-Market Change Beyond Denial
The speaker addresses criticism that earlier warnings about changes in programming never materialized, arguing that many warnings were misunderstood and that the market has changed. He also explains why he no longer spends time answering every dismissive comment.
- 3:32:23
Balancing Ad Creatives with Testing Budget
Responding to a suggestion to add more ad creatives and explain the tool more clearly, the speaker says the appropriate number of creatives depends on the testing budget: larger budgets support more experiments, while limited budgets require a narrower set.
- 3:33:25
The Difficulty of Discussing Technological Change Defensively
The speaker reflects on the difficulty of helping people who react very defensively to technological change. He understands the fear behind that response and sometimes wants to help, but sees discussion as unproductive when there is no willingness to listen.
- 3:35:44
Is AI a Bubble?
The speaker cites others' view that AI may be a bubble but disagrees, comparing the skepticism with reactions to the early internet. He presents the comparison as an opinion rather than a settled conclusion.
- 3:36:45
Forward Deployed Engineering and Adapting to AI-Era Roles
The speaker argues that the decline of some technical roles does not mean the end of a career, because AI may create new positions. He gives Forward Deployed Engineer as an emerging example and recommends recognizing market shifts and adapting one's path and skills. Some later examples of specialties are unclear in the captions.
- 3:39:07
How Industrial Shifts Change the Job Market
A brief opinion that the Industrial Revolution eliminated many jobs while creating others, with the expectation that this pattern will recur and has already begun.
- 3:39:29
A Personal Take on Glasses for Screen Use
Answering an audience question, the speaker described his yellow-tinted lenses as having blue-light filtering and an anti-reflective coating, saying they help him with screens and reduce headaches; he was unsure of the coating’s exact name.
- 3:41:46
Checking an Advertising Funnel’s UX and Marketing Email
The speaker reviewed the user experience of an advertising funnel, identified changes he wanted to make, and then checked whether a marketing email had arrived; it had not arrived at the time of the check.
- 3:45:24
What Does “Creative” Mean in Digital Marketing?
He explained that a “creative” is the advertisement itself—whether video, a static ad, or another format—and argued that strong creative work is important for sales.
- 3:46:22
Staying Focused While Working on a Livestream
The speaker said he cannot maintain full focus while working and reading livestream comments simultaneously. Multitasking is not one of his strengths, but he treats streaming as practice and acknowledges that it is difficult.
- 3:47:29
Comparing AI-Generated Ad Creatives and Applying Feedback
The speaker compared two ad versions, describing one as created from scratch with Codex, Remotion, and GPT, and the other as assembling elements with Higgsfield. He agreed with repeated feedback that prominently showing a shoe could make merchants mistake it for a footwear ad, said he would change it, and leaned toward the Higgsfield version.
- 3:49:23
How CTR, CPC, and CPM Reflect Ad Creative Performance
Explains that each ad creative produces its own results and that click-through rate (CTR) helps assess its ability to attract attention. A higher CTR was described as generally lowering cost per click (CPC) and significantly affecting CPM, which is why ads are designed to encourage clicks.
- 3:50:36
Testing a Better Layout for Product Logos and the CTA
A practical layout iteration for a Burnish for Shopify ad: place the product identity and AI-engine logos at the bottom, then have the final call to action appear below and push those elements upward. The change was framed as an experiment to evaluate the layout and implementation time.
- 3:53:11
Making Progress Clear During Long-Running Operations
A review of a scan flow that takes noticeable time: the existing progress bar may not make it clear that processing is still active. The proposed improvement is a visible animation that keeps users from assuming the scan has finished or stalled.
- 3:56:05
Creative Volume Depends More on Budget Than AI
In response to a question about how many creatives there would be without AI, the speaker said the quantity is driven more by budget than by the use of AI itself.
- 4:01:02
Solo Operations with AI Assistance
Based on observing a one-person company model, the speaker argued that choosing the right niche and using AI effectively may let an individual run a relatively substantial operation. This was presented as an opinion and impression, not verified evidence about the company’s size or results.
- 4:04:36
Testing Multiple Ad Creatives in the Market
The speaker accepted design feedback but explained that the ad creatives would be tested in the market, so several alternatives should be tried before reaching a final judgment.
- 4:04:56
A Unified Analytics Dashboard for Market Validation
The speaker outlined a plan for a data pipeline and single dashboard combining metrics from platforms such as Meta, Shopify, and Google Analytics, with mobile access. He noted that the integrations were not yet wired and the displayed numbers were not valid, and expected the following weeks of data to inform market validation and next steps.
Got a question about something in this stream?
Ask in the group