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
- 0:01:22
Preparing the App for Launch
The speaker reviewed progress on customer support, the help center, a review-request system, and technical visibility optimization, then identified email marketing, ad creatives, and debugging as the session’s priorities.
- 0:03:31
Testing Product-Landing-Page Clarity
In response to first-time viewers, the speaker said the website should quickly explain what he is building; if visitors cannot understand it promptly, that signals a product-communication problem.
- 0:04:49
A Practical Comparison of AI Coding Tools
The speaker shared a personal comparison from using several tools: he said Codex completed design work that another tool had struggled with, while he considered the other tool stronger for debugging and security review. He qualified the design result as possibly specific to that experiment rather than a universal judgment.
- 0:07:03
A Help-Center-Grounded Support Bot
The speaker demonstrated an in-app support flow in which a bot receives a customer question, consults the help-center knowledge base to form a response, and escalates to a person when human support is requested.
- 0:08:37
Hardware and Local Model Use
Answering a memory-capacity question, the speaker said he was not a hardware expert and that his current machine was adequate for his work, but he had not yet reached the stage of running a model locally, so he did not offer a definitive general recommendation.
- 0:09:09
Caution When Giving Models File Access
An audience comment raised token usage and the value of connecting models to additional tools. The speaker then expressed a personal preference not to grant file access to models he does not trust, citing concerns about hallucinations or destructive changes. This was presented as a personal risk posture, not a documented security test.
- 0:10:53
Testing Ad Angles Before Producing Final Creatives
He plans a mix of direct ads, static creatives, and self-recorded videos to test the market. He stresses that these are testing assets rather than scalable production ads: the goal is to identify the angles that resonate before investing in professional production.
- 0:15:04
Direct Install Link or Audit-and-Email Funnel?
He compares sending ad traffic directly to a Shopify app page with a longer landing-page flow that collects store and email details, performs an audit, and follows up by email before installation. He expects the direct link to perform better because the longer funnel adds substantial friction, but presents this as a hypothesis that must be tested using cost per install and longer-term performance.
- 0:17:35
An Early Shopify Ads Underdelivery Observation
He notes that early Shopify Ads campaigns were not spending at the intended rate and that the reason for the underdelivery was still unclear. He frames it as an unresolved issue from an initial experiment, not a confirmed conclusion.
- 0:18:14
Quick Prototypes for Testing Different Ad Ideas
He reviews early creative prototypes including a short video that explains the idea quickly, a version built around a different angle, and a static ad. He notes that finishing touches such as sound effects and background treatment still need refinement.
- 0:19:25
How Can a Campaign Linking Directly to Shopify Be Tracked?
An audience question asks how to track a campaign that sends traffic directly to Shopify and whether measurement would be limited to media performance. He says tracking methods exist and mentions code used through a Chrome-related tool, but the tool name and implementation details are not clear enough in the captions.
- 0:20:37
A unified dashboard for ad and conversion tracking
The speaker discusses consolidating performance metrics in one frequently refreshed dashboard, the difficulty of tracking Shopify Ads, and how platform integrations may support conversion measurement beyond UTM tags alone; the captioned name of the specific mechanism is unclear.
- 0:21:20
Comparing video tools with the same prompt
The speaker says he tried the same prompt in Codex and another video tool and preferred the cleaner result he obtained from Codex. This is a personal assessment from a limited comparison, not a general benchmark of the tools.
- 0:25:30
How digital marketing changed since the mid-2010s
Participants recall that Facebook ad targeting and the Google ecosystem in 2015–2016 were markedly different from today, illustrating how quickly digital marketing platforms and practices have changed.
- 0:26:27
Testing channels when acquisition benchmarks are unavailable
The speaker explains that he prepares sensitivity tables before advertising, but says his research did not reveal a reliable public Shopify cost-per-install benchmark comparable with other platforms. He therefore sees direct experimentation as necessary while acknowledging that market results may outperform or fall short of expectations.
- 0:28:43
Question about recreating a video concept with another tool
An audience participant asks whether the same video concept could be recreated with an alternative tool, compared with the existing render, and what steps such a comparison would require.
- 0:29:23
Learning digital marketing through practice before formal study
The speaker says he began learning digital marketing hands-on in 2016 by experimenting directly with ads, then studied it formally later. In his view, unsuccessful early experiments helped him develop a deeper understanding of the field.
- 0:30:20
Practical Learning in Digital Marketing
Drawing on experience that partly dates to 2016, the speaker says hands-on learning benefited him more than his academy studies and argues that practical work is fundamental to learning digital marketing.
- 0:30:55
Testing Ad Creative Instead of Predicting Performance Without Data
The discussion examines feedback that the dominant visual could make an ad look like a shoe advertisement rather than a message for Shopify merchants. The speaker stresses that CTR and performance costs are difficult to predict before testing when prior data points are unavailable, so he prefers comparing several creative angles. He also explains that audience clarity helps avoid irrelevant clicks and describes CTR as an important metric linked to lower CPC and, in the speaker’s account, lower CPI when traffic and conversion are considered.
- 0:38:23
Using Audience Feedback and Competitor Research to Develop Ad Ideas
The speaker says feedback from the stream community influenced the direction of his work and invites constructive critique. He explains that he reviewed competitors’ ideas, collected them in his files, and developed a set of ad ideas. He was satisfied with one design but wanted to explore an alternative before approving it.
Got a question about something in this stream?
Ask in the group