Building Unfair Advantages
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 ongoing development of Burnish as a GEO application in the Shopify ecosystem. A central workflow uses Claude in Chrome to study competitor products and designs, then passes findings and review feedback to Codex for implementation and debugging, with human inspection still required. Luqman also discusses preparing the go-to-market foundation before spending on ads: analytics and attribution, landing-page, headline, and pricing experiments, lead qualification, and a unified metrics dashboard. He makes clear that many forecasts and experiments were not yet complete. Practical work exposes late-stage difficulties, including a mobile-performance regression after some changes, followed by a clearly measured improvement after a later deployment; whether that gain justified the design compromises remained unresolved. The work also covers deployment and Git-branch friction, pull-request checks, accessibility reviews, and Shopify requirements. Some deployments succeeded, while other technical questions remained open. Audience discussions broaden into multi-agent coordination and usage limits, learning SEO, GEO, and development, the growing importance of marketing and distribution as AI makes building easier, and choosing a familiar market. A substantial secondary thread covers content and audience strategy: batching short-form content or temporarily pausing it to focus on building, building in public, weighing Arabic against North American audiences, YouTube monetization and sponsorships, and livestreaming as both a focus aid and a possible distraction, alongside rest and burnout prevention. The stream documents an unfinished process of development, research, and market preparation rather than a completed launch, with several decisions still requiring measurement and review.
151 moments from the stream
- 4:24:01
Consolidating Marketing Metrics in One Command Center
He discusses gradual progress on go-to-market and customer acquisition work, along with consolidating analytics, product, sales, and customer-management metrics into one command center for healthier monitoring. He notes that the dashboard was unavailable at the time, leaving him without a unified view of those metrics.
- 4:27:00
Reviewing AI Coding-Agent Work and Handling a Problem Outside the Project
The speaker describes passing coding-agent work into a review workflow that opens a URL and checks recent debugging changes. He says this can save time and improve accuracy at the cost of higher token usage. During the experiment, a problem is described as being outside the project and potentially requiring recovery time, so the speaker pauses further action until it is resolved.
- 4:33:42
Debugging a Web Page from a Claude Code Session
The speaker shows an integration that lets the agent see a web page and initiate browser debugging from a coding session after a plugin is installed and enabled. He says an earlier setup required an additional installation that is no longer needed, and finds the newer workflow more convenient.
- 4:37:09
Protecting Work Across Branches and Worktrees
A discussion about preserving ongoing Git work and whether to merge everything into main or continue within separate feature branches and worktrees.
- 4:39:21
Diagnosing Task-Count Aggregation Across Sections
A terminal-assisted check suggested that the dashboard might be combining tasks from separate sections instead of reporting an independent total for each, indicating an aggregation or scoping bug.
- 4:41:38
Safely Saving Work on a Feature Branch
The guidance was to commit and push to the current feature branch—not directly to main—then inventory the branches and worktrees and confirm which work would be saved before proceeding.
- 4:42:22
Could AI Tools Move Toward Dedicated Browsers?
A speculative opinion that some AI-tool companies might develop their own browsers, while explicitly acknowledging that this is only a possibility and their intentions are unknown.
- 4:43:56
Working Directly from the Command Center
A discussion of working directly in the command center instead of maintaining a separate document-center side project for organization and task execution, followed by approval of the proposed approach.
- 4:45:15
Learning Git Branches More Deeply After Problems
A reflection that avoiding the details of Git branches was a mistake, and that encountering a practical problem created the need to understand them more deeply.
- 4:47:04
Resolving a Railway Deployment Error
A brief work update reports that a Railway-related error was fixed and the intended version was deployed successfully, without enough technical detail to establish the cause or exact remedy.
- 4:50:04
How Token Consumption Differs Across AI Models
The speaker argues that token consumption is not uniform across AI models because each model has different token-consumption mechanics. This is presented as a general observation, not a controlled benchmark.
- 4:53:28
Phasing a Shopify App Launch and Marketing Foundation
A three-phase plan covers finishing the app, submitting it to Shopify for review, and scaling after acceptance. During the review period, work continues on go-to-market preparation, marketing foundations, traffic generation, advertising tests, and market validation. The speaker estimated that review could take several weeks, reflecting the expectation at that time rather than a guarantee.
- 4:56:24
Reducing the Risk of Shopify App Rejection
Asked about possible rejection, the speaker explains that a rejection would be tied to specific reasons. Careful adherence to Shopify policies and the submission checklist is presented as the main risk-reduction strategy; no complete fallback route is described.
- 4:57:18
Timing Demand and Focusing on One Product
In an estimate made ahead of Q4 2026, the speaker connects merchants’ preparations with rising interest in digital marketing, SEO, and GEO around late August and September. He uses that estimate to support continued focus on completing and increasingly automating the current product before considering other products.
- 4:59:00
Why the Final Stage of Product Development Slows Down
The speaker explains that progress can be fast early in a project and then slow near completion because the remaining work includes debugging, polishing, and smoke testing. The final stage may therefore be both more important and slower than earlier stages.
- 5:02:34
Controlling Tool Costs and Validating the Funnel Before Ads
The speaker advises monitoring API spending and comparing usage-based billing with a subscription that provides a clearer limit. He also argues against funding ads before building a trusted conversion funnel with A/B testing and a defined objective, aiming for spending to produce a clear return.
- 5:07:09
Competing with an MVP Through Focused Differentiation
The speaker argues that an initial product needs a solid foundation and should meet competitors’ standards while offering a distinct advantage or greater depth in one area. This is presented as a way to begin gaining market share.
- 5:08:25
From Costly Experimentation to a Useful Product
The speakers discuss how experimentation can cost money before a project is mature, while still enabling gradual learning toward something useful that can eventually be marketed. The speaker is willing to accept lower initial margins in exchange for delivering real value.
- 5:13:37
Prioritizing Building Over Continuous Posting
The speaker explains that, at the time of the recording, he had temporarily paused short-form posts and reels to focus his effort on building, noting that it was difficult to work on everything at once.
- 5:15:42
Telegram and Discord in Egypt: A Personal Impression
The speaker shares an informal impression that Telegram is used more than Discord in Egypt, explicitly expressing uncertainty while noting a personal preference for Discord.
Got a question about something in this stream?
Ask in the group