Below The Fold

The episode has occasional useful observations - ROAS ≠ profitability, creative becoming the key lever as AI handles targeting, and the hybrid human-AI workflow - but these are surrounded by extended passages of obvious advice and circular repetition about 'testing and scaling' and 'integration is important.' The density of genuinely non-obvious ideas per minute is low.

“when you evaluate a campaign, it's basically you evaluate how much did you spend versus how much did yield in terms of sales. So this is what we call ROA on return on ad spend. And with clients where we've sometimes done achievements of ROA times 30... but when you go back and look at the costs and look at the cost of food, cost of ads... sometimes the brands are not profitable”

“today instead of brands focusing on setting up the campaigns... they should now start shifting into looking at their creatives because as good your creative is as well as the AI will detect it and push it to the right segment and the right audience”

The conversation recycles widely-circulated digital marketing advice - integrate your stack, test before scaling, data is valuable, AI is changing fast. The ROAS-versus-profitability reframe and the observation that creative is now the primary competitive variable are reasonable, but neither is truly contrarian or first-principles; they circulate commonly in performance marketing circles.

“your best benchmark is your own historical data, and you have to build from there”

“it's getting more complex for the creative because the communication message for each audience has to be really different”

Lubna is a genuine operator - MD and shareholder of a 100-person agency operating across 26 countries with 18 years in the industry - not a career thought-leader, and she draws on real client work. However, she runs a mid-sized regional agency and the depth of practitioner insight shared in this episode doesn't fully leverage her experience.

“eight years back I have decided to step in as a shareholder, a managing director for PS Digital to expand the business from being in Egypt”

“today we're more than 100 employees. We serve more than 26 countries all over the world, different industries”

There are a handful of concrete numbers - 70% efficiency gain in copywriting, 60% in coding, ROAS of 30x, a seven-day-to-one-day copywriting timeline, the Amazon AWS 13-hour outage anecdote - but no named clients, no detailed case studies, and tools are deliberately unnamed. The specificity exists in pockets but is not sustained throughout.

“if we speak about copywriting... it takes our team, let's say seven days to build copies without AI... today with AI tools, they are able to finish an entire copy... in one day. So that means my productivity has actually been more efficient, 70% in copywriting”

“Amazon for example, the AWS had an outage of their service for 13 hours because their AI coding agent has coded something and went into production autonomously”

The host asks a few genuinely substantive questions - pushing on the agency pricing model shift, probing whether clients are becoming more demanding as AI speeds up delivery, and drilling into practical CRM use cases - but he rarely challenges an answer, frequently summarises rather than probes, and allows vague responses to pass unchallenged.

“Are you able to become more competitive when it comes to cost to pricing to your clients with AI... do you feel like maybe advertisers and clients are also being a bit more demanding?”

“should agencies start adding prompt engineers in the structures that they're proposing to clients as an additional layer”