On AI-powered pricing and optimising product strategy

Market Mosaic: Pricing in retail and e-commerce keeps changing fast. What's one big shift you've noticed in how brands are approaching pricing today?

Andrew Dremin: The days of "set-it-and-forget-it" pricing are over. Modern retail, perfected by Amazon, now runs on intelligent, Al-powered dynamic pricing. Amazon's algorithms famously change millions of prices daily, but the Al has evolved far beyond just tracking rivals.

New systems, including Generative Al, now predict demand and analyze your specific browsing behavior to find the perfect price at the perfect moment. This lets Amazon strategically slash prices on popular items to look like the low-price leader, while quietly protecting profits on everything else.

The impact is huge: recent data shows this level of Al personalization can boost revenue by up to 40%.

Market Mosaic: How do you think brands can balance short-term promotions with long-term profitability?

Andrew Dremin: The 2025 balance is about value-add and data, not just discounts. What works: Using AI for hyper-personalisation. This allows brands to send short-term (ST) offers only to specific, at-risk segments, protecting margins from loyal customers who would pay full price.

Promos that build long-term (LT) assets are also key. For example, 2025's successful "immersive" campaigns (like L'Oreal's AR try-ons) drive ST engagement while building LT brand equity.

What fails: Breaking the brand promise. In 2025, Southwest Airlines faced backlash for adding bag fees, breaking its core "Bags fly free" (LT) promise for ST revenue. Similarly, a 2025 Unilever "sustainable" campaign was seen as inauthentic greenwashing, damaging LT trust.

Market Mosaic: With private labels growing and loyalty dropping, how can retailers protect margins without losing customers?

Andrew Dremin: This is a tough spot, but you have to attack it from two angles at once. First, your procurement team needs to build new margin, not just cut costs. Stop treating your store brands as just the "cheap" option.

Go premium, like Kroger's Simple Truth-those have fantastic margins and build store loyalty. At the same time, use your buying power to get exclusive flavours or pack sizes from the big national brands, so customers have to come to you.

Second, your strategy has to make price less of an issue. Stop drowning people in options; curate your aisles. And rethink loyalty. Instead of just discounts, give people real value, like early access to products or member-only events. Finally, the biggest move: launch a retail media network. Selling ad space to your CPG partners is a 70%+ margin business that can protect your entire bottom line.

Market Mosaic: What's one underused metric or signal that can help teams make better pricing decisions?

Andrew Dremin: The most underused signal is what I call "Feature Request Revenue Value." Most teams just count how many users ask for a feature. The smart move is to measure the monthly revenue (MRR) of all the customers asking for that specific feature.

This is your best customers telling you what they need to stay or upgrade. It makes pricing easy. If lots of low-revenue users ask, it's a basic feature. If many high-revenue users ask, you've found your next "Pro" tier feature. If just a few huge customers ask, that's your new high-margin "add-on."

Market Mosaic: And finally, what advice would you give brands trying to get pricing and product strategy right in the 2025 market?

Andrew Dremin: Here's my advice for 2025: stop the one-size-fits-all strategy. It's all about precision. First, your product strategy is your marketing. Use "Good-Better-Best" tiers. This is your first layer of hypersegmentation, letting customers sort themselves by what they value. Then use marketing hypersegmentation to sell. Don't show your "Good" (basic) product ads to your "Best" (premium) customers.

To "go viral," build massive buzz around your "Best" tier. Make it exclusive, aspirational, and create an unboxing experience people want to film and share. For pricing, be transparent. Show why it costs more. And use your Al data to send promos only to specific, at-risk segments, not to loyal customers who would pay full price anyway.