I spent an hour talking discounts, data, and the death of margin health with Peter Gardner on the eCom Collab Club Podcast. If you’ve ever wondered why retailers keep running the same promotions despite knowing they’re damaging their business, or how AI actually helps marketers work faster, we covered it all.
The core problem: discounting without strategy
Most retailers fall into a familiar trap. They launch products without properly understanding demand or optimal pricing. When inventory doesn’t move as expected, they reach for the easiest lever available: a sitewide discount.
This creates two distinct problems.
The immediate one is margin erosion—every percentage point off the retail price comes straight out of profitability.
But the longer-term damage is more insidious. Constant discounting trains customers to question your pricing entirely. When you repeatedly signal that your products are worth less than you claimed, you erode brand value. Future product launches become harder to justify at full price.
Take fashion retailers as the clearest example. An autumn/winter collection that doesn’t sell through at expected rates represents both an inventory problem and a sunk cost.
The natural response—dropping prices to stimulate demand—solves the immediate issue but creates lasting perception challenges.
How AI actually helps
We discussed practical AI implementation, specifically how I use Claude to work more efficiently at RevLifter. The conversation touched on something I’ve found crucial: AI tools are only as good as the training you give them.
I maintain what Claude calls “projects”—detailed environments loaded with our tone of voice, case studies, positioning, and strategic objectives. This transforms the tool from a random answer generator into something closer to a well-briefed assistant.
I don’t ask it to produce finished work. I use it to accelerate thinking, generate initial ideas, and review content I’ve already written.
The critical distinction Peter and I discussed: humans learn through stories and experiences, while large language models predict the next statistically likely word.
This creates a fundamental gap. AI-generated content often feels disconnected because it lacks the narrative thread that comes from genuine experience. The answer isn’t avoiding AI—it’s understanding where it adds value and where human judgment remains essential.
Intelligent offers vs. blanket discounts
Our methodology starts with data collection. We implement a tag that gathers behavioural information—not just what customers buy, but how they browse, where they hesitate, when they abandon. This 90-day initial period isn’t about running campaigns immediately. It’s about understanding the specific patterns that indicate purchase intent or points of friction.
Once we’ve mapped the customer journey, we can diagnose specific issues. High product views but low add-to-basket rates suggest one problem. Strong basket building but checkout abandonment indicates something different entirely. Each pattern suggests different interventions.
The key principle: show offers only to customers who genuinely need an incentive to convert. Someone who’s spent fifteen minutes comparing products and has returned twice might benefit from a targeted discount. Someone who’s already at checkout probably doesn’t need one.
Working with Radley provided a clear example. They knew they were over-discounting and understood the brand damage it was causing. By implementing intelligent, targeted offers, we reduced their promotional spend by 8% while maintaining conversion rates and average order values. The goal is to reach 10% reduction within a year—meaningful savings that go straight to margin.
The Black Friday challenge
Peter asked about peak trading periods, particularly Black Friday. This is where most retailers feel trapped. They know sitewide sales train customers to wait for discounts, yet fear being left out if they don’t participate.
My somewhat idealistic view: if retailers used intelligent offers consistently throughout the year, they wouldn’t need massive end-of-season sales. Problems would be solved incrementally—clearing slower inventory through targeted promotions to specific customer segments rather than broadcasting blanket discounts to everyone.
Realistically, we’re not there yet. Most brands continue running their traditional Black Friday promotions. Where we add value is around the edges—solving the problems that blanket discounts miss.
Perhaps conversion rates improve but average order values drop. Intelligent offers can address that specific issue through strategic product recommendations or threshold-based incentives.
Common mistakes and practical solutions
We discussed discount code leakage—one of the most preventable problems retailers face. Using generic codes like “10OFF” or “WELCOME15” guarantees they’ll end up on coupon sites and in ChatGPT training data. Customers will try these obvious variations regardless of whether they’ve earned the discount.
The answer is straightforward: use unique, non-obvious codes. Most platforms support single-use codes. If you must use generic codes for operational reasons, at least avoid the most common patterns.
Another critical point: understanding what problem you’re actually trying to solve. Discounts often become a catch-all solution for issues that require different approaches.
Customer acquisition costs rising? Product-market fit questions? Pricing strategy unclear? Each deserves specific analysis rather than defaulting to promotional discounting.
The managed service reality
One question Peter raised: won’t this take enormous time and resources to implement?
The honest answer shaped our business model. Commerce teams are already stretched thin. They’re managing email marketing, website optimisation, advertising campaigns, and everything else that comes with running digital retail. Adding another system to maintain isn’t viable.
That’s why RevLifter operates as a managed service. We analyse the data, identify opportunities, present recommendations, and execute campaigns. Clients remain in control—they approve strategies and set boundaries—but we handle implementation and optimisation.
The monthly review cycle allows us to assess performance, plan upcoming priorities, and continuously refine the approach.
Why this matters beyond margins
The conversation kept returning to a fundamental question: what’s the strategic role of promotions in retail?
Too many businesses treat discounts as an afterthought or emergency measure. Radley’s experience showed something different. Once they started thinking about promotions within their trading calendar—as a planned tool rather than a panic button—the entire approach shifted. They could align promotional strategy with inventory management, seasonal priorities, and brand objectives.
This represents the real opportunity. Not just recovering a percentage point of margin (though that matters), but fundamentally rethinking how promotional mechanics support business goals rather than undermining them.
I featured on this eCom Collab Club Podcast episode to talk about RevLifter’s approach to offers and promotions. This article was written with the assistance of AI.
