Price is, in most e-commerce categories, the single variable with the greatest influence on conversion rate, gross margin, and competitive positioning. It's also the variable managed with the least sophistication by the majority of retailers. The dominant pricing model in mid-market Australian e-commerce as of mid-2023 is still some variant of cost-plus with manual competitive monitoring and periodic promotional discounting that isn't tied to real demand signals. That's not a small gap from best practice — it's a structural underutilisation of the most powerful commercial lever available.
Rule-based pricing — if competitor A is at $X, price at $X minus 3%, floor at cost plus 20% — was a reasonable approach when competitive data was hard to collect and pricing changes were expensive to implement. Neither of those constraints applies anymore. Real-time competitor price monitoring is a commodity service. Modern commerce platforms can execute thousands of price changes per day without engineering intervention. The constraint that remains is the intelligence layer: the system that determines not just what price a competitor is charging, but what price your specific customer, in this specific demand context, at this inventory level, with this margin constraint, should see. That's a materially harder problem than rule-based monitoring, and it's where most pricing tools haven't yet delivered on the pitch.
The distinction worth drawing is between reactive pricing — adjusting to external signals like competitor prices and promotional calendars — and predictive pricing, which models demand elasticity at a SKU level and optimises for a specified objective function (revenue, margin, inventory velocity, or some combination). Priceloop, which we backed in 2023, is working in the predictive pricing space, applying ML models to actual demand signals rather than just competitive data feeds. The key technical challenge is data volume: predictive pricing requires enough transaction history per SKU to build reliable elasticity models, which means the category is more tractable for high-volume retailers with broad catalogues than for niche merchants with sparse sales data. That constraint shapes where the earliest high-value deployments sit.
There's a genuine risk in the dynamic pricing category that we think about carefully. Retailers that deploy aggressive dynamic pricing on commodity or comparison-shopped categories without a clear customer communication framework can damage trust — particularly in an APAC consumer context where pricing transparency expectations vary significantly by market. A consumer in Australia who sees a price change between browsing and checkout may simply abandon. A consumer in a market where dynamic pricing is more normalised in e-commerce may respond differently. The systems that get this right are the ones that understand the customer experience dimension of pricing, not just the revenue optimisation function. Intelligent pricing and trusted pricing have to coexist, and the most interesting companies in this space are designing for both simultaneously.
The medium-term opportunity we're watching in pricing infrastructure is the shift toward full margin stack optimisation — pricing systems that coordinate with inventory management, promotion planning, and supplier cost inputs in real time rather than treating each as independent decision domains. That's a more complex integration problem and a harder organisational change for retailers to navigate, but the value unlock from margin stack coordination is substantially larger than from pricing optimisation alone. The companies that build toward that integration layer are the ones we think will define the category.