Digital grocery has a growth story.

1. First, they make deliberate choices about fulfillment models. Home delivery may win on convenience, but it is often the most expensive option. Click-and-collect and curbside pickup can reduce last-mile costs while still giving customers speed and flexibility. Many grocers are finding that the smartest path is not to force a single model, but to shape demand across multiple options. That may mean incentivizing pickup, reserving premium delivery windows for customers willing to pay for certainty or using third-party partners to flex capacity only when demand spikes.
2. Second, they improve picking efficiency. Poor picking economics can destroy profitability long before an order reaches the customer. Leading grocers use better order management, real-time inventory visibility and smarter routing to increase pick rates and reduce errors. Publicis Sapient has helped one of the world’s top retailers improve e-commerce order picking rates by 35 percent, while also enabling expanded delivery slots and better pickup experiences. Those kinds of gains matter because every minute removed from the picking process improves both cost-to-serve and customer satisfaction.
3. Third, they modernize forecasting and available-to-promise capabilities. One of the biggest sources of digital friction is offering products that cannot actually be fulfilled. Without robust available-to-promise logic, grocers risk disappointing customers with missing items, substitutions and delayed orders. Better forecasting changes that. By combining current inventory, inbound supply, committed demand and signals such as booked slots, reorder behavior or items sitting in carts, grocers can make far better decisions about what to promise, where to allocate stock and how to protect service levels. Frequent inventory refreshes across digital channels help ensure customers see what is realistically available for their chosen slot.
4. Fourth, they get smarter about substitutions. Substitutions are often treated as a necessary workaround, but they can be a hidden source of churn and margin pressure. Generic rules can frustrate customers, especially when replacement items do not match their preferences or drive up basket cost in the wrong way. More intelligent substitution logic, informed by customer behavior and preferences, can improve acceptance rates and reduce dissatisfaction. In some cases, limiting substitutions on problematic SKUs can actually improve data quality and forecasting performance over time.
5. Fifth, they treat shelf life management as a strategic capability. Freshness is one of the most important predictors of loyalty in online grocery. Customers who would carefully choose the longest-dated item in store will not tolerate receiving produce or dairy with limited shelf life at home. Better shelf-life management means more disciplined inventory handling, stronger FIFO practices, improved storage design, tighter inbound scheduling and quality checks as a final line of defense before delivery or pickup. These operational changes do more than reduce refunds. They protect trust, reduce waste and improve the perceived value of the digital offer.
6. Sixth, they optimize the last mile rather than simply subsidizing it. Last-mile delivery is often the largest cost line in digital grocery. But not every order needs the same service level. Route optimization, better van and batch scheduling, and flexible delivery partnerships can help grocers match capacity to demand more precisely. Publicis Sapient has helped retailers apply machine learning to delivery scheduling and route allocation, improving delivery efficiency while giving operators more control over dwell time or drop density depending on the day’s conditions. That kind of flexibility is essential when margins are tight and customer expectations are rising.
7. Seventh, they adopt more flexible fulfillment architectures. Store picking alone may work at lower volumes, but it becomes difficult to scale economically as digital demand grows. This is where micro-fulfillment, dark-store models and hub-and-spoke networks come into play. Automation can improve picking speed and accuracy. Dedicated fulfillment space can reduce conflict between store shoppers and online operations. Third-party partnerships can provide surge capacity without overcommitting fixed cost. The right answer will differ by market, density and order profile, but the principle is consistent: fulfillment should be designed as a portfolio, not a one-size-fits-all model.