The Top 10 ways AI can improve Retail Store Operations.
The Top 10 AI use cases that Retail Operators at every level told us would transform their daily work. AI for Retail Operations is...
Data-informed coaching solves an important retail paradox. How can I drive more autonomy and decision making into stores while also ensuring the consistency that a prescriptive approach allows.
The very largest retailers are moving away from being overly prescriptive to how stores should run, driven by the need to motivate employees and be more local in decision making. But retailers still need to capture the benefits of network standardisation and brand consistency.
For over 20 years, Store Ops teams have continuously tried to make retail more standardized and uniform, at its most extreme, planning out entirely the activities of a store minute by minute.
Whilst this provided initial benefits like process standardization and brand consistency, such a prescriptive approach is starting to show cracks.
Stores more than ever need to adapt to local needs and challenges, requiring decision-making at a more local level. Managers and teams are revolting against being treated like compliant robots and want to be freed to be the type of empowered team Sam Walton used to talk about.
So should retailers fully let the reigns go?
Such an approach would remove the benefit of cross-network optimization and reduce the consistency of brand and experience that customers want.
So is retail in a forced choice?
At Quorso we have seen that the value lies actually in a middle ground – providing data-informed alerts to employees that are designed by the center to guide their focus, while still giving them a degree of autonomy on how to take action.
Rather than prescriptively telling employees to do certain tasks, we instead send alerts to the right person at the right time, highlighting what is going wrong, with guidance sourced from the whole network on how best to fix it.
The nuance here can be seen in the following example. Rather than prescriptively telling a store team to clean the store floor every morning, (something a well-drilled team does automatically and will get frustrated by having to check off daily), instead show them that their CSAT scores on cleanliness have been trending down over the last 4 weeks and ask them to focus on it.
To demonstrate the breadth of these types of coaching alerts, we are focusing on the following ones with a national convenience chain:
Which managers are always submitting late schedules, not allocating labor to budget, or letting conversion drop in the evenings?
Which new product areas are underperforming because of lack of training? Which individuals struggle to build baskets or sell attachment items at the checkout?
Which employees have hit absence thresholds and are at risk of leaving? Who has been staffed 12 days in a row and might need a break?
Which stores are opening late/closing early, not completing compliance reports, allowing check-out times to balloon, or not getting self-checkouts repaired on time?
Because alerts are data-informed, and only received when relevant, despite the 80+ different use cases, no manager receives more than 11 highest priority alerts per week.
The Chart below shows how compliance improves over time when stores are sent data-driven operational alerts.
There is clear gap in performance between those acting on alerts and those in the control group (approx. 17%). ‘Critical Alert Incidence’ is the number of times a store was outside the SLA of a critical KPI, rebased to week 1 = 100%.
This approach connects humans and data together with unquestionable effectiveness. It follows the principle of chessmaster Gary Kasparov, who explains that an augmented human trained by AI will outperform a grandmaster, or a computer, on their own.
It’s time we helped all our field leaders become grandmasters.