A leading specialty floral and gifting retailer is improving its customer experience by closely collaborating with independent florist partners.
Charlie Cole, CEO of FTD Companies Inc., recently explained to Chain Store Age how his company is effectively managing unexpected demand shifts and fluctuations in stock levels by intensively sharing data with the local florists that provide its inventory and fulfill its digital orders.
What made FTD decide to collaboratively forecast with florist partners?
After joining the team in March 2020, it became clear that FTD’s business model was truly virtuous - if we give our florists a great experience, they are going to give our customers a great experience, and the customer will continue to come back to FTD.
When looking at how we could be a better partner, we realized that collaborating to improve forecasting was one big way to optimize our customer experience without putting undue pressure on our florists. Our supply chain is fairly unique in the sense that we don’t ‘own’ the inventory nor do we have ‘exclusive rights’ to the inventory our florists sell.
This makes it even more important for us to constantly be communicating with our florists to allow for unexpected shifts in demand or large orders, understand and align on-demand early on, and plan appropriately when it comes to our florist partners’ stock and what’s represented online for customers.
What technolog(ies) and strategies do you use to create collaborative forecasts, and when did you implement them?
Generally, with an e-commerce business, you can rely on a simple revenue equation that factors in traffic, conversion rate, and average order value to determine a fairly consistent unit forecast. This is still true in FTD’s case, but the ability to use geographic location trends is irrelevant given the majority of our customers are purchasing gifts for recipients in a different location.
So, while you can typically rely on customer traffic trends as the first factor in any e-commerce forecasting challenge, FTD has had to rely on the ZIP code of the receiver, which is not shared until much later in the process. With this in mind, we’ve taken a new approach and created a constant feedback loop with our florists - what do they have in stock, what can they easily replenish, what trends are they seeing - and complement this data with our own, such as what are we featuring in emails, on the homepage, advertisements. From there, we can narrow down where the majority of the stem volume will be to estimate where orders are flowing from.
What results have you achieved?
The best metric to evaluate success is our ‘reject rate.’ Whenever we send our florists an order, they can either accept or reject it and provide a reason code such as ‘stems not available’ or ‘vase not in stock.’ If we were absolutely perfect, our reject rates would be zero - something worth striving for.
Since working more collaboratively with florists on forecasting, our reject rate has consistently been down over 25%, and sometimes down over 40%. Over Valentine’s Day season when our florists are strained by volatile demand, and this year challenged with some of the worst weather we’ve ever seen, our reject rate was down 34%.
Any future plans you can discuss?
Our data infrastructure is still catching up to our ambitions but we’re working on employing artificial intelligence and machine learning to make this happen in real-time. We’ll continue to be hyper-focused on rebuilding our relationship with our florists and you can expect to see continued changes being made to FTD’s operational and technology partners in the coming months.