The world is buzzing with discussions around ChatGPT and AI

We adore, revere, fear and despise it. At the same time we are also having a lot of fun with mundane and ridiculous searches. In the spirit of enterprise returns management, I did two separate searches on the respective advantages and disadvantages of using returns label in returns management. The results: I was immediately confused by the facts. As it turns out, returns labels are simultaneously efficient AND inefficient, sustainable AND unsustainable, convenient AND inaccessible.

Like a hungry caterpillar it consumes everything without tasting the contradictions.

So when is ChatGPT and AI more than just a gimmick? (Or a caterpillar, as it were!)

 

You don’t need AI to make some returns predictions

For retailers looking to make immediate reduction in returns, we can predicate with almost 100% certainty and – without any AI – that returns will always happen in two of the following use cases:

  • Multi-size orders or bracketing. Shoppers that order the same SKUs with several variants will return parts of their shopping cart
  • Wardrobers and serial returners. If you know your shopper – and we hope that you do – your returns management software – should be able to track and identify wardrobers, fraudulent anomalies and patterns, that will invariably lead to returns

Spoiler alert! These kinds of predictions present an interesting dilemma for brands. If one can block potential returns ahead of time, but at the same time the customer’s shopping cart is reduced, is it worth it? As long as purchases and returns are measures and reported independently, we can predicate (without AI) that brands will continue to see returns as the price of doing business and decide not to reduce the shopping cart even when they have inevitable proof that a return will happen.

AI is not the only means to become more efficient in the returns process

While we are huge ambassadors on technical innovation, a lot of efficiency can come from functional innovation and the transformation of activities with humans as drivers. Is your returns management software end-to-end? Are you leveraging your entire offline eco-system to handle, process and resell your returns? Do you have roles in your organisation related to the monitoring and accountability for returns and the way they affect your profitability? Is your software specific for returns management? Have you analysed and diagnosed what you do and don’t have? Have you invested time and resources in your returns. We are only as good as our weakest link, and if there is no designated returns learning, investment or leadership in your organization, we can predict (without AI) that returns are a challenge you have yet to overcome.

 

So how should we use AI?

 Let’s paint a scenario.  The scenario is not science fiction and is not a gimmick. It is just one scenario. We can share more. For each shopper, at the point of purchase, your returns management solution – like OtailO – calculates the probability of return for that item. It then predicts the next best shelf for this item. By this we mean, factoring in multiple sources of data, such as:

  • the item specifications and condition
  • the nearest inventories
  • the probability of resale, etc.,

From this point, the return is now routed to the most convenient, logical and profitable best shelf.

Sometimes the next best shelf may be your closest franchise where the item can be inspected and resold. Sometimes the return may be sent straight back to the warehouse.  The best route is not always the obvious one. If an item can be resold quicker from a further destination, perhaps a longer transportation cycle is preferable. Or maybe reducing carbon emissions is a stronger corporate goal than reselling returns. In this case it may make sense to route the returns straight to donation.

When the business logic of your brand meets the returns intelligence powered by OtailO, that’s when AI can transform into the butterfly we expect it to be.

 

Photo by Yuichi Kageyama on Unsplash.