Every once in a while you win one by accident. About a year and a half ago I wrote a post A New Tool for Managing Busy Warehouse Days where I described a system that we built for handling volume spikes in our warehouse, both planned and unplanned. From that post:
- On our back-end we’re able to set a package cutoff for any upcoming shipping day. We use historical data along with staffing information to aid us in picking the number.
- Once our shipping queue contains more than that number of packages, the cutoff is triggered. We’re all sent an automated email so that everyone is aware of the situation.
- Orders from that time until our 1 PM shipping cutoff are estimated to ship the following day, along with a message that the estimated ship date “has been adjusted due to high volume” (see image below). We decided to do this to be 100% transparent to our customers who expect that 1 PM same-day shipping.
- When a customer completes checkout, these new estimated ship and delivery dates are reflected in their My Account page and in the confirmation email that we send to them.
- These orders are placed in a separate queue. At 4 AM the following day, they get moved over to our regular shipping queue.
- If a customer complains or really needs their order sooner, we can move just that single order over to our regular shipping queue and process it. Previously we had no way of doing this without running an entire batch, or manually moving all of the orders you don’t want to ship out of queue.
- If the warehouse team gets ahead, they can move the whole batch over and get those orders out a day early.
- In the event of a late arrival due to snow, or several employees calling in sick, we can trigger this immediately by entering in a number that’s lower than what is currently in our shipping queue.
We’ve experienced crazy high volumes at times these past few months. With our COVID-19 policies in place, it takes a little longer on average to do just about everything in the warehouse. Adding staff is always tough, and for a variety of reasons it has been tougher during COVID-19, although we have added several very good new employees recently.
As I write this, we’re pretty consistently 1-2 business days behind normal (our shipping page now displays this in real-time, under our “Fast Order Processing” heading). Occasionally we catch up, and then a Memorial Day happens where you lose a day and have a monster sale. We will eventually catch up for good, but in the interim being able to strategically control the queue has been a life saver. We can adjust based upon staffing, incoming shipments, and other factors. And since we’re transparent with our customers about the adjusted ship and delivery dates, it significantly cuts down on customer service.
Prior to this, the system was only really used on very busy Monday’s in the Spring and during the holidays. All of a sudden it started getting used daily, and often multiple days out (meaning there are orders queued up for several additional days at once). I added a few new features to our back end to make it more manageable, but otherwise it has just worked perfectly.
At the time we built it, I wasn’t so sure about the need for this system, nor was I sure that we picked the right option (in that post I discuss all of the other things we considered). It turns out that we got very lucky. Had this not been in place, we would have been scrambling to build something like it. I’ve been around long enough to know that rushed features never turn out as well as a planned, well-thought-out feature does.