Purchase order generation is broken....

[JTPets]

I am not sure how this system is designed to forecast orders, but when it takes 37 units 12 months to sell, the system should NOT be telling me to order 180+

I am not sure what logic is being used to generate this order, but I would love to speak to the team behind it and offer some suggestions on how this should work and what data should be used to calculate. Whatever is currently being done, its broken, or not designed correctly.

Often times the system is not even recognizing that the items were purchased on a previous purchase order and is trying to demand a reorder even with EXCESS stock.

Please let me know who i can connect with to work on making this better.

Thank You!
John

[JTPets]

[Ellie_]

Hi there, @JTPets,

We can definitely take a look at this for you. How long has this been happening for you?

[JTPets]

It's gotten worse in the last 3 months but tbh it's never worked quite correctly.

[RobynR]

@JTPetsI understand that this situation can be frustrating, and I'd like to provide some clarification.

Auto POs let you know which of your items could be running low on inventory. At a high level, Square makes decisions based on a few pieces of information:

1. The current inventory amount of an item, both in stock and on order
2. A sales prediction for that item
3. A lead time prediction for that item

Square will suggest reordering an item if, based on those pieces of information, we predict you will sell out before more items will arrive from your vendor.

For example, you have an item “XL Levi Slim Fit Jeans” with a current inventory of 6 in stock, and 6 on order. We predict based on past POs that this item will take 7 days to arrive from your vendor if you order today. We also predict that you will sell 2 per day. This means that in the next 7 days, you will sell 14 units. However, you only have 12 in stock and on order. This means the item will show up on Auto POs.

You can learn more about Managing Purchase Orders With Square for Retail here. 

I hope this helps. 

[JTPets]

None of this addresses a year to sell, 40 but lets order 200.....

[JTPets]

• The current inventory amount of an item, both in stock and on order - The inventory is low, yes that makes sense to suggest a order, but why 180? Never ordered that qty, not selling at that velocity.....doesn't make sense
• A sales prediction for that item - based on what? A logical decision would be to base it on case pack (you don't support that) and average weekly sales over a reasonable time frame (rolling 13 weeks) and show as "average weekly units column - 1.15"
• A lead time prediction for that item - Based on what? A logical decision would be for us to select lead time in our vendor management which you don't support.

So how exactly is this logic set up and how is the average end user supposed to make decisions based on that forecast without having to manually look up each item 1 by 1 by 1 by 1 by 1 by 1 before submitting? IF i was to follow the system PO generation the system would have me in a financial nightmare.

If i am not mistaken purchase orders are part of square for retail, retail users are not going to be able to memorize sales velocity of items to do that sort of work, a purchase order generation tool is useless if the logic behind it is not build on sound logic, and EASILY understood by the end user to make sound decisions based in reality