Running out of stock means losing sales you wouldn’t have had to pay for. On the other hand, excess inventory carries a myriad of costs including warehousing, insurance, markdowns to clear obsolete products, and the future loss of full-margin sales on the item. Stated differently, you end up with a warehouse or a store full of ‘assets’ you can’t sell at full price. Both are deadly to the bottom line, and the problem grows proportionally with the size of your business. Yet stockouts and overstocks continue to be among the top omni-channel issues that retailers face.
The two failure modes are really one problem
Inventory stockouts occur due to late or low ordering, while overstocking is the result of early or excessive ordering. In fact, both of these situations have the same underlying cause – a purchase decision made by someone in the organization without taking into account either the lead time, the demand variation, or the cost of carrying too much stock.
Most founders view these as separate issues that require distinct solutions. It’s easy to rationalize adding more overage when you run out of a product, followed by a cutback when you see some other product gathering dust on a warehouse shelf. This reactive strategy will simply swing you back in the other direction, and you’ll continue to swing back and forth unless you introduce the right amount of slack.
However, as tempting as it may be, the solution isn’t to become more cautious, or less cautious. The solution is to introduce a mechanism into the decision-making process that carefully takes into account risky supplier behavior and unpredictable demand while you place every order.
Lead time is the variable that runs the whole show
Nearly every restocking decision is dependent on lead time: namely, the time that elapses between order and delivery. The longer this is, the sooner you must reorder. And the more it varies from week to week, the more safety stock you’ll need to cope with that uncertainty.
Average lead times are understood by most operators. Far fewer can recite the standard deviation. In general, a lead time of 14 days is easier to manage than one where half the time it is a week and the other half it is three. The variability in lead time is likely the root cause of most stockouts, not the average lead time itself. There are plenty of less obvious choices to make too. But in terms of overall inbound process impact, this is the big domino.
Calculating a reorder point that actually works
The reorder point is the inventory level at which you should reorder more of a product. The formula is based on the average demand for that product and the lead time it takes for you to receive that product after you order it. The formula itself is not too complicated.
However, it only works if lead times are consistent, which they almost never are unless you’re in a business where one party can magically predict the behavior of others 100% of the time. Most reorder point calculations become worth less than the paper they’re printed on as soon as the first delay happens. This is everything from a snowstorm that leaves your warehouse shorthanded for a week to your manufacturer’s sales rep taking well-timed family leave, leaving you guessing on shipping time.
To work as intended, the reorder point formula must also accurately predict lead-time demand. That means the forecast needs to be airtight too. Lead time demand is the number of units you’ll sell or use up on average during the lead time to receive a new order. It also includes the time to manufacture, of course, if you’re doing that. If you deal in particularly fickle products or work with suppliers who don’t plan past this quarter’s raw material costs, good luck with that.
Forecasting doesn’t need to be complicated
Estimating demand is usually seen as something that requires a lot of analysis. But, in fact, for most products, a basic approach will do just as well. This involves taking the average of the last three months’ sales to get a baseline of regular demand. Then adding seasonal adjustments – those will be pretty obvious with most products (in which case you don’t need a computer to calculate them for you). Finally, if you’re running any marketing programs that will shift demand (running up stock levels as the promotion approaches), take that into account too.
And that’s it. This will be enough for most products for most of the year. A simple, regularly revised forecast generally outperforms a complex one that is only updated annually. The number isn’t likely to be exactly right, but you’re looking for a figure that’s good enough to work with, while updating it frequently enough that you don’t stray too far from what you can really expect.
Finding the right order size with EOQ logic
Knowing this doesn’t help much, though, as the “right” vs. “wrong” batch size depends on those carrying, shipping, and procurement costs you estimated and a bunch of other factors. How much can you afford to spend upfront, and how much would you miss out on if you lost a sale due to not having enough product? How well do you know your suppliers, and how many different components or finished goods will you need to build this product (more components means a higher likelihood a missing piece gums up production)? How confident are you in your demand forecasts, and how much space could you free up in your warehouse to become more of a third-party logistics provider?
These considerations shoot off in every direction. Maybe you have low upfront costs, established suppliers, and you’re selling a kitchen gadget with a sharp, reliable demand curve, but you’re also developing a few different product variations and looking to expand production while you handle distribution. In that case, ordering larger amounts of most components to build up a bit of that inventory buffer and save on total shipment costs could make sense. Sound about right?
Segment your catalog with ABC analysis
Not every product in your inventory is equally important. When you treat all SKUs the same, your team is overwhelmed and mentally exhausted by trying to manage every minuscule detail all the time.
ABC analysis separates your SKUs by either sales volume (revenue) or contribution margin. The theory is that, generally, about 20% of your products will account for roughly 80% of your sales or revenue. These are your A-items and they require a lot of attention because they are your top performers and cash generators. Stockouts need to be minimized through precise and timely orders, often with smaller, more frequent deliveries.
B-items – about 30% of your SKUs and revenue – aren’t as critical. You don’t need to order for them weekly or even biweekly, and monthly forecasting should suffice.
Finally, C-items are the bottom 50% of your payment or sales volume. They likely don’t contribute significantly to your revenue and aren’t worth the time and energy needed to continually monitor, replenish, and reorder them. Simple min and max inventory levels should be enough to keep stock of C-items on hand.
Move off informal reordering before it breaks
Many ecommerce businesses begin with reordering through emails, texts, or simply placing a call to a supplier. It gets the job done when you’re ordering small amounts of product. But, as soon as you’re juggling more than one product or one supplier, each with different lead times or reorder cadences, entering even a moderate growth phase, or all of the above: the system parks itself on your lawn with its wheels spinning.
Repeatedly reordering without a structured PO process falls apart in the same three or four familiar ways each time:
- Two people order the same product before you realize it and now you’ve over-ordered.
- A supplier agrees to your order in principle verbally, but there’s no log of the commitment anywhere so the order disappears and your stock doesn’t turn up.
- The price gets misquoted or quantities get mixed up.
- You combine your least and most popular products into one mega-PO that the supplier uses as their private testing ground for how much they can overcharge for shipping before you notice.
None of these problems result from anyone not caring enough about the company – they’re simply how things go off the rails with the high volume running through a system that was never intended to track bulk orders. For merchants running on Shopify, setting up structured purchase orders in Shopify turns reordering from a memory exercise into a system you can actually audit – every open order visible, every commitment logged, nothing depending on someone remembering a phone call from three weeks ago.
Reconcile receiving against the PO, every time
Half of the work is done if the purchase order is created. The remaining half has to be done at the loading dock. As items arrive, they ought to be compared to the purchase order regarding required quantity, quality, and price before updating the inventory.
Overlook this small but important step and you’re in trouble. The supplier delivers 90 units instead of the ordered 100 units but no one keeps track of it; now your inventory has a discrepancy. A damaged box is sent to the storage area and then returned, in the meanwhile the sell-through numbers have got chaos due to the inflow of returns. Receiving reconciliation with the purchase order is the simplest, most effective way of ensuring that your inventory is accurate and the most conveniently skipped step in a fast-paced company.
Hold suppliers accountable with OTIF
Measure the percentage of supplier deliveries that are on time and complete. OTIF (On Time In Full) is the go-to metric for measuring your safety stock needs. A supplier with a 97% OTIF has proven you can trust that they deliver in full when they say they will. So they need less of your safety stock to make up for unpredictability. Conversely, a supplier who has a 75% OTIF rate and is steadily late or short on shipments has exposed the full extent of their unreliability – you will likely need to keep more of your product on hand just in case.
Build a review cadence and stick to it
This is not something you set and forget. Inventory management is like plumbing: it needs to be maintained regularly, or it will get clogged. Review your demand forecast monthly. Reverify your lead times with your suppliers each quarter. Shipping conditions and production capacity change more than you think.
Set a calendar on which to purge dead stock on a schedule – quarterly, semi-annually. Bundle it, discount it, or return it if your contract with a supplier allows. Dead stock will not fix itself as long as it’s in inventory, it will just cost you storage and capital. Get the right habits on a schedule, and the stockout-overstock cycle isn’t something you manage – it’s a machine with known inputs that keeps running despite the inevitable surprises from customers and suppliers.
