Using Demand Forecasting to Grow Your eCommerce Brand

People aren’t robots. They don’t behave according to fixed patterns day after day – and that can be a headache for your eCommerce business. For example, something that sold well last month might now be sitting in a warehouse gathering dust. To avoid this, you need to know about demand forecasting.

 

What is demand forecasting?

In eCommerce, demand forecasting means predicting future sales using data on your business’ past performance. You’re finding out when and why individual products sold well (or poorly) and using that knowledge to optimise your strategy for the future.

To effectively carry out demand forecasting, eCommerce brands need to monitor the right metrics and leverage their more subjective insights to work out what’s going to sell. The goals are better inventory management and a healthier cash flow.

 

How will demand forecasting benefit your business?

It’s hard to imagine a situation where knowing more about how your customers behave wouldn’t help your brand. However, there are four main benefits to proper demand forecasting…

Cut down on risk

Balancing supply and demand can be tricky for big eCommerce businesses who typically hold a lot of stock. That stock has to be stored somewhere and the storage has to be paid for. Demand forecasting gives you a clearer idea of how much stock you’ll need at any given time. This knowledge makes it less likely that you’ll end up holding too much surplus and paying over the odds.

Optimise your pricing strategy

“Everybody knows umbrellas cost more in the rain,” goes the Tom Waits lyric. Rather than charging a flat price for products with fluctuating demand, forecasting gives you extra options. You can drop the price when you know demand will be low, clearing out stock. You can also raise prices if you know people will badly want what you’re selling.

Satisfy your customers

Running out of stock doesn’t only mean you have no way of making more money until new stuff comes in. Customers can also lose faith in your brand if they find themselves unable to get what they need from you when they need it. With proper demand forecasting, you can always keep shoppers happy without overstocking.

Stay consistent all year

Trends, seasonality and public holidays are merely three examples of things that can send demand up and down. Bikinis in summer and sweaters in winter are obvious enough. But some fluctuations in demand for less clear-cut products might surprise you. Demand forecasting lets you see what did well, and when, in previous years so you can plan ahead.

 

What types of demand forecasting are there?

The methods you use to forecast demand for your products comes down to what you’re trying to achieve. In general, you’ll be engaging in one or more of these varieties…

Quantitative demand forecasting

Using quantitative data means working with numbers and objective facts. Numerical data you’ve been collecting over time can give you a solid picture of how your products perform over, say, the past 12 months. Look at:

  • Average sales per month
  • Moving sales averages
  • Seasonal trends
  • Disruption caused by world events

Assuming no external influences, it’s reasonable to assume a similar year-on-year performance for most products. Of course, things aren’t always as straightforward.

Qualitative demand forecasting

New products and emerging technologies mean you can’t always rely on past data. Sometimes, that data simply doesn’t exist. Qualitative forecasting uses more subjective means to predict sales trends.

What do experts think about a new technology’s commercial prospects? How do focus groups respond to new products? How might fashions and fads rise and fall over the coming years? Qualitative data takes a more intuitive approach to gauge the best possible informed estimates.

Macro-level demand forecasting

This is the big picture; how do your products perform against competitors in your entire sector? To do macro-level demand forecasting, you’ll need plenty of broad market research. If you’re unable to capture this yourself, it’s likely you can buy it from third parties.

Typically, macro-level insights will help you expand into new product ranges or introduce new bundle deals. You can see how it went when other companies made similar moves and emulate success as best you can.

Micro-level demand forecasting

Micro-level demand forecasting concerns itself with your business, benchmarked against its own past performance. How are your products performing year on year (both by category and individual SKU)? What profit margins are you hitting? How healthy is your cash flow, and can you anticipate any incoming hurdles?

Long-term demand forecasting

By ‘long term’ we mean anywhere from 12 to 48 months, or even longer for the biggest eCommerce businesses. Knowing how demand might change in the distant future allows you to adjust your supply chain and branch out into emerging sales channels.

Short-term demand forecasting

By thinking about the more immediate future, anywhere up to 12 months, you’ll be able to forecast seasonal demand more accurately. Short-term demand forecasting is a good way of ensuring you don’t overstock or run out of fast-selling items unexpectedly.

 

Key metrics to consider in demand forecasting

The way you approach demand forecasting in eCommerce depends on the nature of your business and the sector you’re in. Your objectives might be optimizing stock levels, introducing new products, opening up new audiences; most business goals benefit from being able to predict demand.

This will dictate the exact formula that’s right for meeting your goals, but some things are always handy to know…

Reorder point formula

Demand forecasting is all about knowing how much of your inventory is going to sell. That’s helped by knowing your reorder point formula. Using this formula, you can work out the point at which you need to restock items to avoid running out.

To work out your reorder point formula, you need to combine your lead time demand and your safety stock level:

  • Lead time demand is the average time in days it takes to restock an item, multiplied by your average daily sales of that item in units
  • Safety stock level is worked out by first multiplying your highest ever daily sales of an item (in units) by its longest ever restock time (in days). Take this number and subtract your average daily sales, multiplied by average restock time

Average order value

Average order value (AOV) is the average spend per customer per visit to your store. Once you know this value, there are plenty of strategies you can use to increase it during times of peak demand. And, of course, with demand forecasting in place, you’ll know when those times should be.

Creating bundle deals, offering free shipping (on orders above AOV) and introducing new, related products are all solid options when you know you’re approaching a busy period.

Return rate

It’s pointless meeting even the most massive demand if customers are only going to send your products back to you and ask for a refund. Knowing your return rate ideally also means knowing why people are sending items back and addressing the root causes. Combined with demand forecasting, it can help you plan a strategy that ensures your revenue stays in your account.

Final thoughts

Early mankind needed to know about the changing of the seasons to properly plant crops. In the same way, your eCommerce business needs to anticipate what customers need and when they’ll need it. Building demand forecasting into your sales strategy is your ticket to anticipating and meeting that demand.