How to Win at E-commerce Demand Forecasting in 2023

How to Win at E-commerce
Demand Forecasting in 2023

The mounting pressure of volatile supply chains, finicky consumer demand, and a bevy of new competitors make e-commerce demand forecasting more important for etailers than ever before.

Etailers need to know, with accuracy, what their customers want to buy, at what time, and for how much. A rigorous e-commerce demand forecast could help maximize their profits and reduce losses in overstocking, understocking, inventory, storage, and other challenges they deal with every day. 

This means an etailer with meticulous e-commerce demand forecasting can bet on success; especially as e-commerce growth is predicted to increase, despite a newly announced recession.

 

Line graph indicating an e-commerce forecast for growth from 2007 to 2021 and estimated growth from 2022 to 2026
Source: Euromonitor, National Data Sources, Morgan Stanley Research estimates

Let’s take a look at what e-commerce demand forecasting involves. 

 

What is e-commerce forecasting?   

Demand forecasting is a process of analyzing historical data (and other factors like seasonality, cannibalization, sentiment, etc.) to anticipate customers’ future willingness to purchase products and/or services. 

The same is true whether it’s demand forecasting for e-commerce or brick-and-mortar sales. Retailers of all kinds attempt to predict the volume and type of products customers will want to buy during a set period of time. 

 

Why is e-commerce demand forecasting important? 

E-commerce demand forecasting is important because of how complex large-scale e-commerce businesses are.  

Sizeable e-commerce businesses are:

 

  • Highly competitive with new, strong players entering the market regularly.
  • Inventory-heavy with a vast array of items.
  • Serving large geographies with complicated fulfillment logistics.
  • Sometimes combined with brick-and-mortar stores, making it more complicated to forecast inventory demands for both channels.

E-commerce businesses are also in competition with giant analytics-first retailers such as Amazon and Walmart. Plus, the speed and competitive nature of e-commerce leave no time for mistakes.  

Without a very good e-commerce demand forecast to help etailers thrive, the implications stemming from these situations could lead etailers to:

 

  • Out-of-stocks; driving customers to competitor websites.
  • Overspending on inventory that no one buys, and underspending on inventory that is in demand.
  • Failing to provide the preferred inventory desired by a diversity of customers in varied geographies.  

Ultimately, an accurate demand forecast is crucial if etailers want to keep customers, reduce inventory costs, and capitalize on the best opportunities.

This is easier said than done however when considering the challenges of e-commerce forecasting. 

 

Demand forecasting challenges in e-commerce

Demand forecasting is never easy. Etailers are constantly contending with the three Vs of big data–volume, variety, and velocity–pouring in minute-by-minute, day-by-day. It’s overwhelming.

Individual overwhelmed with the volume of data. Unable to use all data points for a demand forecast.

All that data is also useless unless etailers can store, process, and analyze it to their benefit. 

Coupled with this, are challenges that include:

 

1. Quality of demand forecasting tools

Not all demand forecasting tools are equal in ability, especially when it comes to the three Vs of big data that relate to customers, competitors, pricing, and other online commerce activities.

Some tools are too basic, capable of only using past sales to forecast demand. A more accurate forecast can factor in a variety of data points to arrive at real-world demand.

Some demand forecasting tools require a lot of manual input resulting in human error, more time, and more labor. And this results in missed selling opportunities. 

Conversely, some forecasting tools are automated, and therefore better for scalability to accommodate complex etailers. 

 

2. Online competition

Comparison shopping is just a click away for online consumers. There’s no need to go from one store to the next online shopping.

Etailers need to be dynamic in real-time and stay ahead of the competition with: 

  • Pricing

Price comparison takes place in mere minutes between multiple products on multiple websites. 

E-commerce retailers must have the right product at the right price in stock.  

 

  • Product visibility  

Consumers have choices galore online with “endless aisles”. What customers see on the first couple of online pages is important. They won’t scroll endlessly looking for what they want.  

Etailers need to know which products are in the highest demand for shoppers, so they are given priority on web pages. 

 

  • Up-selling/Promotions

Similar to knowing what products should appear on the first page, not knowing which products a customer is likely to buy together means lost up-selling opportunities. 

Running promotions that are not optimized for the shopper does not achieve the goal of additional sales, and may result in losses. 

 

3. Balancing inventory for omnichannel retailers

Most omnichannel retailers struggle with the calculations to optimize the inventory quantities they need.

Inventory calculations for brick-and-mortar (A) and e-commerce (B) are unique for each channel. Combining them into one demand forecast is an immense undertaking that falls short of the best human efforts.

 

Calculating a demand forecast for both brick-and-mortar stores and e-commerce combined is too much for human efforts alone

Additionally, omnichannel retailers tend to deal with higher volumes of inventory and wider sales and shipping ranges, requiring complicated mathematical algorithms for demand forecasting.

Without appropriate demand forecasting tools and the ability to process and accurately compute this level of complexity, omnichannel retailers may treat e-commerce as a separate business instead of a separate channel.

Allocating inventory for omnichannel retailers is also a balancing act as they struggle to determine how much inventory to keep in storage for online sales, and how much to sell in stores.

For example, if an omnichannel retailer allocates too much inventory to brick-and-mortar stores and not enough for online sales; they’ll end up pulling inventory from store shelves to fulfill online orders.  And this only increases costly transportation expenses and causes fulfillment delays. 

What is needed is a unified demand forecasting inventory solution to meet customer expectations for all channels.

Unbalanced inventory between channels is inefficient and reduces GMROI

 

4. Cart abandonment rate

It would be convenient to think that etailers could use items added to online carts as a way to predict demand, but it would be misleading. 

Too often a prospective customer does not complete the transaction; abandoning items sitting in their cart. How frustrating for the etailer!

 

E-commerce shopper abandons full shopping cart

Baymard Institute studies calculated data from 41 different studies and found the average cart abandonment rate is 69.82%  

There may be an intention to buy items in carts, but the sale doesn’t materialize. 

This behavior could be an indicator of many things, maybe the price point isn’t right, product cannibalization among the inventory mix, or other reasons. 

 

5. Shipping costs

When it comes to fulfillment and returns, the expense of shipping items out to customers and getting them back in reverse logistics if items are returned is high. 

E-commerce retailers:

 

  • Pay to ship items directly to customers.
  • Pay again to transport and manage returns.
  • Offer same-day, next-day, and free shipping to stay competitive.

Speed and efficiency for shipping and returns are costly for etailers, but customers expect no less.

All of these challenges, if not well managed through e-commerce demand forecasting, cut into profit margins.

The question remains, how can etailers win at e-commerce forecasting?

 

How leading e-commerce retailers forecast demand 

Today’s leading etailers leverage AI-driven analytics to find opportunities in the challenges and yield success by harnessing their data.

They recognize that e-commerce is high-tech retail and therefore needs high-tech solutions. 

 

High-tech e-commerce business needs high-tech solutions for demand forecasting; as represented by an analog phone versus a cell phone

Armed with a highly accurate demand forecast, etailers can predict which goods are needed for which store locations and channels; ensuring customers get what they want while minimizing their e-commerce challenges.

When they transition their technology strategies towards AI-based predictive analytics to create an effective e-commerce demand forecast, they can drill down to channel, store, and SKU levels with specificity and clarity.

 

Overall, an advanced analytics platform helps e-commerce retailers:

  • Plan confidently with an accurate demand forecast
  • Know what to stock, where, and when to maximize margins
  • Set optimal prices for every SKU at every location

Advanced analytics also helps with their common challenges in specific ways.

 

7 ways AI-based analytics improves e-commerce demand forecasting 

Etailers benefit from AI-powered advanced analytics because it can handle and dissect enormous amounts of data and automatically make recommendations.
   

And that’s not all, with AI-driven analytics etailers get:    

 

1. Greater accuracy with quality tools 

AI-powered predictive analytics solutions are developed to address the challenges of modern retailing. 

It’s the most advanced technology that has the capacity to store, organize, and analyze mass amounts of data that flows through e-commerce retail to increase:

 

  • accuracy
  • efficiency
  • visibility across the business

As a powerful tool in the hands of etailers, it eliminates the headache of manually analyzing and reconciling large, complex spreadsheets to optimize business.

And because this solution is automated, it reduces the time and cost of demand planning while reducing human error. 

 

Man assesses multiple data points for e-commerce demand forecasting based on AI-driven analytics

The more accurate the e-commerce demand forecast the better etailers can identify opportunities for optimization, consider next-steps recommendations, and engage in better planning for a more profitable sales cycle.  

 

2. Set dynamic competitive pricing 

Dynamic pricing is a strategy that uses big data and AI to automatically change pricing for products after advanced analytics carefully analyzes current pricing trends and competitor prices. 

In fact, some e-commerce giants use dynamic pricing, and they change the prices of their products in real time (for different people in different locations). 

For example, ride-sharing giant Uber uses a dynamic pricing strategy through which it controls the prices that customers pay. 

As a result, Uber can adjust prices frequently–up to every five minutes–when surge prices are in use. To reach an optimal fare, Uber has to access an enormous amount of real-time data.

Wal-Mart and Amazon are examples of etailers using dynamic pricing models. 

Offering competitive prices to customers results in increased e-commerce revenue as customers get competitive prices moment by moment.

 

3. Get a holistic view of omnichannel retailing

A big mistake some omnichannel retailers make is treating e-commerce as a completely separate business that requires separate tools and novel approaches. The reality is, e-commerce is simply another channel.

Yes, e-commerce has nuances that other channels don’t (and it pays to be mindful of those nuances). But it is more similar than different, as you still have to order inventory people want, set prices to maximize your profits, replenish your DCs like they were stores and warehouses, etc.

Many savvy omnichannel retailers have already figured this out — and they’ve unified their analytics across their channels using AI. This allows them to have one, clear view of their entire business — while automatically accounting for all of the complexities of each channel.

In other words, omnichannel retailers can now see the trees without losing sight of the forest.

 

Forest with trees indicating various channels for omnichannel retailers with both brick-and-mortar and e-commerce channels

The ability to see and understand omnichannel demand forecasting for all channels of the business and as a whole removes siloes and improves: 

 

  • communications between and across all channels.
  • inventory management, planning, and operations. 
  • accuracy and reliability of the demand forecast.  

 

4. Flag reasons for cart abandonment 

A robust analytics solution that dissects and analyzes big data gives etailers real-time insights into customer online shopping habits and flags probabilities for cart abandonment. 

As a result, etailers can make critical business decisions faster from data based on:

 

  • competitive pricing data
  • individual and mass shopper behavior
  • checkout behavior
  • patterns in the sales funnel
  • threats and anomalies
  • SKU behavior

A prescriptive analytics solution can even provide optimal next-step recommendations to mitigate lost sales due to cart abandonment.

 

5. Improve fulfillment and returns 

Happy customers visit sites again. That means making sure they are satisfied with rapid order fulfillment, but also dealing with the necessary evil of returns. Etailers can use AI-powered software to:

 

  • Speed up delivery  

An accurate AI-based demand forecast quickens fulfillment. 

Products are brought to the best storage location in advance of online orders. This leads to quicker delivery options–even next-day delivery. 

Omnichannel retailers with a unified inventory demand forecast can offer more fulfillment options such as BOPIS. (buy online pick up in-store)

 

  • Reduce shipping costs 

AI-based demand forecasting software can identify batching opportunities. When several items are grouped by location and time, it mitigates shipping costs by cutting down on excess transportation and storage costs.

 

  • Find gains on returns

With an AI-powered analytics forecasting tool, e-commerce retailers can quickly identify where there is demand for returns; and send (vetted online) returns for re-selling online or to stores where it’s wanted.  

 

6. Create a desirable personalized online shopping experience

 

  • Targeted product mix  

Considering geodemographics, customer shopping history, and other available data enables AI-driven demand forecasting to determine what product mix should appear on the first few pages of an e-commerce website. 

The first few pages are valuable “real estate” that should promote inventory with the highest demand.
  

 

  • Upselling 

With good data and AI-based analytics e-commerce retailers can identify the best product pairing offers. E-commerce retailers can effectively upsell to specific customers by finding new patterns and relationships in the data.  

 

Upselling products using AI-based analytics to identify the best product pairing offers for a demand forecast

Retailers will be able to proactively get smart promotional and customer-specific recommendations.  

 

7. End-of-life product options

Deciding on what to do with end-of-life products can be difficult. 

For example, when an omnichannel retailer is left with 10 widgets at the end of the sales season, they have to decide what to do with them. Should they pull them from the shelves? Make them available online only? Bulk them all in one store location?  Place one widget each in 10 separate locations? 

AI-based analytics can identify the best options right down to SKU and location levels. That means omnichannel retailers can get the highest rate of return for their end-of-life inventory.

So, a widget with demand online is not languishing on a shelf somewhere. 

 

E-commerce retailers win with AI-powered analytics for demand forecasting

When etailers need accurate, dependable demand forecasting, AI-driven analytics software is a promising tool that offers solutions to challenges. 

Having the right AI-driven analytics demand forecasting tool drives more evidence-based decisions and positions etailers to gain better insights into consumer demand to align their assortment, merchandising, and promotion strategies accordingly.

This unified e-commerce platform enables them to maximize opportunities, revenues and profits while reducing costs. 

Are you an etailer looking for the right demand forecasting solution for your business? Let our team help.

At Retalon we provide software for predicting the future, automating high-value tasks, and fixing your entire commerce process—from factory to consumer.