Opinions expressed on this blog reflect the writer’s views and not the position of the Capgemini Group

Forecasting in a multi-digi-social-augmented channel world

In my previous blog I discussed the opportunity for retailers to increase store footfall through using Pokémon Go as a channel to market.  I touched upon the difficulty retailers face through the many channels they can now interact with their customers. As retailers begin to broaden their reach across these channels, it can become very challenging to forecast sales volumes as customers tend to switch and flow between channels of convenience.

In this article I explore the implications different channels have on a retailer’s forecast and look at how to build an effective forecast in this new retailing age.

Change forecasting

We live in a world where change is constant. We naturally endeavour to improve our lives, and consequently this drives new consumer offers, that in turn change our shopping behaviours. As a result retailers need to continuously refresh their range to keep up with the latest consumer trends and frequently bring in new products. With this comes a need to forecast the right buy quantities to ensure products are available to the customer at the right place, the right time and in the right quantity.

Knowing quantities to forecast for new product lines is not the only challenge. There are now so many influencers that impact sales with weather, media, showbiz, Twitter and Facebook just to name a few - all of which can sway a customer’s buying decision. Coupled with this, customers now also have the ability to read product reviews and compare prices on their mobile device. They can do this whilst standing in front of the product in-store, helping them make an informed decision.

On top of these challenges, retailers must also think multichannel (stores, online, mobile, click & collect etc). It’s not enough to forecast a total volume, but also to plan how to distribute that volume across channels to ensure availability where customers demand it. A retailer’s own infrastructure can strain its ability to get this right, variables such as physical capacities, minimum credible displays, purchase quantities and speed of network are just a few of the constraints a merchandiser has to work within; it becomes quite a job!

Data spoilt

Now with multiple channels to market and the continuous evolution of technology, retailers are spoilt/ drowned in customer data. Not only is rich data available by selling channel, but also from other sources to identify future consumer trends or specific customer buying habits, for example, social media listening or customer loyalty schemes.

As retailers continue to innovate their products and services, they begin to find new ways of interacting with their customer base. This also enables them to collect more specific data to help them forecast customer buying needs and buying behaviours. The new service launched in the UK by Amazon called Amazon Dash, is a good example of this. The service allows customers to plant mini ‘Dash’ devices around their home close to frequently purchased products. The devices are linked to the customer’s internet connection and when pressed will re-order the product automatically for the customer without needing to log in to their account.

Amazon Dash offers ease and convenience to ordering in real-time. Not only does this guarantee the sale for Amazon, but also provides detailed data points specific to that customer. These new data points could be used to create customer specific forecasts, predicting when the customer is likely to run out of a particular product (thus likely to order) and ordering a new one automatically – what if your fridge could automatically re-order your food for you, when you needed it?

Using all this new data in the right way can be very powerful, helping to improve forecast accuracy and ultimately customer satisfaction through ensured product availability.  The down side to this rich data now available to retailers is that it can be even harder for them to separate the relevant from the irrelevant [the ‘wood for the trees’], thus creating a cumbersome task without sophisticated analytics and clean input data.

Playing catch up

Technology is moving fast and retailers need to move quickly to keep pace as customers look for the next innovation to make their everyday shopping easier. As retailers start to adapt their front end customer interactions (e.g. in-store apps, delivery tracking) to the latest technology, they continue to put strain on their existing core merchandising system. These are the work horses of the retail business and been operating for many years prior to the invention of the smartphone.

These systems make it hard for retailers to be agile in a world of evolving consumer trends and buying channels. Instead, it restricts their ability to analyse data effectively and build a forecast to serve their customers. For most retailers, they will need to decide what to do about their core merchandising system in order to ready themselves for future growth. Can they afford not to?

Quick wins

Realistically, heavy investment in a sophisticated planning system is not a feasible option for most retailers. There are, however, a few quick solutions to help support better forecast accuracy:

  • Breakdown the walls – promote active engagement across departments, removing the silos and getting different parts of the business to share insight. Zara does this well; they establish a clear feedback loop between stores and head office to share information on what customers are buying and what trends they are looking for.


  • Get analytical – retailers have vast repositories of rich customer data. Use the reporting tools already available, upgrade the logic/ analytics, to group and manipulate data in to specific reports with actionable insight e.g. group and calculate trends across product hierarchies to inform buy or sale profiles for new lines.


  • Understand the small events – a better forecast is a consequence of a better understanding of demand drivers in the past. Capturing the smallest event in the past can help de-mystify the reason for sales and help predict events in the future.


  • Include social – leverage social media to capture what customers are saying. Integrate a social media team within your buying and merchandising function to not only enable better forecasts upfront, but also to shape your demand through active listening to trending behaviours.


  • And finally, know your levers – using levers such as pricing and promotions to influence your sales volumes will also influence your forecast in a big way. Putting a product on promotion does not necessarily mean an increase in sales, but maybe a shift in sales from one time period to another. Being sure merchandisers are aware and therefore understand the impacts of their actions will be critical to shaping an accurate forecast.


So, change doesn’t have to be expensive, but one thing is for certain, and that is that change will continue and customer behaviour will become more complex. As a result, forecasting will become trickier and even more critical to the success of any retail business. Doing nothing is no longer an option.

About the author

Chris Long
Chris Long
Chris is a Senior Retail Consultant at Capgemini having joined from industry as an Operations Manager. He has experience in delivering complex programmes across the end to end retail supply chain, specialising in demand driven merchandising and store operations.

Leave a comment

Your email address will not be published. Required fields are marked *.