Using AI to predict stock shortages

– Originally written by Veena Giridhar Gopal, Founder of salesBeat –

AI driven platform is pioneering sales intelligence in FMCG

If you have ever done a search on Google for sales intelligence for FMCG companies, you’ll find several academic articles on how AI can benefit sales teams in FMCG companies, along with links to several companies that provide AI driven sales intelligence to any other sector than FMCG.

And if you search for demand planning solutions there are an abundance of them, those that use AI and include seasonality, holidays etc, to provide forecasts. Ironically demand planning is not used by sales teams. In fact, one of its input factors is a sales forecast from the sales team!

So who uses Demand planning software?

The supply or production planning teams typically use demand planning software to ensure there is enough stock of the final product, so that they can fulfil orders and meet demand. Typically, the annual sales plan or the joint business plan drawn up/agreed by the sales team is one of the inputs into the demand planning software/application.

According to an article by Gartner (Aug 2019), demand planning is typically used by the Sales & Operations Planning team (S & OP team) and focuses on a tactical horizon from three to 24 months.

The primary goal of demand planning is to ensure availability of stock on the supply side. The process results in a plan that feeds into the Sales & Operations Planning meeting that agrees on the stock levels that need to be maintained at a company warehouse level for each SKU to be able to meet demand for the period being planned.

Now here is why demand planning is not the same as sales intelligence. Have you ever planned your grocery list three months in advance? Also, have you ever known a supply team to influence a customer’s decisions?

Sales intelligence in FMCG

Even a few years ago, sales intelligence in FMCG would have been complex and time taking. Insights teams would look at similar information as marketers do, when considering innovations or ad campaigns. However, due to the sheer volume of information, and the time it takes to process the information and gain insights, this happens only for new product development or for market entry initiatives.

Due to recent advances in AI, it is now possible to build applications that process the same quantum of information on a daily basis to provide valuable insights to sales teams, helping them take advantage of opportunities that were never identified by the strategy team or by the demand planning application.

Sales intelligence apps are meant to be used by any sales team who sell to customers, at a national account level and at a store level.

Economies of scale are critical to success in the FMCG sector. When the production planning team plans production, they need to ensure that the production runs maximise equipment capacity. So they plan three to 12 months in advance. Most companies plan three to four months ahead and maintain enough stock to cater to demand.

Demand planning is based on the assumption that sale forecasts are a result of ‘push’ tactics, that current year sales reflects prior year sales plus an uplift under ordinary conditions, adjusted for seasonality and other medium term factors.

Why is sales intelligence becoming increasingly important?

On a broad basis this still holds true. However, these days, the sales forecasts that go into the demand plan are no longer reflective of expected sales. So there are short term fluctuations within medium term forecast numbers. Unless these short term fluctuations are tightly managed, there are frequent stock outs of brands at supermarkets.

With Millennials and Gen Z now forming a significant percentage of the adult population, the influence social media and news have on how these two generations buy and consume products, sales intelligence apps and platforms are becoming increasingly important to help sales people understand how much to sell.

By empowering sales teams with information on what consumers want and are likely to buy in the next four to six weeks or even days, sales teams can ensure 100% in-store availability of the brands they sell. Providing timely sales intelligence helps sales teams take advantage of opportunistic upsides that are in line with long term strategy, leading to increased revenues of up to 40%.

AI predictor eliminates stock-outs and loss of sales

Since the COVID-19 pandemic, it’s inevitable that retail organisations need to be prepared for consumer stockpiling. The first national lockdown saw a nationwide shortage of toilet paper, hand sanitiser, pasta and many other goods which caused chaos across the country.

However, this is not the first time that we have witnessed stockouts, everytime a heatwave hits many shops run out of icecreams, cold beverages, BBQ ingredients and the like.

Humanity may face more lockdowns and without a doubt there will be other situations which encourage people to buy more or stockpile. We have hardly felt the true impact of Brexit and how that will affect our favourite products becoming unavailable due to stockpiling and stock-outs.

Technology is one way of eliminating this issue and maintaining steady stock levels by predicting consumer movements. AI can be cleverly designed to tell you what your customers want before they even want it!

Having seen the stock outs at the beginning of the pandemic last year, the salesBeat team launched Beat, an AI driven consumption predictor which works out consumer actions via numerous indicators; including but not limited to, information on air pollution levels, weather circumstances, water levels, consumer movements, current affairs, popular trends etc.

Currently sales are simply gauged on what was sold last year, which doesn’t take into account any important consumer factors, which can change vastly from day to day, let alone over a year. Companies within the FMCG industry use demand planning or financial forecasting tools which make longer term forecasts over 12 to14 weeks based on seasonality and what was sold last year. salesBeat makes precise predictions based over 4 to 6 weeks, making it the best consumption predictor available.

This is a great tool for sales representatives and buyers within the FMCG industry. It helps eliminate waste as well as eliminating stock-outs and therefore will not only save businesses money but will also improve on profit margins.

Why is this a great solution now?

We didn’t have the technology or resources available before and now it’s available for everyone! You don’t need to hire an expensive data science team to do incredible things in tech.