The BBQ and outdoor eating category are highly seasonal, with a sharp rise in Customer Purchase Intent between February and April worth £3.3M per week in market growth.

Using Purchase Predictor retailers can identify the exact product mix and distinct customer missions happening in their categories, pinpointing the exact date when Customer Purchase Intent starts to grow (10^th February), the size / speed of the growth (10.3% avg. growth per week) as well as the impact of key trading events such as Easter and Bank holidays. Even more valuable are the insights which highlight which sub-groups and products are driving the growth and at which point in the season.

Using Purchase Predictor we can identify:

• Demand for key products within the ‘BBQ’ and ‘Alfresco Dining’ mission grows by 319% between the beginning of February and end of May
• ‘Gazebos & Parasols’ are the most important with the largest share of Customer Purchase Intent (35% of the category) in the run up to Easter
• ‘BBQ’ growth accelerates after Easter, with the fastest growth rate (+36%) into Mayday Bank Holiday replacing ‘Gazebos & Parasols’ as the category with the largest share of Customer Purchase Intent in the category (34%)

This shows how quickly customer priorities can change, and marketeers must adapt to these changes if you’re to grow sales and win market share. Purchase Predictor even enables you to drill down to a product level to understand the growth in Customer Purchase Intent likely around key events.

Yes – we forecast a 5-7% growth in Customer Purchase Intent (worth £100m to the UK cleaning market). The Customer Purchase Intent spikes across two key missions;

1. The “Make it Sparkle” mission peaking in February (week commencing 10^th), including floors, windows and upholstery cleaning equipment
2. The “Deep Clean” mission peaking in April week (week commencing 14^th), including carpets, furniture & appliances

With this insight retailers can better time promotions, giving customers the best offers, when they most want them; and maximise sales from featured space by placing the most in demand items at front of store and on gondola ends (and online equivalents). Category managers can go one step further using Summit’s ‘headroom methodology’ pinpointing category revenues left on the table during last year’s trading. This will help in building a business case for budgeting.

As you can see in this case study – revenue is at its lowest point for products included in the ‘Make it Sparkle’ mission when Customer Purchase Intent is at its highest. Based on proprietary methodologies we have been able to calculate that the retailer should have yielded £165K more sales in the week commencing 9^th February, with even more revenue left on the table week commencing 24^th February. It’s a slightly different story for the ‘Deep Clean’ mission, where the retailer is doing a good job driving sales during the April peak in Customer Purchase Intent but has missed a golden opportunity to drive sales week commencing 17^th February.

In a very practical way, the headroom calculation enables retailers to see the amount of money they have left on the table in each category in each week. Therefore, Purchase Predictor enables categories to identify the one week where it will be easiest to drive incremental sales. To capture this headroom, the category manager could:

• Run a multibuy promotion on household cleaning products from week commencing 10th February, with support in a weekly flyer or circular
• Utilise seasonal space with specialist cleaning products in February and early April

The Science behind Purchase Predictor

Purchase Predictor makes it easy for retailers to plan promotions more effectively and maximise profits. But behind the simplicity of the platform, there’s a world of complicated data science powering the insights. Processing and analysing millions of data points across 40,000+ products, in more than 10,000 categories, the platform uses machine learning to identify and cluster customer behaviour patterns. Digital fingerprints are also analysed to understand Customer Purchase Intent. This data is then used to predict future behaviour through a Bayesian statistical modelling system. Purchase Predictor decomposes data in order to identify long-term trends, underlying seasonality and the effects of key marketing dates or events. This enables the platform to forecast behaviour and demand up to three years out, using linear regression analysis to understand areas of growth and decline. The platform’s recommender engine is able to identify the most appropriate activity and focus product categories for each week. Taking the retailer’s last year of sales data, the engine identifies incremental sales revenue opportunity using headroom analysis based on trends and customer behaviour. Crucially, this is used to provide a monetary value on the identified potential sales opportunity, allowing retailers to make planning decisions backed by statistics.

*http://nationalbbqweek.co.uk/facts/