Vispera - Image Recognition Solutions for Retail


One of the global confectionery manufacturers visited us to seek an innovative solution to track the performance of their retail execution regarding the chocolate and biscuit categories in modern channel. 

Their existing solution was based on having hundreds of field agents to track, monitor, and replenish the retail displays on a periodic & reactive manner. They have realized that visual assessment and manual reporting of SKU counts, facing & shelf shares and planogram compliance is very open to human errors especially in categories like biscuit and chocolate which consist of at least 1000+ items per shelf on average. Moreover, the collected data was not properly stored in a database to build an institutional memory and to serve smart and actionable analytics. Last but not least, this operational time was stealing from the time to be spent to ensure customer interaction and satisfaction in modern channel. 

We ran a pilot project during August -November 2016 where 7 stores of a national retail chain account (X) and 5 stores of a local modern trade account (Y) have been visited periodically. The following results which allowed the company to take immediate actions, were identified and reported by Vispera Image Recognition Service. 

Reported Output 1

The comparison of KPIs (Planogram Compliance, Facing Share and on-shelf availability scores) for 2 different customers were shared (Figure 1). 

Actions Taken

  1. Both Key Account Managers were able to calculate the potential loss due to out-of-stock and share with top management. Accordingly, they have started taking preventive actions to increase on-shelf availability scores to above 90%. 

  2. Based on the list of Most out-of-stocked SKU lists, an internal analysis was done to understand the root cause of not being on shelf (due to production, distribution or customer).

  3. Regional Sales Manager set one-on-one meetings with field executives to discuss the reasons of low planogram compliance scores and informed them that their incentives would be given based on the overall store scorings.


Reported Output2

A dashboard consisting of overall results of all visits, high performing stores, low performing stores, most out-of-stocked SKUs and the frequency of stock-out information was shared (Figure 2). In addition to this, the visited stores were shown on a color coded map depending on the overall KPI scores (Figure 3).

Actions Taken

Sales Director was able to take regular snapshots of the field via dashboard and of the complementary map showing all store scores to have a general overview of the field execution performance.


Reported Output3

The trend of KPIs per store was shared (Figure 4).

Actions Taken
Sales/store manager responsible for several stores was able to analyse the trend of KPIs and take preventive actions to keep them at desired levels by holding field executives accountable.

Vispera Image Recognition Service provides their customers with more accurate field execution scoring powered by actionable analytics. This allows different levels of managers to take relevant and immediate actions within their responsibility areas. Moreover, Vispera Image Recognition Service reduces the not-value-added time spent in the store to complete visual assessment and manual reporting and allows more and value-added time to invest in customer satisfaction and engagement.




Inconsistent data 

Inaccurate scoring 

Data not transformed into actions 
Accurate Scoring 

Actionable Results