Retail Execution
Understanding the term "Retail Execution" in todays realities
Articles
The most important part of the Retail Execution takes place in the planning period.
Part 1: Meaning of “Retail Execution”
Clients do tend to ask the meaning of "Retail Execution".
The answers is — it is a predefined series of activities at point of sale to reach one single target, a "Perfect Store". It is long and tough journey, for that matter company will be needing a Retail Execution Software (RES).
RES is a combination of real-time Image Recognition and Retail Business Analytics. It is real-time solution alternative to the manual compilation of retail data for retailers and consumer packaged goods (CPG) companies. Data is collected as real-time snapshots at the physical outlets via mobile devices owned by merch/audit teams or fixed cameras.
Analytics are presented to the field team and the HQ in a few seconds in form of preinstalled reports, to take actionable decisions based on the KPI results as the field operation goes on real-time as well.
The answers is — it is a predefined series of activities at point of sale to reach one single target, a "Perfect Store". It is long and tough journey, for that matter company will be needing a Retail Execution Software (RES).
RES is a combination of real-time Image Recognition and Retail Business Analytics. It is real-time solution alternative to the manual compilation of retail data for retailers and consumer packaged goods (CPG) companies. Data is collected as real-time snapshots at the physical outlets via mobile devices owned by merch/audit teams or fixed cameras.
Analytics are presented to the field team and the HQ in a few seconds in form of preinstalled reports, to take actionable decisions based on the KPI results as the field operation goes on real-time as well.
Benefits from RES
TIME SAVING
RES’s biggest benefit is the time saved manually compiling data. Where it was once commonplace to gather data and compile it in a spreadsheet or label the shelf images manually by operators, eventually get the KPI results approximately 24 hours later; many brands now see this as inefficient — preferring to spend this time on other, more important activities like taking action based decisions on informative real-time data visible both at the sales point and the HQ.
RES’s biggest benefit is the time saved manually compiling data. Where it was once commonplace to gather data and compile it in a spreadsheet or label the shelf images manually by operators, eventually get the KPI results approximately 24 hours later; many brands now see this as inefficient — preferring to spend this time on other, more important activities like taking action based decisions on informative real-time data visible both at the sales point and the HQ.
HIGH ACCURACY
The second benefit of a RES is the ability to improve analytical accuracy, ultimately, your understanding of your real-time physical performance in categories across the physical world. Automated RES data is void of human error, providing you with your planogram compliance, prices and where your products sit in comparison to competitors'.
The second benefit of a RES is the ability to improve analytical accuracy, ultimately, your understanding of your real-time physical performance in categories across the physical world. Automated RES data is void of human error, providing you with your planogram compliance, prices and where your products sit in comparison to competitors'.
MORE DATA
Lastly, a RES empowers you to track the real-time physical performance and having real-time KPI results opens multiple doors such as getting snap Retail Execution decisions or taking automated orders. The fluidity of this data also makes it easier to spot trends, survey price changes and make data-driven decisions from the HQ.
Lastly, a RES empowers you to track the real-time physical performance and having real-time KPI results opens multiple doors such as getting snap Retail Execution decisions or taking automated orders. The fluidity of this data also makes it easier to spot trends, survey price changes and make data-driven decisions from the HQ.
Part 2: What is “Artificial Intelligence-AI”
Another popular question from potential clients is about the wording "Artificial Intelligence-AI" as a platform for image recognition in Retail.
There are hundreds of "physical shelf recognition" companies providing image recognition services with AI infrastructure of their own. But do they really? Word "AI" is a good slogan for any company to smarten and give mystery to their product thus increasing product value in front of their potential customers. The answer is simple "NO".
Most companies do not have an AI platform unless they provide the recognition in "real time" within the expected level of accuracy. Any type of delay in providing the KPI results of a visit, means they are using human operators to label each image in each photo. So it is more likely a "Natural Intelligence" platform.
There are hundreds of "physical shelf recognition" companies providing image recognition services with AI infrastructure of their own. But do they really? Word "AI" is a good slogan for any company to smarten and give mystery to their product thus increasing product value in front of their potential customers. The answer is simple "NO".
Most companies do not have an AI platform unless they provide the recognition in "real time" within the expected level of accuracy. Any type of delay in providing the KPI results of a visit, means they are using human operators to label each image in each photo. So it is more likely a "Natural Intelligence" platform.
What are the limitations of "Natural Intelligence" model?
DELAY IN RESULTS
First of all, delay in results cause incapability to take instant actions within a visit in terms of Retail Execution. The field team member should revisit the same store to be able to take actions based on the results as the KPI results are available in the future.
First of all, delay in results cause incapability to take instant actions within a visit in terms of Retail Execution. The field team member should revisit the same store to be able to take actions based on the results as the KPI results are available in the future.
LOW ACCURACY
Secondly, the results will be open to human mistakes. As the labeling of the visit photos are done by human operators in the shadows, practically there is no difference with collecting the data manually on the field, which both of them open to human mistakes with an average of 70%.
Secondly, the results will be open to human mistakes. As the labeling of the visit photos are done by human operators in the shadows, practically there is no difference with collecting the data manually on the field, which both of them open to human mistakes with an average of 70%.
HIGH COSTS
Thirdly the cost of the service, an AI platform should be less costly as being an automated system, but using human operators brings a huge constant cost for the provided service.
Thirdly the cost of the service, an AI platform should be less costly as being an automated system, but using human operators brings a huge constant cost for the provided service.
Part 3: Why “Retail Execution” is important
The answer is simple: "Shoppers can easily be influenced, thus may change their mind at the moment of purchase." Consumer goods companies spend an average of 20% of their revenue with strategies and promotions to address that fact.
Companies Plan: Define planogram and promotion for each category and segment.
Execute: Maintain planogram and promotion compliance.
Audit: Check planogram and promotion compliance. Collect competitors information.
Companies Plan: Define planogram and promotion for each category and segment.
Execute: Maintain planogram and promotion compliance.
Audit: Check planogram and promotion compliance. Collect competitors information.
The most important part of the retail execution takes place in the planning period.
Alas for the planning, accurate and dynamic results from the latter are required to optimise and revise the results. For the execution part, a real-time Image Recognition Solution is the key element for taking instant actions.
And the minimization of the time gap between the execution and the auditing is compulsory.
Alas for the planning, accurate and dynamic results from the latter are required to optimise and revise the results. For the execution part, a real-time Image Recognition Solution is the key element for taking instant actions.
And the minimization of the time gap between the execution and the auditing is compulsory.
Part 4: Methodologies used to collect KPI Data
What are the methodologies to collect KPI data by a field force at retail execution? This is the question which each CPG company should decide for their retail execution processes.
Many companies are already doing it for their products available at online shopping platforms. But it is easy, they just stream the platform data to their reporting infrastructure and create any report based on their desired KPIs because the data iS already in digital form. But for the retail execution throughout the physical outlets, the first issue is to transform the analog physical world into the digital data. There are ways of doing it and the companies shall choose one out of 4 different methodologies for their Retail Execution. And the main differences are:
1) Manual collection: Field force simply writes down the shelf metrics on a page and later at the office they are typed by operators into a database (or excel).
Accuracy: Open to human mistakes, average accuracy is around 70%
Cost: Consumes a big time of field force with an average of 1 hour for each visit.
3) Automatic collection with "Image Recognition-IR" solution supported by operators in the shadows: Field force simply takes photos of the physical shelves but the IR engine is not strong enough to reach to the desired accuracy, so the provider uses operators in shadows to increase the accuracy to 95% by correcting each photo manually.
Swiftness: KPI results are available hours later because of the hidden operators of the provider. Impossible to take instant actions during the visit according to the real time KPI results, should revisit the same outlet in the future again to take an action.
Accuracy: 95%
Cost: Consumes a big time of provider’s operators, thus the IR solution will be expensive for all traffic values.
4) Fully automatic data collection with "Image Recognition-IR" solution: Field force simply takes photos of the physical shelves. as the IR engine is fully AI based. the accuracy successfully reaches 95% without any operators in the shadows.
Swiftness: KPI results are available instantly. Possible to take instant actions during thevisit according to the real time KPI results.
Accuracy: minimum 95% at the beginning. Increases with the time, as the engine gets more experience by unsupervised machine learning during the usage.
Cost: Consumes negligible time of both the field force and the IR provider. Provides huge labor cost savings for both sides thus also the IR solution will be inexpensive, with a decreasing rate according to traffic values.
Ailet had chosen the 4th methodology as core business model and for the last 6 years implemented many successful projects together with the world’s leading CPG companies.
Many companies are already doing it for their products available at online shopping platforms. But it is easy, they just stream the platform data to their reporting infrastructure and create any report based on their desired KPIs because the data iS already in digital form. But for the retail execution throughout the physical outlets, the first issue is to transform the analog physical world into the digital data. There are ways of doing it and the companies shall choose one out of 4 different methodologies for their Retail Execution. And the main differences are:
1) Manual collection: Field force simply writes down the shelf metrics on a page and later at the office they are typed by operators into a database (or excel).
Swiftness: KPI results are available hours later. Impossible to take instant actions during the visit according to the real time KPI results
Accuracy: Open to human mistakes in 2 layers both at writing and typing stages, average accuracy is less than 60%
Cost: Consumes an enormous time of field force and the operators, the cost is the highest of all.
2) Manual collection using a sales force application: Field force simply fills the required fields or chooses from the multiple choices for the specific outlet at the application screen.
Swiftness: KPI results are available instantly. Possible to take instant actions during the visit according to the real time KPI results.Accuracy: Open to human mistakes, average accuracy is around 70%
Cost: Consumes a big time of field force with an average of 1 hour for each visit.
3) Automatic collection with "Image Recognition-IR" solution supported by operators in the shadows: Field force simply takes photos of the physical shelves but the IR engine is not strong enough to reach to the desired accuracy, so the provider uses operators in shadows to increase the accuracy to 95% by correcting each photo manually.
Swiftness: KPI results are available hours later because of the hidden operators of the provider. Impossible to take instant actions during the visit according to the real time KPI results, should revisit the same outlet in the future again to take an action.
Accuracy: 95%
Cost: Consumes a big time of provider’s operators, thus the IR solution will be expensive for all traffic values.
4) Fully automatic data collection with "Image Recognition-IR" solution: Field force simply takes photos of the physical shelves. as the IR engine is fully AI based. the accuracy successfully reaches 95% without any operators in the shadows.
Swiftness: KPI results are available instantly. Possible to take instant actions during thevisit according to the real time KPI results.
Accuracy: minimum 95% at the beginning. Increases with the time, as the engine gets more experience by unsupervised machine learning during the usage.
Cost: Consumes negligible time of both the field force and the IR provider. Provides huge labor cost savings for both sides thus also the IR solution will be inexpensive, with a decreasing rate according to traffic values.
Ailet had chosen the 4th methodology as core business model and for the last 6 years implemented many successful projects together with the world’s leading CPG companies.
Part 5: Importance of "Brand Blocking"
Why is "brand blocking" the secret weapon to sell more products in retail?
Brand blocking is the display of a brand all together on a gondola, aisle bay, end cap or even a wall shelf area.
It’s importance comes from the composition of the human eyes which are effective only to the straight forward central point, moreover the receptor cells which pick up colour and motion are also concentrated at the centre and both thin out towards the periphery.
So to sell a product, the shopper’s very narrow focal point should be catched. By grouping the products together, the target is to form a "brand block" and catch the attention, which actually "blocks" the shopper from skipping to other products.
As humans acknowledge the shelf from left to right scanning each row as reading the lines of a page, there are two ways to form it:
At Ailet, we provide the "brand block" results instantly for both horizontal and vertical formations, as soon as the photo of the shelf is taken. Merchandising teams can get instant actions according to the results which will be visible with the details on their mobile devices.
Brand blocking is the display of a brand all together on a gondola, aisle bay, end cap or even a wall shelf area.
It’s importance comes from the composition of the human eyes which are effective only to the straight forward central point, moreover the receptor cells which pick up colour and motion are also concentrated at the centre and both thin out towards the periphery.
So to sell a product, the shopper’s very narrow focal point should be catched. By grouping the products together, the target is to form a "brand block" and catch the attention, which actually "blocks" the shopper from skipping to other products.
As humans acknowledge the shelf from left to right scanning each row as reading the lines of a page, there are two ways to form it:
- Horizontal: If the brand is well known, horizontal formation is better because the shopper might stop searching further with reaching to the block, during the row scan.
- Vertical: If the brand is new to the market, vertical formation is better because the product is catched many times by the eye at each row scan. Retailers often choose this formation to drive the shopper to check the entire shelf from top to bottom.
At Ailet, we provide the "brand block" results instantly for both horizontal and vertical formations, as soon as the photo of the shelf is taken. Merchandising teams can get instant actions according to the results which will be visible with the details on their mobile devices.
Ertugrul Elmastas has been part of Ailet team since 2019. His focus areas include the Retail Execution and technologies such as Image Recognition, Machine Learning and Robotic Process Automation.
Since 2019, Ertugrul Elmastas has been part of Ailet Solutions team as Business Development Manager in Turkey.
Connect to our LinkedIn to get all the market insights and be informed about upcoming events in your region.
Since 2019, Ertugrul Elmastas has been part of Ailet Solutions team as Business Development Manager in Turkey.
Connect to our LinkedIn to get all the market insights and be informed about upcoming events in your region.
The contributor to the article
Ertugrul Elmastas
Ailet Turkey
Ertugrul Elmastas
Ailet Turkey
