EXPERTISE / DISTRIBUTION

The sector of distribution has given us many challenges we have beenable to solve
challenge
1

Consumer goods: probability of recommending a brand

o assess what influences customers when recommending a brand, we were asked to develop a simulator that would allow us to calculate the probability of recommending a brand based on socio-demographic variables, brand image and levels of satisfaction.

Solution:

We applied logistic regression models to research data that tracked branding, buying habits, attitudes, consumer satisfaction levels, etc..

Actions carried out:

Data from a questionnaire directed at a representative sample of consumers were used, and metric variables for the application of logistic regression were created.

With all this information, a logistic regression model was built.

Results:

We created a final simulator that calculated the likelihood of brand recommendation in terms of socio-demographic characteristics, opinions and habits of the respondent, and which identified the "levers" to activate via marketing.

Simulator

Simulator
return to
top
challenge
2

Distribution: geo-marketing and POS segmentation

A major beverage brand wanted to segment 17,000 retail outlets in the FOOD channel into 4 categories (IMAGE, RICH, CONVENTIONAL, AND POPULAR), according to the socio-demographic and socio-economic profile of the premises.

Solution:

We analyzed data from the 35,000 Spanish census tracts and developed a geo-marketing application to define the areas of influence of each retail outlet and the characteristics of its environment.

Actions carried out:

A mapping of retail outlets was carried out and, using the Huff method, the areas of influence of each one of them was defined (based on its size and its distance from other outlets).

We identified the socio-demographic and economic profile of each outlets environment.

Results:

We succeeded in segmenting the 17,000 retail outlets into the 4 categories, with a cluster analysis at national, regional and provincial level, allowing our client to optimize distribution and trade marketing activities.

return to
top
challenge
3

Retailer: Sales analysis by product categories

A leading retail company wanted to quantify the contribution of the main key drivers influencing sales in each of their product categories.

Their objectives:

  • Determine media mix effectiveness by product category.
  • Quantify external factors affecting sales.

Solution:

We applied descriptive analysis in order to identify and classify variables that might affect sales of the different product categories.

We developed explanatory econometric models to test media mix and other variables previously identified.

Actions carried out:

From the models: The recommended optimal effectiveness was calculated for each category.

The contribution of other variables (intangibles, web searches and seasonal factors) with different effects in each category was quantified.

Results:

It was successfully demonstrated that the effectiveness of the media mix was different in each product category.

Using effectiveness response curves, the advertising saturation point for each of the categories was identified.

Sales analysis by product category

Ejemplo driver ventas
return to
top
challenge
4

Distribution: Customer Segmentation

Consumers have certain common needs and attitudes. This allows us to segment them into different interest groups for companies.

Our client had a classification of its buyers, but it was made up of too many segments (15) rendering it too unwieldy to establish action plans. Therefore, they needed to reduce the number of segments.

Solution:

Information relating to the purchase ticket of a group of customers, already segmented, was collected and a survey about uses and attitudes was launched to complete the investigation.

From these data, the segments were grouped based on qualitative and quantitative criteria and multivariate techniques.

Actions carried out:

It was intended that each group should have similar characteristics in terms of lifestyle, ticket and purchasing behavior.

Furthermore, a representative percentage of customers should be grouped, having similar expenses and purchase frequency.

Finally, multivariate statistical analyses were performed to create factors that would maximize the homogeneity within a group and that would ensure the heterogeneity with the other groups.

Results:

We managed to regroup the initial 15 segments into 6 more closely related segments. The new segmentation offered certain advantages over the former one: greater operability, greater homogeneity and greater business opportunity.

Client segmentation

client segmentation
return to
top