Analytical CRM Solutions for Product Sales

Relying on our experience in international projects, the data mining background for a company’s product sales is constructed using our solutions as follows.


A core element of analytical support for product sales is the expected profitability model, which can be construed by combining product affinity scores (or the probability of a purchase) and the customer value effect (the business gain of a purchase).

By using uplift modelling, the extent to which a sales campaign can bring forth changes in individual customers’ purchase intent can be demonstrated.

When multiple marketing channels are being utilised, campaign channel optimisation can assist in defining the optimal pitching formats.

With the help of the above, the effectiveness rates of various ad-hoc and regular campaigns, whether outbound or inbound, can be improved tremendously, while also resulting in significant savings in campaign costs and optimising sales commissions for the possible highest profits.

A systematic evaluation of campaigns helps keep track of, and perhaps even intervene in, the implementation of the sales strategy.

Product Affinity Modelling

Product Affinity Models define the customer-specific likelihood of a product being purchased. Based on previous purchases and customer characteristics relevant before the purchase, these models yield a percentage indicator shedding light on the customer’s likelihood of purchasing the given product within the next defined period. This method can be put to work in cross-selling, service upgrade and upsell applications alike.

Price Elasticity Modelling

 Setting up a price elasticity model can be an effective supplement or even input to cross-sell and upsell models, simulating shifts in customer demand as a function of service/product price.

Customer Value Effect Modelling

Relying on pre-purchase customer characteristics, short-term profitability estimation is used to model the expected changes in customer value for the time a given product is purchased by the customer under review.

By mapping out the relevant customer lifecycles and modelling the possibility of shifts in between, long-term profitability simulation provides information on the expected long-term effects of product purchases.

Expected Profitability Score

By combining the product affinity score (the probability of a purchase) and customer value effect (business gain of a purchase) we can establish the expected profitability per customer-product pair. If this business potential is known for each of these pairs, comparisons can be made and products can then be ranked for each customer based on expected profitability. 

Campaign Uplift Modelling

Uplift Modelling is an analytical method whereby campaign efficiency can be further enhanced by means of additional filtering of the target group. Using past purchase data and relying on campaign results, the essence of this model is to distinguish, among potential buyers, those who can be convinced (the real target audience), those who are willing to purchase without a campaign (useful for lowering campaign costs) and those who are either inconvincible or are negatively affected by campaigning. As such, uplift models are a refinement to affinity models.

Campaign Channel Optimisation

For marketing activities taking place over multiple channels, companies can access their customers via different networks (small and large retailers, call centres, internet, text messaging, email etc.). This gives an opportunity to handle each consumer group separately based on needs and various characteristics.

Campaign Optimisation

Expected profitability models set up for customer-product pairs, as well as possible campaign uplift and campaign channel models, provide the basis for successful campaign management. Existing and potential customers listed in the system can be segmented into target groups or even different campaigns by means of scoring and alongside various other criteria.

Modelling can ensure the long-term optimisation of campaigns, both ad-hoc and recurring. The method is suitable for both optimising outbound campaign offers and appropriately serving inbound customers.

For the purposes of boosting the effectiveness of the sales process even further, our models can also be applied to optimise the sales incentive and commission system as well. That is to say, customer value effect modelling provides an excellent means for fine-tuning the commission structure used in the remuneration of sales personnel.

Campaign Evaluation

Following up the quality of sales activities and drawing the right conclusions are of vital importance for planning subsequent campaigns. Our effectiveness analyses and reports are always carried out by means of comparison with corresponding control groups. Our solutions allow sales campaigns to be monitored from various aspects:

With the effectiveness analysis of sales offers, we compare the purchasing habits and behaviour of pitched customers to those of members of an appropriately selected control group.

With the efficiency analysis of campaign targeting, we specifically gauge the performance of the models on the basis of which the target group has been assessed (affinity and uplift models).

Measuring the success rate of sales associates helps compare the effectiveness of campaign activities between individual associates.

The results of the efficiency surveys and comparisons are both made available to campaign managers and sales managers in the form of regular reports.



Antal Kerekes

Antal Kerekes

Partner, PwC Hungary

Gábor Oltyán

Gábor Oltyán

Senior Manager, PwC Hungary

Follow us