In this article we will take a look at the results that modelling can achieve, which ultimately will impact sales. The example below will help give a better overview of how this works in practice. Modelling can be an excellent basis for decision-making and help with budget allocation for companies in the online as well as offline sector. The re-distribution of budget will then be based on statistical facts rather than gut feel.
QUISMA has used sales modelling for a number of customers, who will then base their budgetary decisions on the results produced by the marketing intelligence unit. Resulting in higher sales, a higher turnover, lower CPOs and lower marketing ratios.
An example from the clothing industry:
QUISMA has been managing all the online activity of a leading German fashion store since the beginning of 2009. The services include year-round measures to increase awareness and sales as well as complete media planning, including optimal management of the marketing budget.
QUISMA used sales modelling, namely regression analysis, to manage the budget. Sales were modelled as the result of customer activity. With spend being distributed between the following channels: affiliate marketing, display advertising, retargeting, search engine advertising and search engine optimization. To ensure optimal management of the budget, QUISMA evaluated the performance of the individual channels as well as the influence of the channels on the overall performance. These procedures made it possible to consider complete user paths and recognize cross-media effects between the channels when allocating the budget.
Modelling was conducted between January 2009 and June 2010, on the basis of which the effects of the channels on the total performance were determined, taking cross-media effects into account as well. Based on these results, the budgets for the last two quarters of 2010 were distributed anew and the campaigns for every single day were optimised in terms of sales and marketing ratio.
The results achieved through modelling are remarkable. A comparison of the last two quarters of 2009 and 2010 reveals that sales increased by 24 percent while the marketing ratio fell by 43 percent. It should be noted that the budget during these time periods were almost identical.
Modelling does not evaluate channels individually, but rather looks at them in their overall media context therefore the control lever can be better detected and applied. This holistic approach shows the essential advantage over conventional methods such as the customer journey analysis. These differences will be explained in more details in the coming blog entries.