際際滷

際際滷Share a Scribd company logo

Marketing Experiment Assignment - Christian van 't Hof

Christian van 't Hof
Christian van 't Hof

This document outlines a marketing experiment for Viking Beds of Sweden to test two TV advertisements. The experiment will use a before-after design in the Dutch cities of Laren and Naarden, which have similar populations and incomes. Both cities will see Ad 1 in the first period and in the second period Laren will see Ad 1 again while Naarden sees a new Ad 2. Sales data will be analyzed to see which ad increases sales more and if either city responds better to the ads. Potential issues like seasonal spending are also addressed. An adaptation using online ads and a full factorial design to study gender differences is proposed.

1 of 2
Downloaded 27 times
Student: Christian van t Hof 11th of January 2016 Marketing Experiment Assignment 1 Introduction This report is about a marketing experiment for Viking Beds of Sweden. This Swedish boxspring brand has just been introduced on the Dutch market in the upper-middle and higher segment. The marketing will be a TV advertising campaign for their boxspring type Royal Viking. Since boxsprings are gaining more market share, these high quality beds are being the real deal. The experiment that is going to run will be a before-after design experiment. Since there hasnt been any brand exposure yet, a basic experiment wont be working because the city with the TV ads will always be selling more. Moreover, a full factorial experiment wont be working either, because its about 1 product with the same pricing for both cities. In this case, the before-after design experiment will be working by showing two different ads. 2 Experiment Design The variables in this experiment will be independent as well as dependent. The dependable variable will be the amount of units sold and the independent variable will be the TV advertisements. This experiment will run in the cities Laren and Naarden in the Netherlands. These cities are both in the same geographical area, share almost the same amount of population and have both on average about 50,000 income per household1, which is almost the highest in the Netherlands. Both Laren and Naarden have above 10,000 inhabitants, which is enough for statistical significance. The entire cities will see that TV ads, so this makes this a valid experiment. The first round of TV ads will run from March till April and the second round of TV ads will run from May till June. In the first round, both cities will see TV Ad 1. In the second round Laren will see TV Ad 1 and Naarden will see TV Ad 2. The lift in sales will be calculated from the differences between the two periods. In terms of causality three rules will at least be applicable, the other one is in this case not applicable. The two applicable rules are the first, second and the third one: Change in the marketing mix produces change in sales, no sales increase when theres no change in the marketing mix (Round 2, Laren) and the time sequence. In this high-end market, competitors will probably respond, so that will be the external factors. 1 Centraal Bureau voor Statistiek. (2015, December 8). StatLine - Centraal Bureau voor Statistiek. Retrieved
Student: Christian van t Hof 11th of January 2016 3 Anticipated Issues 3.1 - 2 Issues to be considered In this experiment there could be issues. In the Netherlands people get around May and June, their holiday fee from their jobs, this could increase their budget in those months, which could be an external factor that could affect the net lift. Another issue could be our competitor Carpe Diem. Carpe Diem is also from Sweden and has already entered the market. If they launch a new campaign, than this would definitely affect the net lift as well. 3.2 - 2 Things the experiment demonstrates This experiment could point out a difference in outcome of both TV Ads. It could indicate which TV Ad would work better. On the other hand, a result of this experiment could also indicate which city would be buying more products from our company compared to the other. If the outcome of both cities shows very different numbers, the most interesting city could be appointed. 4 Experiment Adaption In this 2nd version Viking is going to show their TV Ads on the Internet. This campaign will span in the same months in the second year. In this case, a full factorial experiment will assist in providing more specific metrics, which can be controlled more precise. In this case the video ads will be an independent variable as well as gender. According to metrics Sweden, males tend to buy these boxsprings over females. The dependent variable will still be the amount of units sold. This more specific experiment will show more relevant information on the gender of our customers, which we can use to improve the targeting of customers by more specified videos that will be more focussed on the differences in gender. The results from this experiment will be easier to measure and to analyse, which saves the company money and gives the company better results. On the other hand, just like TV ads, it only measures the people that use the Internet on their computer or watch TV.

More Related Content

Marketing Experiment Assignment - Christian van 't Hof

  • 1. Student: Christian van t Hof 11th of January 2016 Marketing Experiment Assignment 1 Introduction This report is about a marketing experiment for Viking Beds of Sweden. This Swedish boxspring brand has just been introduced on the Dutch market in the upper-middle and higher segment. The marketing will be a TV advertising campaign for their boxspring type Royal Viking. Since boxsprings are gaining more market share, these high quality beds are being the real deal. The experiment that is going to run will be a before-after design experiment. Since there hasnt been any brand exposure yet, a basic experiment wont be working because the city with the TV ads will always be selling more. Moreover, a full factorial experiment wont be working either, because its about 1 product with the same pricing for both cities. In this case, the before-after design experiment will be working by showing two different ads. 2 Experiment Design The variables in this experiment will be independent as well as dependent. The dependable variable will be the amount of units sold and the independent variable will be the TV advertisements. This experiment will run in the cities Laren and Naarden in the Netherlands. These cities are both in the same geographical area, share almost the same amount of population and have both on average about 50,000 income per household1, which is almost the highest in the Netherlands. Both Laren and Naarden have above 10,000 inhabitants, which is enough for statistical significance. The entire cities will see that TV ads, so this makes this a valid experiment. The first round of TV ads will run from March till April and the second round of TV ads will run from May till June. In the first round, both cities will see TV Ad 1. In the second round Laren will see TV Ad 1 and Naarden will see TV Ad 2. The lift in sales will be calculated from the differences between the two periods. In terms of causality three rules will at least be applicable, the other one is in this case not applicable. The two applicable rules are the first, second and the third one: Change in the marketing mix produces change in sales, no sales increase when theres no change in the marketing mix (Round 2, Laren) and the time sequence. In this high-end market, competitors will probably respond, so that will be the external factors. 1 Centraal Bureau voor Statistiek. (2015, December 8). StatLine - Centraal Bureau voor Statistiek. Retrieved
  • 2. Student: Christian van t Hof 11th of January 2016 3 Anticipated Issues 3.1 - 2 Issues to be considered In this experiment there could be issues. In the Netherlands people get around May and June, their holiday fee from their jobs, this could increase their budget in those months, which could be an external factor that could affect the net lift. Another issue could be our competitor Carpe Diem. Carpe Diem is also from Sweden and has already entered the market. If they launch a new campaign, than this would definitely affect the net lift as well. 3.2 - 2 Things the experiment demonstrates This experiment could point out a difference in outcome of both TV Ads. It could indicate which TV Ad would work better. On the other hand, a result of this experiment could also indicate which city would be buying more products from our company compared to the other. If the outcome of both cities shows very different numbers, the most interesting city could be appointed. 4 Experiment Adaption In this 2nd version Viking is going to show their TV Ads on the Internet. This campaign will span in the same months in the second year. In this case, a full factorial experiment will assist in providing more specific metrics, which can be controlled more precise. In this case the video ads will be an independent variable as well as gender. According to metrics Sweden, males tend to buy these boxsprings over females. The dependent variable will still be the amount of units sold. This more specific experiment will show more relevant information on the gender of our customers, which we can use to improve the targeting of customers by more specified videos that will be more focussed on the differences in gender. The results from this experiment will be easier to measure and to analyse, which saves the company money and gives the company better results. On the other hand, just like TV ads, it only measures the people that use the Internet on their computer or watch TV.