This demonstration conjoint analysis study of cereal is the outcome of a
lecture on market modelling by Charlie Nelson at Sydney University. Third
year marketing students participated in the study by:
1. specifying the attributes which were important to them;
2. observing the research design process; and then
3. acting as survey respondents (89 respondents).
A number of simplifications to the study design and execution were made in
order to complete the exercise within the lecture period.
Despite these minor limitations, the study and its findings are considered to
be of interest - as a demonstration of conjoint analysis generally and as an
indication of the preferences in breakfast cereal of third year university
students. In fact, despite the sample being small and not representative
of the overall population the market preference predictions are remarkably close
to actual market share data.
The attributes and levels for the study were:
| Attribute | Levels |
| Brand | Kellogg, Sanitarium, Uncle Tobys, No Frills |
| Price (relative to category average) | -30%, -10%, +10%, +30% |
| Sugar content | Low, High |
| Fibre content | Low, High |
| Fat content | Low, High |
| Toys in pack | Yes, No |
Bretton Clark software was used for the design and analysis. The design comprised 16 cards as shown below. In practice, these are on separate cards and respondents rank them. For this demonstration, students rated each of the 16 products on a 0 to 100 scale, where 0 is "no chance I would buy" and 100 is "definitely buy"
| Card | Brand | Sugar | Fibre | Fat | Toys | Price |
| 1 | Kellogg | high | high | high | no | -10% |
| 2 | Uncle Tobys | low | low | high | no | +10% |
| 3 | Sanitarium | low | high | high | no | -30% |
| 4 | No Frills | high | low | high | no | +30% |
| 5 | Kellogg | low | high | high | yes | +10% |
| 6 | Uncle Tobys | high | low | high | yes | -10% |
| 7 | Sanitarium | high | high | high | yes | +30% |
| 8 | No Frills | low | low | high | yes | -30% |
| 9 | Kellogg | high | low | low | no | -30% |
| 10 | Uncle Tobys | low | high | low | no | +30% |
| 11 | Sanitarium | low | low | low | no | -10% |
| 12 | No Frills | high | high | low | no | +10% |
| 13 | Kellogg | low | low | low | yes | +30% |
| 14 | Uncle Tobys | high | high | low | yes | -30% |
| 15 | Sanitarium | high | low | low | yes | +10% |
| 16 | No Frills | low | high | low | yes | -10% |
The chart below shows the relative influence on preference - and the strength of brands.
Several scenarios have been evaluated. The first shows brand strength when all else is equal. The second scenario incorporates realistic prices. The third scenario shows the potential of a low fat product with this market.
| No Frills | Uncle Tobys | Kellogg | Sanitarium | |
| Sugar | high | high | high | high |
| Fibre | low | low | low | low |
| Fat | high | high | high | high |
| Toys | no | no | no | no |
| Price (0% = average) | 0% | 0% | 0% | 0% |
| Preference share | 4% | 25% | 62% | 9% |
| No Frills | Uncle Tobys | Kellogg | Sanitarium | |
| Sugar | high | high | high | high |
| Fibre | low | low | low | low |
| Fat | high | high | high | high |
| Toys | no | no | no | no |
| Price (0% = average) | -15% | -5% | +5% | -5% |
| Preference share | 8% | 35% | 49% | 8% |
| No Frills | Uncle Tobys | Kellogg | Sanitarium | |
| Sugar | high | high | high | high |
| Fibre | low | low | low | low |
| Fat | high | high | high | low |
| Toys | no | no | no | no |
| Price (0% = average) | -15% | -5% | +5% | -5% |
| Preference share | 2% | 19% | 31% | 48% |
Note that in this simplified example, it has been assumed that all brands have the same price sensitivity. It may be that the more popular brands have a lower price sensitivity - but a more complex design would be needed to capture that effect.