APPENDIX

CASE STUDIES

Effect of the new LSM RIM weight structure and Universe update on TAMS data

CASE STUDY 1: Target Groups comprising LSM5 – 7

  • Adults LSM 5 – 7
  • Housewives LSM 5 – 7
  • Black Housewives LSM 5 – 7

When using target groups that include LSM 5 – 7, it should be noted that the RIM weight structure that was applied prior to Week 32 (i.e. before 6 August 2012) grouped LSM 5 together with LSM 1 – 4. In the new RIM structure, LSM 5 has been separated from this group and can be selected in isolation. As a result of this change, the LSM 5 universe on an all individual level has been fixed and will remain constant over a selected period of time.

The trend-line below for several LSM 5 target groups, from Week 10 to 39 in 2012, shows that the Universe for LSM 5 increased slightly after the Universe was updated in Week 27. However, after the application of the new RIM weighting structure in Week 32, a significant decline in the LSM 5 Universe is clearly observed. The decrease in the LSM 5 Universe is explained in detail in theEffect of the new LSM RIM weight structure on TAMS data article.

On further analysis of the TVR 000s for the same LSM 5 – 7 target groups, it can be seen that the impact of the change on TVR 000s mirror those seen in the Universe. A slight increase in TVR 000s occur after the Universe Update in Week 27. However, TVR 000s decline after the RIM Weight change in Week 32. The audience has been re-distributed to the LSM 1 – 4group. See the Effect of the new LSM RIM weight structure on TAMS data article.

Although the Universes and TVR 000s for LSM 5 – 7 target groups show a decline after the RIM weight change in Week 32, viewing levels, which are percentage off these lower base numbers, show a healthy increase.

CASE STUDY 2: Target Groups comprising LSM 8 – 10

  • Adults LSM 8 – 10
  • Housewives LSM 8 – 10
  • White Housewives LSM 8 – 10

When using target groups that include LSM 8 – 10, it should be noted that the RIM weight structure that existed prior to Week 32 (i.e. before 6 August 2012) grouped LSM 7 and LSM 8 together. In the new RIM weight structure, LSM 7 and LSM 8 have been separated and weighted in isolation. As a result of this change, each of the LSM 7 and LSM 8 Universes on an all Individuals level has been fixed and will remain constant over a selected period of time.

The graph belowdepicting Universes from Week 10 to 39 in 2012shows that the Universe update had little impact on LSM 8 – 10 target groups. A marked change for these target groups, however, can be observed after the implementation of the new RIM weighting structure in Week 32. Universe sizes increased significantly for all the target groups represented below.The increase in the LSM 8 – 10 Universe is partly explained in the Effect of the new LSM RIM weight structure on TAMS data article.

The fluctuations in the TVR 000s after the Universe Update are minimal, and suggest that the Universe Update had little impact on the TAMS data. The dramatic increase in TVR 000s after the RIM weight structure change however, reflect the re-distribution of audiences that occurred in the re-distribution of the LSM 7 and 8 audiences.

Viewing levels expressed in TVR percentages show some fluctuations directly after the Universe Update, which is naturally influenced by the TV content that was being broadcasted at the period. After the RIM weight change, an overall increase in viewing is seen as expressed in percentages.

CASE STUDY 3: Target Groups including Income Categories

  • Adults R 1 – 3999
  • Adults R 1 – 7999
  • Adults R 4000 +
  • Adults R 8000 +

Although income is not a variable included in the calculation of LSMs, there is a correlation between LSM group and Household (HH) Income. With the Universe Update in Week 27, slight increases in the Universes of the higher Income categories i.e. R 4000 plus and R 8000 occurred. The Universes for the lower HH Income categories, i.e. HH Income R 1 – 7999 and R 1 – 3 999 declined slightly. After the RIM weight structure change was applied, a slight increase in the Universes of HH Income R 4000 plus and R 7000 could be seen. We can therefore, conclude that generally the higher income categories grow in Universe and that this reflects the growth in Universes in LSM 8 – 10. Conversely, the general effect of the RIM weight structure change on the lower income categories is an overall decrease in the Universes of these groups.

When looking at TVR000s, the effects of the Universe changes are accentuated, where higher income categories show a growth in 000’s in viewing (as is seen with LSM 8 – 10) , and lower income categories show lower viewing levels in TVR 000’s.

In terms of viewing levels by percentage, the higher income groups show a stable, if slight increase, in their viewing levels that mirror the increased viewing percentages for the LSM 8 – 10 Group. Lower Income groups show a marked decline in viewing levels.

CASE STUDY 4: Target Groups that include Housewives

  • Housewives
  • Housewives LSM 5 - 7
  • Housewives LSM 8 - 10
  • Housewives R 1 – R 3999 HH Income
  • Housewives R 1 – R 7999 HH Income
  • Housewives R 4000+
  • Housewives R 8000 +

The Universes for total housewives increased slightly after the 2nd July update and then decreased slightly after the LSM unbundling which was applied in Week 32. LSM 8 – 10 housewives, as well as higher income housewives (i.e. R 4000 + and R 8000 +) show increases in Universe sizes after Universe Update and then increase slightly again after the LSM unbundling. LSM 5 – 7 housewives, as well as lower income housewives (R 1 – R 3999 and R 1 – 7999) show increases after the Universe update, but Universes decline after the LSM weighting structure change.

In the table on the next page the TVR000’s for total housewives increased after the 2nd July update and remained steady after the LSM unbundling which was applied in Week 32, despite the small decline in Universe sizes. This suggests that actual viewing levels for this group increased (see comments on TVR % below for further confirmation). LSM 8 – 10 housewives, as well as higher income housewives (i.e. R 4000 + and R 8000 +) show increases in TVR 000s after Universe Update and then again after the LSM unbundling. LSM 5 – 7 housewives, as well as lower income housewives ( R 1 – R 3999 and R 1 – 7000) show TVR 000 increases after the Universe update, but decreasing TVR 000s after the LSM weighting structure change in Week 32.

For total housewives, viewing incidenceincreases slightly despite the Universe update and the RIM weight change. After these changes to the TAMS data, LSM 8 – 10 and higher income housewives ((i.e. R 4000 + and R 7000 +) show increases in TVR %, indicating increased viewing. Despite the drop in Universes and TVR 000’s for the LSM 5 – 7 housewives, viewing levels show increases, as the TVR percentages are higher after the changes were implemented in the TAMS data. Lower income housewives follow the same trend as the LSM 5 – 7 Housewives.

Summary of Findings

  • Universe figures and TVR 000’s for LSM 5 – 7 have declined due to changes in the LSM RIM structure. Viewing levels expressed in percentages, however, show an increase in viewing on the new bases.
  • The Universe sizes for LSM 8 – 10 groups have increased after the RIM structure changes, and TVR 000’s have increased accordingly. Percentage viewing levels have also shown an increase for LSM 8 – 10 viewers.
  • In accordance with the above observations, Higher Income groups (i.e. R 4000 + and R 8000 +) have shown increases in their Universe and TVR 000’s. Viewing levels in percentage terms have also increased after the RIM weighting structure change. Lower Income groups (i.e. R 1 – 3999 and R 1-7999) have declining Universe and TVR 000’s. Viewing in percentages have also declined markedly for these groups.
  • Despite the small drop in the Universe size for total housewives after the LSM unbundling, the TVR 000s remain fairly stable, and viewing levels actually show increases. While Universe sizes and TVR 000’s for lower LSM and lower income housewives decrease after the Week 27 and Week 32 changes, actual viewing levels increase as expressed in percentages. The lowest income level (R 1 – R 3999), however, show decreases in TVR 000’s and viewing levels after the changes.

END

* Housewives: Adult females who are wholly or partly responsible for household purchases

*All data is run on live + VOSDALPage 1