MRG Meeting – Satellite Universes

At last night’s MRG meeting speakers from BARB, Continental Research & Carat UK all spoke about satellite and cable universes – looking at how figures were currently arrived at, and what they thought the future held.

Bill Meredith from BARB looked at each of the base methods used by the three companies producing estimates; BARB, Continental Research and GfK.

BARB:

UK random sampling covering 3500 in home interviews per month, plus monthly projection using dish sales figures and ITC cable research (Narrow & Broadband)

C.R.:

UK random sampling covering 4000 telephone interviews per month, plus dish sales data from retailers and ITC cable research. (Broadband only)

GfK :

Random sampling in GB (excludes N.I., IoM and Channel Islands) of 10,000 individuals at 6 monthly intervals, plus monthly projections from home audit purchasing and Lek-trak sales data.

He stated that because BARB used personal interviews its sample was partially clustered, whilst C.R.’s telephone interviews meant its had no clustering.

At the end of the day, according to Meredith, there was not much difference between the estimates provided by all sources and press speculation, on the whole, was based on individuals not taking account of the different methodologies used. The figures at the end of ’92: Barb 2.3m, CR 2.6m, Gfk 2.1m, showed the market was between 2.1 and 2.6m which was good enough. Dish homes growth for 1992 was predicted by each source as follows:

John Clemens of Continental Research went on to look at their methodology in more detail, and how over time they had amended their forecast to account for the changing environment:

Their monthly figure is currently based on:

• Weekly retail audit from independent retailers – sales and rentals.
• EPOS data from high street multiples such as Dixons, Currys and Comet.
• Philips/Thorn estimates – rentals out versus rentals in.
• 1000 weekly telephone interviews.

Dave Chilvers of Continental examined the main factors affecting dish sales, and went on to give C.R. predictions up to 2000. He stated that the three key demographic factors affecting dish sales were:

1. Size of household – The larger the household, the more likely to purchase.
2. Age of head of household – The younger the heads of household, the more likely to purchase.
3. Social class – C1/C2 more likely to purchase. (This factor is seen as less important that the others.)

Forecast figures to 2000 were as follows:

He emphasised that in his opinion by the end of the decade replacement dishes would be more important than new dish sales and that there would be a much higher growth seen in people accessing via cable rather than by dish.

Phil Gullen of Carat UK ended the evening by talking about the affect of ‘churn’ on satellite penetration.

From research carried out by Carat UK, which took the form of re-interviewing TGI respondents, ‘churn’ amongst individuals between April 1991 and November 1992 was 19%, equating to a 15% ‘churn’ in homes. Based on a ‘churn’ factor of 15% he estimated by 2000 there would be a 32% satellite and cable penetration of all homes; if this factor were reduced to 5% the penetration would be 48%. Forecasts to 2003, accounting for ‘churn’, he gives as follows: