As a general term an "optimiser" simply means a system that produces the best result for a given set of inputs.

 Many years ago, the only so-called television optimisers were ones that simply provided a set of programmes (or spots or breaks or advertising blocks) that achieved the largest number of GRPs for a fixed amount of money. This could be done simply by estimating the average GRP of each programme, dividing this by the cost of the programme, ranking the results and picking whichever programmes came highest. However, nowadays, it is commonly accepted that the true effectiveness of an advertising campaign is measured by its reach - not simply the total number of GRPs obtained. It is important to point out that other so-called TV optimisers of 10-15 years ago just used a model based on GRP levels per daypart to estimate reach - and this simply wasn't accurate enough and would often produce meaningless results. Around 1989 it was realised that any practical TV reach optimiser must use respondent level data.

          Many other optimisation systems have come and gone. However, X*Pert, launched in 1994, is one of the most scientific television optimiser being available in excess of 40 countries world-wide.

One of the traditional criticisms of systems such as X*Pert is that is does not give a definitive list of programmes (or spots) to buy - it merely gives daypart allocations of money (even though the dayparts may be defined as programmes or groups of programmes). Consequently there have been some attempts to develop "spot optimisers" as opposed to "daypart optimisers".

However, it has been found that spot optimisers cannot ever work in realistic conditions. The reason is that it is very easy to optimise a set of spots (or set of programmes) if you look in hindsight at exactly what each panel member viewed - but producing such wonderful schedules on historical data gives little or no real indication of how successful the same schedules will be in the future. This is why X*Pert aggregates data across a statistically-robust period (i.e. a daypart) so that the resultant information is not only accurate on past data but also gives a reliable and repeatable guide to what will happen in the future.

The purpose of X*Pert is to provide "better" schedules. A better schedule is one that delivers more reach for the same budget (or number of GRPs) or the same reach for a lower budget (or less GRPs). X*Pert is not a spot optimiser (and why it isn't explained in Appendix 2). X*Pert provides recommendations on the volume of airtime in dayparts that act as a guide to spot buyers in order to ensure that a good reach level is attained.