In this issue, Meier [1] proposed a scientific research agenda to study the effects of alcohol marketing and addressed relevant issues concerning marketing complexity, target groups and timing of effects. She offers an important overview of knowledge gaps in the field of alcohol marketing research, in particular on the lack of theoretical basis of most studies. As she discussed our studies as well—in the realm of direct marketing effects in groups of (heavy) drinkers—we will elaborate specifically on that part.

Besides the study of indirect, delayed effects of alcohol marketing on alcohol use, we are interested in whether people pour themselves a beer or another glass of wine when they are exposed to alcohol portrayals on TV or in the movie theatre. Concerning alcohol commercials, we found mixed evidence. Our first experiment among a sample of 80 male university students [2]—to which Meier refers—indeed demonstrated immediate effects on alcohol consumption while watching alcohol advertisements. In a similar vein, Koordeman et al. [3] demonstrated that alcohol commercials prior to a movie (Watchmen) in a movie theatre led to increased consumption of alcohol, but only in heavy drinkers. However, in a new experiment aiming to replicate our first study [2], 160 male participants watched an informative movie (Planet Earth) interrupted by three commercial breaks containing alcohol or non-alcohol advertisements (Koordeman et al., unpublished data). They were allowed to drink alcohol and non-alcoholic beverages while watching. We did not find immediate effects of alcohol commercials on drinking, and also not in subgroups for which alcohol portrayals might have had more profound effects, such as heavy drinkers or those who like and remember the alcohol commercials. Therefore, the immediate effects are not as straightforward and clear as one might expect. We sincerely hope that other groups are setting up experiments to replicate our findings, as this is the most important step to move forward.

This probably depends upon how alcohol cues are embedded in the media. In another experiment [4], in which people were exposed to a movie clip (a carefully edited clip of What Happens in Vegas), it appeared that those men watching a movie with alcohol cues drank more alcohol than those watching a version without the cues. No effects were found for women. The effects for men's drinking were in line with previous work [2]. Currently, we think that because people are transported into a storyline and identify with the actors and characters, they are more prone to be affected by alcohol cues. People probably watch a movie with more attention than a commercial break, which is often used to do something else. Consumer research showed that consumers' awareness and scepticism towards advertising predicts whether people buy or consume certain products [5,6]. In general, consumers know they are being persuaded [5], and scepticism might make people aware of advertising strategies and develop reactance [7]; therefore, immediate media effects might be much stronger for alcohol portrayals in movies than for explicit alcohol advertising.

Conversely, the context of viewing alcohol commercials might be essential. In a movie theatre, for example, participants are probably more concentrated on and transported into both movie and advertisements than in our laboratory setting or at home. If they have the opportunity to drink while watching—in many theatres people can order drinks while watching a movie—or directly after the movie in a pub or club, the emotional state instigated by the movie might strengthen or focus attention on the alcohol cues in drinkers. High programme involvement might rivet people's attention to the screen, causing a higher advertisement exposure rate [8] which might affect alcohol intake.

Further, we do not know whether brand or drink preference plays a role in immediate effects of alcohol commercials. It might be that beer advertisements exert more influence on men than liquor advertisements because beer advertisements are shown more frequently on TV and also because beer is popular among male young adults [9,10]. In line with this, alcohol brands try specifically to address certain target groups [11], and it is unknown whether fit between brand, type of alcohol and target group is important for immediate effects on consumption while watching.

We fully agree with Meier that not everyone is affected by alcohol marketing. This strongly urges for the need to specify and test theoretical models that elaborate on individual differences in susceptibility. Dual-process models [12]—on the interplay between automatic processing of alcohol cues and deliberate control mechanisms—in combination with models on behavioural imitation [13–15] are highly informative and should steer developments in this field. When we expect huge effect sizes in general population samples we will be disappointed as, based on these models, we can only expect effects in specific groups. We do not expect a 45-year-old female teetotaller to engage in a big drinking fest when she incidentally watches a soccer game interrupted by a beer commercial. The same expectation probably accounts for the company who developed this commercial and decided to broadcast it during this game.

Experimental research designs are well suited to test immediate and causal effects of alcohol marketing on alcohol consumption. However, this line of research has just begun, and before firm conclusions can be drawn replication in other samples and with other designs (e.g. different movies and alcohol advertisements, cumulative exposure, extended exposure effects) is warranted.

Declaration of interest