Predicting the best time to do things online is an ongoing challenge, whether it's the best time to write a tweet or put out an advert. A new model believes it has many of the answers when it comes to the best time to advertise on a variety of platforms.
The model was created as part of research from the University of Michigan and it predicts when people use various media each day, and even when they're using multiple media at the same time.
This is nothing new, but whereas previous models have laboured to achieve accuracy figures of 60-70%, this new model claims to be accurate 97% of the time.
“For businesses, our model does a much better job of predicting where your customers are at any given time,” says lead researcher Chen Lin, assistant professor of marketing at Michigan State University. “It represents a significant advancement over other models because much of that work assumed people consumed one type of media at a time.”
The researchers studied the media consumption habits of 2,000 American residents, and used this data to create their forecasting model.
The research revealed that 35% of our time is spent consuming one form of media or another, with television still the most popular, followed closely by the computer.
This changes at the weekend however, with more time spent watching television and reading print publications, with a subsequent drop in computer usage.
People spend about 1.5 hours a day consuming multiple media at the same time (e.g., surfing the web while watching television). This happens more at during the start of the workday and before bed—at about 9 a.m. and again at 9 p.m.
Mixing up your media
The researchers found that mixing media can often have a good result. For instance, print media remained popular early in the morning, and was also particularly popular when paired with other forms of media.
“The old thinking is that print is endangered, but we found that it doesn’t need to be eliminated,” Lin says. “Print can have a second life if it’s cleverly paired with new media such as personal computers and smart phones.”
Lin goes as far as to say that different media are in fact complimentary, and as such should not be regarded as competing for our attention.
For example, print ads should be partnered with radio and Internet media forms in the key time slots when consumers are likely to be using all three forms.
“Our findings underscore the need to move away from a competitive mindset to a coordinated viewpoint,” Lin says, “as consumers increasingly use combinations of media forms in short periods.”
Whilst the model is not commercially available at the moment, Lin confirmed to me that she is happy to trial it with companies if any are interested.