The draft of this blog post was written the old fashioned way today.
Pen and paper. Drafted whilst sitting on a rock at the beach - a
detour after a meeting out of town - on an unseasonably nice afternoon.

The weather guys probably didn't know this afternoon was going to be
quite as sunny and pleasant as it turned out to be. I'm pretty sure no
marketing department predicted it.

It reminded me that no matter how much data you use to predict
people's behaviour there is always the element of biology that will
affect what people do and how they act. Using data to time marketing
on a detailed individual consumer level sounds great in theory, but
will be subject to a heap of factors that no data can predict for.
Including biological responses to external factors. Miss these, and
you can end up misled.

Companies that send electronic marketing messages measure performance
and relate this back to the message and the offer, and the target
market details.

There are plenty of other factors though. Weather is one of these and
it's usually overlooked. On a sunny day, people will generally prefer
to be outside, doing things other than reading email.

Big data is being used to predict individual preferences, including to
help determine the exact time of day that individuals are more likely
to be receptive to advertisements.

Great in theory. Some behaviour is entirely predictable. An office
worker will be at his or her desk at 9am Monday. And again at 9am

But their responsiveness will be different.

People operate on regular cycles of alertness, rest and sleep. It's
the circadian rhythm, the body clock. Our alarm clocks work on a 24
hour cycle. Our body clocks generally work on a 25 hour cycle... the
natural 24 hour cycle gets extended out by exposure to artificial
light at night. Some people prefer to operate on even longer cycles.
We're all different, and we're all different to our alarm clocks.

This means that at the same times but on different days, we'll feel
different. This means we'll act different.

Seasons and latitude make a difference too.

Timing behaviour predictions on a 24 hour clock will always make the
result a little inaccurate.

Using data helps business, and data-driven insights can be of enormous
value, but be aware that data doesn't drive behaviour and can't
predict all of it either.

There is only so much detail you can get into before you need to stop,
and recognise that biology is important.

At ZenBus we are all for market forecasting and consumer segmentation.
We know that economic trends affect business success so some element
of timing is important. We know that certain people can be prompted to
act in certain ways. Our business is based on this.

But we also think that making tactical marketing decisions based on
information such as an ideal customer being a 35 year old woman who
likes a,b and c and has previously done x,y and z can backfire if the
bigger picture isn't taken into account. The most highly refined and
targeted marketing can sometimes miss the mark.

The solution in such cases is not always a more complicated predictive
model or more complex algorithm. Sometimes all you need to do is look
outside. Is it sunny? It matters.