Using The Past To Predict The Future

I often use
the past to help predict
the future.

I’ll look at past sales
for vampire romances, for example,
to help predict
how many sales
my vampire romance
might have.

In the past,
this prediction
was dang close.

Today?
It could be
business-flatteningly off.

The world has changed
significantly.

In the book world,
AI and COVID
has shifted the baselines.

But we are also in a new era
for climate
and pandemics
and politics
and almost everything else.

Being VERY cautious
when using the past
to predict the future.

Does Your Model Still Work?

Weather prediction models
are based on a stable jet stream.

The jet stream
is no longer stable.

And the weather prediction models
no longer work
for any predictions longer
than a day out.

I suspect
we are all basing decisions
on models
that no longer work
in today’s world.

I know I am.

In the past,
as one example,
if readers downloaded
one of my free books,
11% of them
would read the entire series.

I built my rate of return on promotions model
around that read-through
(sell-through)
number.

But that read-through number
is no longer applicable.
Very few readers
now read the entire series.

When I modified
my rate of return on promotions models
and applied the new
read-through number
(plus some other factors),
none of the promotions
I was considering
made a return.

Do your models
still work?