We can multiply, that's going to be the likelihood associated with
each of the incremental triers.
So just to walk through the approach that we're going to be taking,
we're going to first calculate out what those incremental triers,
how many there are, that will be straight forward for us.
In column D, we're going to calculate that survival probability.
So what's the probability that someone has not yet
tried the product by the beginning of that week?
We're going to then calculate the incremental probability
of trying it in a particular week in column E.
In column F, we then going to combine the number of incremental triers we have with
that probability of try on a particular week to produce our log likelihood.
We're also going to have to take into account the likelihood associated with
the number of people who have not yet tried the product.
And then what you see in columns G and H is we're going to predict how many people,
according to our model, should have tried the product by a particular week.
That is, what's the expectations?
Now see, we provided up the data for 52 weeks,
but we're not going to use necessarily that full dataset.
What we may want to look at, for example, is splitting that data so
that we have a portion to be used for forecasting.
The alternative, we could use that full dataset and
make our forecast based on that.
Now what you'll see later in this worksheet,
I've also put in completed versions of what we're going to be working through,
so you can go back and look at the videos the steps that we are taking.
You can also refer to the completed sheet of the corresponding module, all right.
So what we are going to do in building this up is,
we are going to try to build the model using the first 26 weeks of data.
And then we are going to see how well the model does at testing the rest of
this data.
All right, so to get started,
let's calculate out the number of incremental triers.
Well that's going to be the difference between how many people tried it,
in the first week, it's just going to be the number of triers that we have,
that's the first week it's offered.
In subsequent weeks,
it's going to be the difference between the cumulative triers in adjacent weeks.
So cell C7, that's going to be the difference between how many
people tried it in week 2, and how many people tried it in week 1.