Percentage change in Clustered chart when aggregating years

[mcdyl]

I have a survey answers and I wanted to compilate them in a cluster column chart and see the answers by percentages. 

But when I started adding more years then the calculation of the % calculates it for the whole databse and not year by year. So my first answer which was chosen by 61% of the people appears in the bar as 7,3% of the total (sum of samples of all the years included). 

Is there any way to be able to have all the years and maintain their percentage they represent, only of the data of their year (2017), instead of calculating the percentage for the whole dataset(2015-2022)?

[lbendlin]

Have you considered using small multiples?