Before reading this, please read the tutorial about table-based projects.
Ok, so as you remember we've had this table in that tutorial:
We have spent $1.65 to process 55 rows. Can we do it cheaper? Yes, but it comes at a cost of more processing of results for you.
First we need to re-format the CSV data like this:
Basically, we took one column and moved data into 3 columns and renamed our "city" column to "city1", "city2" and "city3".
Next I'll copy the previous project
And edit the draft project
Next I'll edit the substitution to refer to "city1" field and copy the field two times.
Add your new table data, like in previous tutorial via "Add table-based input" and run the project!
The project was done in 5 minutes and at three times less cost.
However, the price is that now we have to unwrap our batches in answers.
You can definitely notice a few lazy users in the results that answered that Brunswick is not a city or that Alfa Romeo is a city. To learn to deal with that without resorting to manual approvals, please read our tutorial about multi-stage projects