diff --git a/README.md b/README.md
index db32ce1..ee359fa 100644
--- a/README.md
+++ b/README.md
@@ -24,23 +24,30 @@ Check the version of the package after importing it:
 0.3.0
 ```
 
-Currently only "raw" datasets are implemented.
-
 You can see available datasets and load them directly into Pandas:
 
 ```python
->>> from tsdata.raw import available_data, load_data
->>> available_data()[:1]
-['LakeHuron']
->>> load_data("LakeHuron").iloc[:2]
-                   Time       Demand  Temperature        Date  Holiday
-0  2011-12-31T13:00:00Z  4382.825174        21.40  2012-01-01     True
-1  2011-12-31T13:30:00Z  4263.365526        21.05  2012-01-01     True
+>>> from tsdata.fpp3 import raw
+>>> "Tourism" in raw
+True
+>>> raw["Tourism"].head(2)
+   Quarter    Region            State   Purpose       Trips
+0  1998 Q1  Adelaide  South Australia  Business  135.077690
+1  1998 Q2  Adelaide  South Australia  Business  109.987316
 ```
 
-In the future, these will be available in `pandas` and `xarray` structures with proper indexes set.
+## Supported Datasets
+
+The currently support datasets are grouped into the following sources:
+
+- `tsdata.fpppy`, with data from [Forecasting: Principles and Practice, the Pythonic Way](https://otexts.com/fpppy/)
+- `tsdata.fpp3`, with data from [Forecasting: Principles and Practice, 3rd Edition](https://otexts.com/fpp3/) (extracted from the R package)
+
+Currently only "raw" datasets are implemented, i.e. as-is.
 
 ## Contributing
 
 If you have time series datasets you would like to add (that you have the rights
 to contibute), please create a pull request!
+
+Preferred formats are `.parquet` or `.csv`, though if it can be read by Pandas - we can add it.