We propose a backreaction toy model with realistic features and confront it with the Union2.1 supernova data .
The model provides a good fit even though the expansion history is quite different from the \Lambda CDM model and the effective equation of state is far from -1 .
We discuss compatibility ( or lack thereof ) with other observations .
We show that the FRW consistency condition between distance and expansion rate is violated with an amplitude that is slightly below the current observational limits .
We expect that this is also the case in the real universe if backreaction is significant , providing a distinct signature of backreaction .