Unlike the ill-advised and since-revoked US dietary recommendations against cholesterol, these processed-meat cancer-risk estimates are based on years and years of empirical research.
The 18% estimate comes from a 2011 meta-analysis, published in the open-source journal PLOS One, of nine different studies of colorectal cancer.
It wouldn't be unreasonable to make changes in your diet based on this evidence (and similar but weaker evidence on the link between red meat and cancer).