Automating Prospect Segmentation with Regex and CSV Pipelines

The solo SEO operator who has graduated from manual copy-paste outreach knows the real bottleneck is not writing emails—it’s extracting, normalizing, and categorizing the raw data that feeds those campaigns. You’ve already scraped a list of 2,000 prospective domains from a competitor’s backlink profile, or maybe you’ve exported LinkedIn Sales Navigator results. Now you need to slice that mess into meaningful segments: high-authority bloggers, broken-link targets, resource-page candidates, and guest-post opportunities. Doing this by hand is not only soul-crushing but also error-prone, and it introduces latency that kills momentum. The solution is a purely deterministic, regex-driven segmentation pipeline that runs in seconds on your local machine or inside a CI/CD action.

Think of your raw CSV as a blob of unstructured text wearing a thin veneer of structure. A column named “Domain” might contain `https://www.example.com/blog/` or `example.com` or `sub.example.org:8080`. A column named “Description” could hold anything from a single sentence to a full paragraph of SEO fluff. You need to extract the root domain, classify the URL path (is it a homepage, a /blog/ directory, a /resources/ page, a /contact/ page?), and infer the site’s type—personal blog, e-commerce store, government agency, or industry publication. This is where regular expressions become your scalpel.

Start with a Python script that reads the CSV using the `csv` module (or `pandas` if you prefer the overhead). For each row, apply a series of compiled regex patterns. For domain extraction, use `re.search(r’^(?:https?://)?(?:www\.)?([^/:]+)’, row[’URL’])` to grab the hostname without protocol and www prefix. Then split on dots and check the TLD—but beware of country-code TLDs like `.co.uk` or `.com.au`. A proper pattern for those looks like `r’\.([a-z]