Does your dataset include names or personally-identifiable information (IMO being able to key by all those 1-person companies counts)? If so, that seems a little... scuzzy? Would you consider pulling the dataset or at least making it coarse and pseudonymous? I would guess many of the people included and monetized here wouldn't expect or wish to be.

If the response is public scraping is legal, then I might suggest your company's framing of the product is making incorrect claims about this group. From your blog page:

"[N] people indicated a change of employment in the last 4 months (from the date of writing i.e. 14 February, yes Valentines’ Day) - aka ex-Googlers that got laid off."

That 'aka' doesn't necessarily follow, though? People leave companies for many reasons, and therefore this is not a valid "laid-off Googler" dataset.

(Edit: I might be misunderstanding but Google's layoffs were announced less than a month before Feb 14. Why look at the three months before that? Edit 2: noting sibling comment already asked this, didn't mean to duplicate)

Less important, while you do mention some biases inherent in the source (LinkedIn use by country, who has/updates a profile), please note that some of your blog article's headline claims are contradicted by publicly-available data available from press and WARN act filings.