What is Source Influence?

Domain authority is a general, category-agnostic estimate of a site’s link strength. Source influence is specific to your question set: it counts how often a source is actually credited when assistants answer the questions you care about. A famous domain that never appears in your category’s answers has high authority and no influence, while a niche forum thread credited in most answers has the reverse. Only the second one tells you where a mention would be read.

Sometimes, and never predictably. Assistants re-retrieve on their own schedule, weigh multiple sources, and may summarise a page without repeating every brand it names. Treat a mention as improving your chances of appearing, then verify by re-asking the same question later and comparing. Any single before/after is one directional sample rather than proof, so look for a pattern across repeated runs.

No — they drift, sometimes quickly. Engines change their retrieval mixes, publishers update roundups, review platforms restructure category pages, and new community threads outrank old ones. That is why source influence is worth re-measuring on a schedule against the same question set rather than captured once as a static target list.

Run a stable set of category questions against the assistants on a schedule, capture the URLs each answer credits, and aggregate them across questions and runs. The ranking that falls out — which domains and which specific pages recur — is the measurement. Repetition matters more than any single run, since assistants vary between requests, and a source that appears once may simply be noise.

It becomes an outreach target rather than a content task. Depending on the source that might mean claiming and correcting a review-platform listing, contacting the author of a roundup that describes your product inaccurately, or participating honestly in a community thread. The value of measuring first is that it makes the pitch concrete: you are asking about a specific page that a specific assistant is demonstrably quoting for a specific question.

Frequently, yes — a rival’s documentation or comparison page is often credited when assistants explain how something in the category works. You cannot get mentioned there, so the useful response is to publish the equivalent explanation more plainly on your own domain and to strengthen the neutral sources where both of you can appear. Knowing that a competitor page is doing the explaining is itself a finding.

It replaces the prospecting step rather than the process. Instead of exporting a list filtered by authority metrics, you start from the pages your tracked answers already credit, which is typically a much shorter list with a much stronger pitch attached. Everything downstream — contact discovery, the email, follow-up — stays as it is.

Effectively per question set, which usually means per category and per market. Two clients in the same industry with different buyer questions will surface overlapping but distinct source rankings, and the same brand in another country can see an almost entirely different set. Agencies get more from maintaining a question set per client than from a shared master list of sources.

Fewer than most teams expect. Aggregated across a tracked question set, credited URLs tend to concentrate heavily in a short head — a review platform or two, a handful of editorial roundups, one or two community threads — with a long, thin tail that appears once and never again. Focusing outreach on the concentrated head is what makes the work tractable.