Skip to content
SourcepathBook a Demo

Competitors

The answer has sources. Your domain isn't one of them.

Sourcepath reads every answer to the buyer questions you track and records the pages it cites. When a competitor's domain is among those sources and yours isn't, that gap is captured with the question, the provider, and the citation behind it.

Book a Demo
An illustration of one observed AI answer and the pages it cited. The tracked buyer question is "Which customer data platforms handle enterprise identity resolution best?", and the answer shown is ChatGPT's. A region of light marks the set of sources that answer cited. Inside it sit 5 pages: competitor-one.com, analyst-brief.org, industry-review.com, data-standards.org, practitioner-forum.net. One of them, competitor-one.com, is a page on a competitor's own domain — its title is "Enterprise Identity Resolution: A Buyer's Guide" — and it is the citation this observation is about. Outside the lit region, on the same plane and at the same scale, sits yourcompany.com: the brand's own domain, which is not among the sources this answer cited. The sources are drawn in no particular order. Sourcepath records the position of a citation but nothing here ranks them, and being cited is not a claim that the page caused the answer, that the provider preferred it, or that the competitor is stronger. The only difference asserted between the two is whether the domain appears in this answer's citations. The environment shown is representative, not customer data.

ChatGPT “Which customer data platforms handle enterprise identity resolution best?”

The landscape

You track six. AI cited forty.

Sourcepath records every domain cited across the buyer questions you track. The ones you configured are marked. It doesn't guess which of the rest are competitors.

An illustration of the domains cited across the buyer questions this brand tracks, drawn as the same landscape of circles the Sourcepath product uses. In the latest scan, 40 distinct domains were cited across 8 tracked buyer questions, and each is drawn as one circle. 6 of them are marked as configured, drawn as solid circles: a person added them to this project's competitor list, which is the only way a domain becomes a tracked competitor in Sourcepath. The other 34 were cited but are not on that list, and are drawn as dashed rings. They keep their own colour, because colour here is identity and material only — it is not a category, a score or a class. An unconfigured cited domain may be a competitor, a publication, a community, a marketplace or another kind of source, and the product's own threshold cannot tell those apart, so none of them is presented as a competitor. A circle's size is the number of tracked questions on which that domain was cited. The largest circle in the field, analyst-brief.org, is not on the configured list: it was cited on 7 of 8 tracked questions, by 4 of 4 providers. Size means citation breadth only, never quality, rank or strength. There is no axis, no grid and no quadrant: position is composed for the picture and reflects no measurement at all. Nothing here ranks or scores these domains, and nothing claims why any of them was cited. The environment shown is representative, not customer data.

AI Search Landscape

Every domain cited across your tracked buyer questions — bubble size is tracked questions cited on.

opensource-docs.orgindustry-wire.netdata-council.orgintegration-hub.ioproduct-brief.combenchmark-lab.orgsystems-journal.netapi-directory.iotrade-press.comcommunity-answers.netresearch-notes.orgplatform-digest.comspec-archive.orgengineering-blog.iomarket-notes.netreference-desk.orgtoolkit-index.comcomparison-site.netadvisory-group.orgfield-guide.iocompetitor-five.comcompetitor-six.comcdp-guide.netenterprise-weekly.comschema-registry.orgbuyers-desk.comdev-notes.iostackforum.netcompetitor-four.comvendor-compare.comtechdigest.ioknowledge-base.orgmartech-index.comcompetitor-three.compractitioner-forum.netdata-standards.orgcompetitor-two.comindustry-review.comcompetitor-one.comanalyst-brief.org

analyst-brief.org

Cited — not on your configured list

Cited on
7 of 8 tracked questions
Cited by
4 of 4 providers
  • 6 configured competitors — solid
  • 34 other cited domains — Sourcepath doesn’t classify these
  • Bubble size = tracked questions cited on · position means nothing

The evidence

One question. Four answers. Exactly one cites a competitor.

Sourcepath keeps every completed answer to the buyer questions you track, and the pages each one cited. So a competitive gap resolves to something you can open: this question, this provider, this citation — and your own domain in none of the four source lists.

An illustration of one Sourcepath prompt-detail surface. The tracked buyer question is "Which customer data platforms handle enterprise identity resolution best?". The recorded observation reads: ChatGPT cites Competitor One (competitor-one.com) answering "Which customer data platforms handle enterprise identity resolution best?", without citing Your Company's own site. Beneath it are the 4 completed answers to that question. ChatGPT names Competitor One in its answer text, and cites competitor-one.com. Its cited sources are: competitor-one.com, analyst-brief.org, industry-review.com, data-standards.org, practitioner-forum.net. Claude names Competitor One in its answer text, and does not cite competitor-one.com. Its cited sources are: analyst-brief.org, practitioner-forum.net, vendor-compare.io. Gemini does not name Competitor One, and does not cite competitor-one.com. Its cited sources are: techdigest.io, analyst-brief.org, data-standards.org. Perplexity does not name Competitor One, and does not cite competitor-one.com. Its cited sources are: industry-review.com, data-standards.org, schema-registry.org, cdp-guide.net. Two different records are shown for each answer and are not the same thing: whether the competitor's name occurs in the answer's text, which is a mention, and whether the competitor's domain occurs in that answer's cited sources, which is a citation. 2 of 4 answers name the competitor; 1 of 4 cite their domain. yourcompany.com is not a source in any of these 4 answers. Nothing here ranks the answers, the providers or the sources, and being cited is not a claim that a page caused an answer, that a provider preferred it, or that the competitor is stronger. The environment shown is representative, not customer data.

  • ChatGPT

    Names Competitor OneCites competitor-one.com

    For enterprise identity resolution, buyers typically shortlist a small number of platforms. Competitor One is usually in that comparison, and the practical differences tend to come down to how each one handles…

    Sources citedcompetitor-one.comanalyst-brief.orgindustry-review.comdata-standards.orgpractitioner-forum.net

  • Claude

    Names Competitor OneNot in sources

    There are several established options in this category. Competitor One and a handful of others come up often, though the right choice depends heavily on the data infrastructure already in place and on…

    Sources citedanalyst-brief.orgpractitioner-forum.netvendor-compare.io

  • Gemini

    Not mentionedNot in sources

    Enterprise identity resolution generally relies on deterministic matching, probabilistic matching, or some combination of the two. Most platforms in this category support both, so evaluation usually turns on…

    Sources citedtechdigest.ioanalyst-brief.orgdata-standards.org

  • Perplexity

    Not mentionedNot in sources

    At enterprise scale this is usually assessed on match rate, latency, and how conflicting records across sources are reconciled. Vendor documentation and independent comparisons differ noticeably on…

    Sources citedindustry-review.comdata-standards.orgschema-registry.orgcdp-guide.net

AI Search Patterns

One cited page is an example. Seven start to show a pattern.

Sourcepath fetches and analyses the pages cited for the buyer questions you track, then records what recurs across them — and only calls something a pattern once enough of the set shares it.

An illustration of Sourcepath's AI Search Patterns. Nine pages were cited across the answers to the tracked buyer question "Which customer data platforms handle enterprise identity resolution best?". 7 of them have been fetched and analyzed: competitor-one.com, analyst-brief.org, industry-review.com, data-standards.org, practitioner-forum.net, vendor-compare.io, techdigest.io. The remaining 2 — schema-registry.org and cdp-guide.net — have not been analyzed and are not part of any count below. 5 of 7 analyzed cited pages answering this question include an FAQ section. That is 71% of the analyzed set, which the product classes as high confidence. 6 of 7 analyzed cited pages answering this question include a comparison table. That is 86% of the analyzed set, which the product classes as high confidence. 7 of 7 analyzed cited pages answering this question include a published date. That is 100% of the analyzed set, which the product classes as high confidence. 4 of 7 analyzed cited pages answering this question include FAQ schema. That is 57% of the analyzed set, which the product classes as medium confidence. These are characteristics observed across the analyzed pages. Nothing here claims those characteristics caused the pages to be cited, that any provider prefers them, that the pages are better, more authoritative or more relevant, or that adopting them would cause a citation. A characteristic is only reported when at least 3 cited pages have been analyzed and at least half of them share it. The environment shown is representative, not customer data.

  • An FAQ section

    5 of 7

    High confidenceHeuristic

  • A comparison table

    6 of 7

    High confidenceHeuristic

  • A published date

    7 of 7

    High confidenceDeterministic

  • FAQ schema

    4 of 7

    Medium confidenceDeterministic

5 of 7 analyzed cited pages answering this question include an FAQ section.

Based on 7 analyzed cited pages. 2 more were cited for this question but haven't been analyzed.

Questions

What teams ask about AI competitive intelligence. Answers about the competitive set AI actually names.

What is AI competitive intelligence?

It is competitive analysis built from what AI engines actually answer, rather than from a list you maintain by hand. For every question your buyers ask, Sourcepath records which companies were named, which pages the answer cited, and where you stood — so the competitive set is observed rather than assumed.

Can Sourcepath find competitors we are not already tracking?

Yes. Any site cited often enough across your tracked questions is surfaced alongside the competitors you added yourself. It is identified by its domain, because a domain is what was actually observed — Sourcepath will not invent a company name from a hostname. In practice this is where the surprises come from.

Why does the competitive set change depending on the buyer question?

Because the answers do. Ask a category question and you get the names you expect; ask about one specific requirement inside that same category and the field can reconfigure underneath you. Sourcepath keeps each question's set separate rather than averaging them into one list, which is the only way that shift stays visible. See that play out in a security evaluation.

Can we compare competitor visibility across AI answers?

Yes. Each tracked organisation is measured the same way over the same set of questions and engines, and every figure carries the counts behind it. Sourcepath reports how often a company appeared — never a ranking, a score of who is winning, or a share-of-market claim it has no basis to make.

Is this the same as monitoring competitors on traditional search?

No. There are no ten blue links to compare. An AI answer names a handful of companies and cites a handful of pages, and the useful unit is the question a buyer asked — so the work is reading answers and their sources, not tracking positions.

Now see what Sourcepath finds about your brand.

Book a Demo