This Week on the Citation Index

Vector Databases has four real brands stacked between 91% and 94%. SecurityScorecard is one missed prompt away from a perfect score in Third-Party Risk. DocuSign and three competitors tie in eSignature. UiPath runs alone in RPA. Here's the week.

The 94-94-94-92 cluster in Vector Databases

Vector Databases is the tightest top-of-table we've seen since launching the Citation Index. Four brands, 157 prompts, all four engines — and the spread from #1 to #4 is 2.5 percentage points:

Weaviate leads at 94.3%, Pinecone and Qdrant tie at 93.6%, Milvus sits at 91.7%. Four open-source-leaning infrastructure brands, near-identical AI footprints. Below them, Zilliz Cloud at 80.3% is the only other brand even in the same conversation, and the next named brand — Chroma at 64.3% — is 30 points back.

What this tells you: in emerging infrastructure categories, AI retrieval doesn't yet pick a winner. The models have indexed all four brands' documentation, blog posts, and benchmarks roughly equally, and none has a content moat large enough to break out. Whoever publishes the next round of authoritative comparison content — performance benchmarks, RAG cookbooks, framework integrations — has a real shot at separating from the pack. A 2.5-point race is movable.

---

SecurityScorecard's near-perfect score

SecurityScorecard surfaces in 95.5% of Third-Party Risk discovery prompts — 150 out of 157, across all four engines. That's the highest single-brand citation rate in any enterprise security category we track this week.

UpGuard follows at 87.9%, Prevalent at 84.1%, Bitsight at 83.4%. The top four are all genuine TPRM specialists — no generalist GRC tools or repurposed identity platforms — which means buyers asking AI about third-party risk get sent to the right kind of vendor every time. That itself is notable. In categories where buyer intent is less mature, you see HubSpot or Microsoft 365 show up as filler in the top ten. Not here.

A 95.5% rate isn't quite ZoomInfo-in-Sales-Intelligence territory, but it's the same shape of dominance over a larger prompt pool. SecurityScorecard's defensive position is strong. The interesting question is whether UpGuard can close the 7.6-point gap with content focused on the discovery-stage queries where SecurityScorecard currently runs alone.

---

eSignature is a four-way race

Last week we flagged Sales Engagement as the tightest race on the Citation Index — three brands separated by 2.8 points. eSignature just took the title. Four brands inside 1.3 points, on 155 discovery prompts:

Adobe Acrobat Sign and DocuSign tie at 65.2%. Dropbox Sign at 64.5%. PandaDoc at 63.9%. Salesforce — yes, in eSignature — comes in fifth at 54.8%, which is its own story.

eSignature is a mature category with stable feature parity, and the citation data reflects that. AI models have no consensus on a single "default" signature platform — they hedge across the top four every time a buyer asks. For DocuSign, that's a problem: a category-defining brand sitting at the same number as Dropbox Sign suggests its name recognition isn't translating into AI retrieval preference. For PandaDoc, it's an opportunity: being the #4 brand within a percentage point of the leaders, in a category where the leaders are stuck, is a winnable fight.

---

UiPath leads RPA — Make is closing on Zapier

Robotic Process Automation has a clear #1 and an interesting fight underneath it. UiPath surfaces in 80.6% of RPA discovery prompts — 18 points ahead of the next brand. That's a gap with staying power: UiPath's brand is so deeply associated with enterprise RPA that AI models consistently reach for it first regardless of company size or use case.

Microsoft Power Automate follows at 62.4% — expected, given its native position inside Microsoft 365 deployments. But the number that stands out is Make at 58.2% vs Zapier at 43.6%. Make (formerly Integromat) outpaces Zapier by 15 points in AI citations, a meaningful reversal of the traditional perception that Zapier owns the no-code automation space. AI models appear to be routing prosumer and mid-market automation buyers toward Make more reliably than Zapier — likely a reflection of Make's broader scenario support and the volume of technical content comparing the two.

For brands in adjacent categories (HR, Finance, Sales), this pattern matters: the automation platform your buyers are hearing about from AI models isn't necessarily the one with the largest install base.

---

Three giants, identical to four decimal places

Video Conferencing produced the most quietly remarkable number this week. Zoom, Microsoft 365, and Google Workspace each surface in exactly 64.97% of Video Conferencing discovery prompts. Same denominator (177 prompts). Same model diversity (all four engines). Identical shortlist rates to the fourth decimal.

This isn't coincidence — it's what perfectly-balanced category authority looks like. When buyers ask AI models about video conferencing, the answer is functionally always "Zoom, Microsoft Teams (via Microsoft 365), or Google Meet (via Google Workspace)." The three brands are bundled together so reliably that the AI engines no longer make a choice between them. They surface as a triplet.

Webex follows at 62.1%. The actual product names — Microsoft Teams (47.5%), Google Meet (47.5%) — sit lower than their suite-level parents. The lesson for B2B brands operating inside enterprise suites: the suite name dominates AI retrieval. If your product is bundled into Microsoft 365 or Google Workspace, the buyer never hears your product name first — they hear the suite. That changes the SEO-for-AI playbook entirely.

---

What we're watching next

Salesforce's eSignature visibility. Salesforce hit 54.8% in eSignature — a category it doesn't really compete in, on the strength of Salesforce-adjacent content alone. Same pattern we flagged last week in ABM and Marketing Automation, except now it's leaking into adjacent categories where Salesforce has no native product. That's either content overreach about to be corrected, or a sign that AI models are reading Salesforce's gravitational pull more broadly than its actual product surface area.

Whether the Vector Database cluster breaks. Four brands at 91–94% can't all stay there indefinitely. Watch for the first one to publish a breakout piece of authoritative content — benchmarks, integration depth, or a high-citation analyst report.

Make vs Zapier over the next month. The 15-point gap in RPA is striking. We'll track whether this holds across more automation-adjacent categories (HR tech, finance ops, sales ops) where both brands compete for the same buyer's attention.