HubSpot AEO vs Otterly: Measuring and Acting on Citations

HubSpot AEO vs Otterly is a choice between two monitors, and the better one for measuring AI visibility is the one that covers the engines your buyers use and lets you act on each cited URL. HubSpot AEO tracks three engines and Otterly four on its published plans. Noma by Nurdd tracks 13 AI engines, including ChatGPT, Gemini, Perplexity and Claude, and reports each engine separately.

Key takeaways

• HubSpot AEO covers ChatGPT, Gemini and Perplexity from $50 a month for 25 prompts. Otterly Lite is $29 a month for 15 prompts on four engines. Both were read on 2026-11-09.
• A citation list is not a plan. Triage each cited URL by who owns the page, whether your brand is named and whether the pattern repeats across runs.
• One missing citation is not a finding. Parse measured that consecutive cited answers shared only about a third of their hostnames on ChatGPT Search, so act on patterns across repeat runs.
• Compare tools on price per prompt, and separately check which engines the base price covers. Prompt limits decide which plan you need.
• Before you buy, run the same 20 prompts on each trial and check that every cited URL is clickable, exportable and tied to a prompt, an engine and a date.

What do HubSpot AEO and Otterly actually measure?

HubSpot describes HubSpot AEO (labelled beta on its product page) as tracking a brand across ChatGPT, Gemini and Perplexity, with a visibility score, share of voice and prioritised recommendations. Its own comparison article gives a 28-day trial with 25 prompts for HubSpot AEO.

Otterly's pricing page lists tracking of four AI search engines on its plans: ChatGPT, Google AI Overviews, Perplexity and Microsoft Copilot. It sells Claude, Google AI Mode and Gemini as extra add-ons. Its plans list brand visibility and sentiment analysis, citation source tracking, brand profiling and AI search optimisation with recommendations. Standard and above add API access, MCP access and agent analytics, and the page lists crawlability and content check audits. Our Otterly.AI review covers it in more detail.

The table compares what each vendor's own pages state, as checked on 2026-11-09. Where a cell says "not confirmed", the page we read did not say, which is not the same as the feature being absent.

Sources: HubSpot product page, HubSpot blog, Otterly pricing and Noma pricing, all as checked on 2026-11-09.

Whatever a monitor shows, the question content teams keep asking is what to change once the citation data arrives. The section on acting on citations below answers it.

How do the plans compare on price and prompt limits?

HubSpot AEO is $50 a month for 25 prompts. Its product page says you can buy more prompts anytime, but it publishes no add-on price. A HubSpot blog roundup states AEO is included in Marketing Hub Professional and Enterprise. Otterly's pricing page lists Lite at $29 for 15 prompts, Standard at $189 for 100 prompts and Premium at $489 for 400 prompts. Prices are in USD, read on 2026-11-09.

Price per prompt is the monthly price divided by the prompt limit. It ignores add-on engines (Otterly and Noma both sell extra engines) and extra prompt packs, so treat it as a first filter, not a verdict. For other vendors, see our AI visibility tool pricing comparison, which lists twelve tools with the date each price was read.

Worked example: tracking 50 prompts

A content team wants 50 prompts, a realistic set for one product line with several buyer intents. On HubSpot AEO, 50 prompts means buying beyond the 25 included, and HubSpot publishes no price for that, so the cost cannot be calculated from public pages. On Otterly, 50 prompts does not fit Lite's 15, so the team moves to Standard at $189 for 100 prompts. On Noma by Nurdd, Starter at $79 includes 50 tracked prompts across four engines.

That is $79 against $189 for the same 50 prompts on published plans. Standard's extra 50 prompts are headroom you may not need yet, and Standard also brings API, MCP and agent analytics. Size your set first with the prompt scoring sheet, then price every tool at that number.

HubSpot AEO vs Otterly: which is better for measuring AI visibility?

The better tool for measuring AI visibility is the one that covers the engines your buyers use and keeps each engine's result separate. A score pooled across several engines can look healthy while one engine names a competitor. The engines do differ: our write-up on engines naming different top brands shows how to measure it.

Noma by Nurdd is the stronger choice on these rows:

• Engine range: Noma tracks 13 AI engines, including ChatGPT, Gemini, Perplexity and Claude, so you can add the engines your buyers use as your prompt set matures.
• Per-engine reporting: visibility, position and sentiment reported separately for each engine, so a drop on one engine is not averaged away.
• Prompts at the entry tier: 50 prompts at $79, against 25 at $50 for HubSpot AEO and 15 at $29 for Otterly Lite.
• MCP at the entry tier: Noma connects Claude, Claude Code and Cursor through its MCP server on every paid plan, from $79. Otterly lists MCP and API access from Standard at $189.
• From gap to draft: Noma finds prompts where competitors appear and you do not, then drafts the page that would answer them.

On citation source tracking, sentiment and site audits, Otterly's page lists comparable features, so those are not rows to decide on. Test them in a trial instead. Noma's own breakdown against suites is in Noma vs Semrush vs Ahrefs, and a wider by-job table is in AI visibility tool comparison.

Why do citations change between runs?

Engines rebuild each answer, so cited sources move even when nothing on your site changed. Parse reported that consecutive cited answers on ChatGPT Search shared an average of 33.3% of their combined hostnames, and 24.7% on Google AI Mode, in its volatility research, October 2026.

Noma's citation half-life analysis of 2,545 answers from ChatGPT, Gemini, Perplexity and Claude found that two in three cited URLs appeared in one run only. The practical rule is the same: run each prompt at least three times before you assign any work, and act on patterns that repeat. For a fuller treatment, read AI citation stability.

This also tells you what to demand from a tool. A tool that samples once a week gives you noise. A tool that runs the same prompts daily from fresh sessions lets you see which citations persist.

How do I act on AI citation data for my content?

Treat every cited URL as a row with three attributes: who owns the page, whether your brand is named in the answer, and whether the citation repeats. The action follows from those attributes, not from a dashboard score. The triage method below is our own, built to turn a list of URLs into a work queue.

Step 1: split mention from citation

A brand can be named without a link, and a page can be cited without the brand being named. Citation versus brand mention explains why the two need separate counts. Record both columns for every prompt and engine.

Step 2: classify each URL by owner

Sort cited URLs into yours, a competitor's, third-party editorial, and community. Counts by owner tell you where the work sits. If most citations are third-party lists, rewriting your own pages will not move the result; earning a listing will. Our guides on getting listed on third-party lists and earned media for AI answers cover that route.

Step 3: check your own pages can be read

Google states that "optimizing for generative AI search is optimizing for the search experience, and thus still SEO" in its Search Central guide to generative AI features. A page that is not crawlable or clear will not win a citation either. Run it through the free Noma AI SEO Checker, which shows whether ChatGPT, Claude, Perplexity and Google AI Mode can read the page, with its top fixes.

Step 4: change one thing, then re-measure

Change one page, wait for a recrawl, rerun the same prompts and compare per engine. Changing five things at once leaves you unable to say what worked.

Step 5: read referrals as a lower bound

Clicks alone undercount the effect. Pew Research found users clicked a traditional result in 8% of visits with an AI summary against 15% without, and clicked a link inside the summary in 1% of visits, per Pew Research Center, July 2025. Measure mentions and citations, and treat referrals as a floor. See GEO referral traffic measurement for a four-layer plan.

What should I test before choosing an AI visibility tool?

Run the same 20 prompts on each trial and score the output on the checks below. For how to pick the 20, use the by-job comparison guide.

1. Engine match. List the engines your buyers use. Count how many the base plan covers and what add-ons cost.
2. Prompt-to-URL traceability. Pick one answer. Can you see the prompt, the engine, the date and every cited URL on one screen?
3. Repeat runs. Does the tool sample each prompt daily, or once? One-off checks cannot separate noise from change.
4. Owned versus earned split. Can you filter cited sources by your domain, competitors and third parties?
5. Export. Pull the cited URLs into a spreadsheet. Check for API or MCP access and which plan includes it.
6. Next action. Pick a gap. Does the tool give you a brief or draft, or only a score?
7. Cost at your prompt count. Price the plan you will really need, including add-ons, not the entry tier.

Trial lengths differ. HubSpot's blog states 28 days for HubSpot AEO. Otterly's pricing page says it offers a free trial but shows no length. Noma's trial is 3 days with 25 prompts. Prepare the 20 prompts and a spreadsheet before you start, so a short trial is enough to run the checks. To measure share of voice during the test, follow how to measure AI share of voice across engines.

Where Noma fits

Noma by Nurdd asks 13 AI engines, including ChatGPT, Gemini, Perplexity and Claude, the questions your buyers ask every day. It reports visibility, position and sentiment per engine, lists the sources each answer cites, shows competitor share of voice and finds topics where rivals appear and you do not. For content teams, it then drafts the page that would answer the gap.

See it on your own brand: book a demo or start the 3-day free trial. A card is taken at signup so billing can start on day 3; cancel before then and nothing is charged.

How we researched this

We reviewed HubSpot's AEO product page and two HubSpot blog articles, Otterly's pricing page, Google Search Central documentation, Pew Research Center and published citation-volatility research. Facts and prices were checked on 2026-11-09. Where a vendor's public page did not state a figure, such as Otterly's trial length or HubSpot's prompt add-on price, we left it out or marked it not confirmed. We did not run hands-on tests of HubSpot AEO or Otterly.

What to do next

Take the 20-prompt test to each trial, then apply the triage table to the cited URLs you get back. To run the same test across 13 AI engines on your own brand, book a demo or start the free trial.