Skip to content

Put Agents to work on your AI search data.

An Agent is a set of instructions Quotable Agent runs on its own: every week, or the moment a metric moves. Each run reads your AI search data and comes back as a report you can reply to.

Weekly visibility briefDone
  1. 1TrackingVisibility, mentions, share of voice7 days vs 7
  2. 2PromptsThe prompts that moved most21 prompts
  3. 3CitationsDomains that gained or lost12 sources
  4. 4CompetitorsWho took the answers you lost2 brands
  5. 5ReportThree actions for this weekDone
Every Monday · 09:00 · Read onlyVisibility held at 41%. Two comparison prompts went to a competitor.

Share of voice is flat week on week. The loss is on two comparison prompts in ChatGPT, where a review site now cites a competitor and not you.

Actions for this week

  • Update the comparison page2 prompts
  • Get listed on the review sitecited in 9 answers
  • Correct the pricing claim3 answers
Illustrative runReply to this run
  • Detus

Get started with ready-to-use Agents

Choose a template, read the instructions it comes with, and change the schedule or the wording before you save it.

  • Weekly visibility brief

    A Monday morning read on how AI answers treated your brand last week, and what to do about it.

    Every Monday

    Use template
  • Competitor watch

    Which competitor gained ground in AI answers this week, and the sources carrying them.

    Every Thursday

    Use template
  • Source gap sweep

    The domains AI engines cite for competitors but never for your brand: your outreach list.

    Every Tuesday

    Use template
  • Sentiment and facts check

    A Friday check on how AI engines describe your brand, and any claim that is wrong.

    Every Friday

    Use template
  • Monthly opportunities review

    On the first of the month: what the team shipped, whether it moved anything, and what is stuck.

    First of the month

    Use template
  • Visibility drop investigation

    Runs only when visibility falls. It finds the prompts, engines and sources behind the drop.

    When visibility drops

    Use template

If you can write the brief, you can set up the Agent.

Describe the work and Quotable Agent drafts the Agent for you. Or write the instructions yourself. Either way you decide when it runs and what it is allowed to change.

  1. Tell Quotable Agent what should run and when. It drafts the Agent and shows it to you. Nothing is saved until you create it, and you can edit the draft first.

  2. Run it every day, every weekday, every week or every month, at a set time in your workspace's time zone. Or start it when your visibility score, mention rate or share of voice moves.

  3. Choose Read only, Asks before changes, or Changes without asking. Attach up to three Skills, choose how deep each run goes, and decide whether the Agent may stop to ask you a question.

  4. Each run lands as a report in its own conversation. Reply to ask for more, open past runs to compare, or run the Agent by hand whenever you want a fresh read.

Quotable AgentDraft ready

Every Monday morning, tell me how AI answers treated us last week and what to do about it.

The draft

  • NameWeekly visibility brief
  • RunsEvery week, Monday 09:00
  • May changeRead only
Edit firstCreate Agent
Visibility drop investigationEditing

Schedule

Every week

On

Monday, 09:00

Or when a metric moves

Visibility score drops by 5 points against the 7-day average

Checked after each complete day of tracking.

Source gap sweepBehaviour

What it may change

  • Read onlyReports, changes nothing
  • Asks before changes
  • Changes without askingOff

Depth

Standard

Skills · 2 of 3

  • Brand voice
  • Claim correction
Weekly visibility briefDone
Mon 10 Nov · On schedule
Thu 6 Nov · Metric change
Mon 3 Nov · On schedule
Wed 29 Oct · Run now
4 past runsRun nowReply

Give your Agents the knowledge behind your brand.

Bring your positioning, product knowledge, and editorial guidelines into Quotable IQ. Give Agents the context to report on your brand accurately, in the terms your team uses and to the standards you set.

  • Brand CenterProfile
    PositioningAI search visibility for B2B
    AudiencesContent leadsSEO teams
    DifferentiatorsReal consumer AI answers

    Define your brand in Brand Center

    Set out your positioning, audiences, and differentiators. Give Agents a clear understanding of who you serve, what you offer, and how your brand should be represented.

  • Knowledge baseIn use
    Product information
    Company documents
    Approved claims
    Pricing and plans

    Bring your knowledge to every run

    Connect product information, company documents, and approved claims from your knowledge base. Give Agents trusted material to check answers against when they report on your brand.

  • SkillsApplied
    Brand voice
    Terminology
    Article structure
    Editorial rules

    Apply your standards with Skills

    Write your brand voice, terminology, and editorial rules as reusable instructions. Attach up to three Skills to an Agent, and every run follows them.

Run Agents in bulk with Tables

Process multiple pages, topics, or prompts in one table. Each row holds an input, and Agent columns perform the work, with outputs feeding the next stage.

Content refresh · Q332 rowsRun all columns
PageTopicResearchAgentBriefAgentDraftAgent
/ai-visibilityAI visibility tools12 sources · 4 gapsName the criteria buyers compare1,480 words
/brand-monitoringBrand monitoring9 sources · 3 gapsAdd the engines each alert covers1,210 words
/citationsAI citation tracking14 sources · 5 gapsExplain how a citation is countedWriting…
/guides/aeoAnswer engine strategy11 sources · 2 gapsDrafting brief…Queued
/prompt-volumesPrompt researchReading sources…QueuedQueued
/compare/profoundCompetitor comparisonQueuedQueuedQueued
Showing 6 of 322 drafts ready · 3 running

Read the run, reply, and stay in control

An Agent does not disappear into the background. Every run leaves a report, shows its state, and waits for you when the next step is yours to decide.

Source gap sweepNeeds you
  1. Read the cited sources
  2. Rank the gaps
  3. Write the report
  4. Question for you
ReportTen domains cite your competitors and never you3 realistic targets

Review sites too, or editorial only?

Open the runReply
Past runs
  • Tue 11 NovNeeds you
  • Tue 4 NovDone
  • Tue 28 OctDone
  • Tue 21 OctSkipped
  • Every run comes back as a report

    A run is a Quotable Agent conversation. Read what it found, see what it read, and reply to ask for more.

  • It stops when it needs you

    Let an Agent ask you a question, or make it ask before it changes anything. The run waits as Needs you until you answer.

  • The result finds you

    Get a notification when a run needs you, and turn on email to receive each report as it lands.

Put your AI search strategy into action.

Set up Agents that watch your visibility, competitors and sources for you, and report back with what changed and what to do next.

Frequently Asked

What is an Agent in Quotable?

An Agent is a set of saved instructions that Quotable Agent runs on its own, on a schedule or when a tracking metric moves. Each run reads your AI search data and comes back as a report you can reply to.

What can I use Agents for?

Work you would otherwise repeat by hand: a weekly visibility brief, a competitor watch, a sweep of the sources that cite competitors and never you, a check of how AI engines describe your brand, or an investigation that starts when visibility falls. Each run can read your tracking data, citations and sentiment, and search the web.

What is the difference between Agents and Quotable Agent?

Quotable Agent is the conversation you open to investigate a question. An Agent is work Quotable Agent repeats without being asked. Every run of an Agent is an ordinary Quotable Agent conversation, so you can reply to it and keep going.

Do I need coding skills to build an Agent?

No. Describe what should run and when, and Quotable Agent drafts the Agent for you. Nothing is saved until you create it. You can also write the instructions yourself or start from a template.

Can Agents run automatically?

Yes. An Agent runs every day, every weekday, every week or every month, at a set time in your workspace's time zone. It can also start when your visibility score, mention rate or share of voice drops, rises, falls below or rises above a level you set. You can run any Agent by hand as well.

Can an Agent change things without asking me?

Only if you allow it. Each Agent has one of three settings: Read only, Asks before changes, or Changes without asking. You can also let an Agent stop to ask you a question. A run that is waiting for you shows as Needs you.

How do Agents use my brand’s information?

Agents work from the brand context you keep in Quotable IQ, the same context Quotable Agent reads. You can also attach up to three Skills to an Agent, and each run follows them.

What are Tables?

Tables run Agents in bulk. Each row is an input, such as a page, topic, or prompt, and each column can be an Agent whose output feeds the next column. Every intermediate result is visible and editable.

What does an Agent run cost?

Runs draw on your plan's monthly credits. A deeper run uses more credits than a standard one. If the credits are used up, the run is skipped and shown as Skipped.

Does using Agents guarantee more AI citations?

No. Agents help you see what changed, find the sources and prompts that matter, and decide what to do next. AI platforms determine which sources they cite. You can monitor subsequent answers in Quotable to see how your mentions and citations change.