Eighteen jobs. One system. One login.

• –One line of script, on every product you run
• –Who is on the site now, by country or state
• –A limit your team can tighten for each product, and nothing can raise

Worth knowingIn the US a visit can be claimed as a person — on evidence they volunteered, or through an identity provider you connect and switch on yourself, on your own key — and a teammate approves it before it becomes a contact. In Australia, Canada, the EU, Japan, Korea and the UK the same is possible only for a visitor who turned Recognition on in the site's notice themselves. Everywhere else you get the company, or just the count.

• –One update reaches every site through the embed already in place
• –A scheduled check of each listed site's main address, with a review queue for what is running but not declared
• –A preference centre visitors can return to
• –Every choice kept with the version they saw

Worth knowingThe scheduled check reads a site's markup and headers on a few pages — it does not run the page, so a tag a tag manager loads at runtime is only governed once you declare it, and a site you have not listed is not checked at all. The ad industry's TCF string is built in but stays off until Runner's own CMP registration with IAB Europe is complete.

• –Accounts, contacts and opportunities
• –Owners, next actions and notes
• –Stage history, and for contacts from a connected system, the record they came from

Worth knowingReady-made record types and pipeline stages make Runner useful on day one — a place to keep the record, rather than a platform you have to build first.

• –Runner reads a company's own website and keeps what it says, with the source
• –Every claim lands with its source and the time it was read
• –A fit scorecard your team fills in, with the evidence behind each score

Worth knowingRunner does the reading on a research account you connect yourself, then files the evidence for your team to review. What comes back is evidence for a decision — never permission to contact anyone.

• –Lifecycle, consent, geography, recency and delivery history, from a set list
• –An estimated size before you commit to it
• –Per contact, the reason they matched
• –A sync plan against the Google Ads account you connect, to review

Worth knowingConditions come from a set list, in plain language, one condition per phrase — a phrase it does not recognise comes back to you. Runner preps a plan against the Google Ads account you connect yourself, and uploading from Runner is switched off.

• –One teammate gives the go-ahead before anything goes out
• –Sending windows, pacing and a daily cap on each mailbox
• –Opt-outs and bounces handled for you
• –Timed multi-step follow-up, in the journey engine

Worth knowingLive sending ships off for your organization and for each product, and an admin turns each one on, so your team chooses when automation begins. A platform switch sits above both, and while any of the three is off no campaign email leaves.

• –Hosted booking pages on the hours you set
• –Slot holds, timezones and daily limits
• –A team rota, sharing bookings out by turn or by who has waited longest
• –One script tag puts the booking inside your own page
• –Each booking kept as a meeting

Worth knowingA rota shares bookings out by two countable rules: each member in turn, or whoever has waited longest. Weighting one person's share, or ranking the team by past performance, sits outside it. The embedded booking takes your brand colour rather than a full theme.

• –A hosted form page, or your own site's form posting to Runner
• –Spam guards, and a list of the domains allowed to post
• –Consent recorded only when asked for, ticked and current, with the exact wording kept
• –A question that opens or hides based on an earlier answer
• –Answers a question was never put to are counted apart from answers left blank

Worth knowingA rule shows or hides a later question from one earlier answer, and that is the whole of the logic: no scoring, no routing on an answer, no fields calculated from one. Each form still hands the new lead to the email, chat or calling workflow your team chooses for follow-up.

• –One script tag, limited by default to the domains you list
• –It tells every visitor it is an AI
• –One tap asks for a person, and once a teammate takes over the AI stops replying

Worth knowingIt runs on an AI account you connect and arrives switched off until you turn it on; live replies also wait on Runner's platform switch and a production sign-off for the widget. When a pricing, legal, security or similar answer has no source, or the model isn't confident, it says a teammate will follow up rather than guessing, and no audio ever reaches Runner.

• –Plans and creates your GA4 property in your own Google account, then reads the setup back to check it
• –Reconciles it against your own first-party events
• –Holdout groups on the audiences you build

Worth knowingEvery number here is worked out from data you already hold — your own events, rather than a tracking script watching the rest of your tools.

• –Branded short links that count their clicks, with link-preview crawlers left out
• –A partner portal showing each partner their programs and codes
• –Commissions with hold periods and disputes

Worth knowingRunner works out what each partner is owed from the orders your team records. The money moves through whatever you already pay people with — never through Runner.

• –Duplicate pairs, each with the reasons it was flagged
• –Where both contacts sat on the same product, the more protective email, call and recording consent is kept
• –Each merge records the rows it moved, so undo can put them back

Worth knowingRunner brings each likely pair to a person — one reviewed pair at a time, never in bulk on a schedule. A contact undo puts the moved rows back unless one has changed since, and what was combined into the kept contact — consent, and a membership, tag, property or link both held — stays with it. An account undo moves its contacts, opportunities and evidence back as they are now.

• –Sends wait for a go-ahead from an owner or admin
• –Whoever prepared a send is refused as its approver, outside a one-person team or a time-boxed exception
• –Delegation for a set period, never beyond the delegator's own authority
• –One page listing decisions from several areas, with who decided and when
• –Every approval and rejection kept in an audit trail nothing in the product can edit or delete

Worth knowingThe decision history records what was decided; it never decides. Approvals are given by a person on the page that owns the work, the few entries the rules decide — such as whether a meeting was recorded — are marked automatic, and each entry links back.

• –An MCP endpoint your agent connects to, with OAuth
• –Permissions checked fresh on every call, never from a cached copy
• –Every tool is graded, with consequential work routed to a person
• –Each answer checked against the shape its tool declares before it goes back
• –A stop control for the workspace, a project, one tool or one person's connections

Worth knowingWhat an agent may call follows the permissions of the teammate it signed in as, checked again on every call, and anything irreversible or destructive stays off the list by design. A tool that changes a record is served only once it declares the go-ahead it needs, the evidence it carries and how it is undone, and that go-ahead is given by a person elsewhere, never over the connection — these are the controls around an agent, not an agent deciding for you.

• –Discord, Telegram, your chat widget, a connected email inbox, WhatsApp, and Messenger and Instagram DMs in one record
• –Replies to Messenger and Instagram inside Meta's 24-hour window, or with one of Meta's message tags after it — the composer states the rule rather than letting a send fail
• –Replies to WhatsApp inside the customer's own 24-hour window — the composer states the rule, and a message that would need an approved template goes through the outreach approval path instead
• –SLA tracking and next-action ownership per conversation
• –Human handoff from any channel

Worth knowingMessenger and Instagram messages arrive once your Page or account has passed Meta's messaging review and you have recorded it in Runner; until then the connection verifies each delivery and stores nothing. WhatsApp messages arrive as soon as your business number is connected and delivering, and a reply leaves the inbox while the customer's own 24-hour window is open, to a contact whose opt-in you have recorded; after that window closes WhatsApp accepts only an approved template, which Runner sends from the outreach approval path rather than from the inbox. LinkedIn is connected for lead forms and publishing only — LinkedIn offers no inbox API to a product like this, so nothing from LinkedIn messages is read or kept.

• –Every stage counted against one denominator, not five different ones
• –Open any stage and see the deals inside it, with their accounts
• –Outcomes recorded when a deal closes, win or lose

Worth knowingThe counts use six fixed stages and the deals your team records: no predicted close dates and no win probabilities, which is exactly why the numbers reconcile. An inbound lead can carry one labelled suggestion beside it — Runner's reading of a named, versioned qualification policy, with its reasons shown — and that suggestion never moves a stage, a count or a total.

• –Contacts with their email consent and opt-outs, in one versioned export
• –Where each contact came from, carried with it
• –A read-only snapshot that leaves the live record untouched

Worth knowingOne export covers one product's contacts, read in a single request — check its row count against your directory — each carrying the email consent state already on its record — including unknown where no basis was recorded — and any opt-out. Permission for calls and for each messaging channel is kept separately and stays out of the file.

• –It announces itself when it joins, in wording that stays the same on every call
• –Notes land on the same record every other part of Runner reads from
• –A teammate accepts each agreed item before it becomes a next step

Worth knowingRunning on Runner's own calls today, with customer access next. The joining announcement is signed off for the internal pilot while the customer version completes review.