The weekly brief in Slack
The brief already assembles itself and already lands in the workspace inbox every Monday. What is not built is the second channel: the same brief posted into Slack.
Resonue is not a dashboard you have to remember. Three agents run on a clock: one sweeps every voice talking about your company every six hours, one rescores every lead and every partner overnight, one assembles Monday's brief and emails it to your team before anyone is awake.
Resonue structures your resonance. Then turns it into revenue.
The agents watch, score and draft. A person still hits publish, and a person still hits send. That line does not move.
Illustrative log: the timestamps, the counts and the workspace are made up. The cadences are the real ones: a sweep every six hours, a rescore nightly at 03:00 UTC, the brief every Monday at 08:00 UTC, all running in production today. If a run fails it retries twice more, a minute apart, then reports itself as failed rather than going quiet.
Resonue is not a place you go to do the work. It is three named agents that do the work, on a clock, in your workspace, and hand you the result. They are processes, not screens: there is no agent list to open, nothing to configure per agent, no switch to flip. What you get is what they left behind. Here is every one of them, what sets it off, what it is allowed to read, what it writes down, and where it stops.
Goes and looks, four times a day, whether or not anyone is in the app.
0 */6 * * * UTC, the real expression. Nobody presses anything.Redoes the arithmetic on everything, every night, so nothing on your screen is a stale number from last month.
0 3 * * * UTC, once a day, per workspace.Assembles the week into one email and sends it to the workspace. Shipped, and running in production right now.
0 8 * * 1 UTC, before anyone opens anything.Turning reactions into named people costs money on every provider call, so the capture agent is not on a clock. You decide when it runs, from the voice or the post you are looking at. After that nobody touches it again.
On a voice, or on a single post. That is the whole human contribution.
Only posts whose engagement actually grew since the last look. It caps itself at ten posts a run and tells you how many are left.
Every like, comment and repost turned into a named person and the company they work for, deduplicated against what it already had.
Only profiles it has never seen before, so it never pays twice for the same person.
Heat and ICP grade computed on the spot, then kept fresh by the scoring agent from then on.
Four sweeps, one nightly rescore, and once a week an email that writes itself. That is the whole clock, and the three expressions below it are the ones actually deployed, not an illustration of them. There is nothing here you have to remember to do.
A Voice is any account whose posts talk about your company: your own company page, your employees, outside people mentioning you unprompted, or contracted content partners. You define the roster once. The watch agent then covers all four, together, as one scored signal, which is something almost nobody measures because almost nobody can.
Point a tool at the company page alone and it is measuring a fraction of what is actually resonating. The roster is the scope, and the scope is yours to set.
Company page, employees, external voices and contracted partners, tracked together. Compliance against the cadence you agreed, engagement against your own workspace average, brand mention rate. Nobody refreshes any of it. The scoring agent does, and the status pill on the roster is simply what it found.
Drill into any voice and the verdict row shows the arithmetic behind the grade, not a summary of it.
Share of voice, share of topic and share of sentiment against the competitors you track, computed from the same captured data as everything else. Topics and hashtags tiered against your own baseline rather than against a generic best practice. No agent is writing you a paragraph about the market. It is doing arithmetic on what actually happened.
The same ten people posting about the same launch on the same day make more noise than those same ten people posting once a month all year. You set the window and the priority. From there it is automatic: every post captured in that window gets matched back to the brief it covered, and the engagers it produced roll up under it.
Rosa reads what the agents wrote overnight and surfaces it where there is actually a decision to make: one card at the top of your dashboard, the engager you just opened, the draft you just asked for. Not a chat box bolted to every screen waiting for you to think of a question.
Reason crossed with resonate. It is the word the product uses for the moment Rosa is working, and it is deliberate: she is not fetching, she is working out what a pattern in your own data means. It always comes with what she is doing underneath it, never a bare spinner.
What comes back is a finding, the reason for the finding, and one thing you can do about it. Your dashboard opens on exactly one of these, the most important thing in the workspace this morning. On a slow week it says “Quiet morning. Nothing urgent.” and leaves you alone, which is the part most tools cannot bring themselves to do.
Before a post goes out, Rosa scores the draft against your workspace's own resonance history: topic fit measured on the topics that already worked for you, plus hook strength, format fit and tone match on the draft itself. She can also generate a starter draft for one voice from that person's own best posts and recurring topics.
The goal is never the post that collects the most likes. It is the post built to land with the accounts already warming up.
An agentic product is only as good as its stopping conditions. These are Resonue's. They are not a setting you can loosen, and they are not a phase we intend to grow out of.
Rosa scores a draft and can write a starter one from a voice's own best posts. Resonue has no write access to LinkedIn at all: the integration interface has five methods and every one of them reads. The product could not publish if we asked it to.
Rosa turns a hot engager into an opener built from what they actually engaged with. Under 700 characters, no greeting flourish, no signature, no buzzwords, and if the only signal is one old like she says so plainly instead of manufacturing enthusiasm. Then it sits there until you copy it.
Every grade on every screen is measured against an ICP you wrote in your own words. The agents never decide who your customer is. Change the definition and the next nightly run regrades everything against it.
Every Voice talking about you throws off an echo. On its own it is noise on a feed. The agents read the echo, score it, and stand it up as pipeline you can actually work.
One like means nothing on its own. A pattern across people, posts and months means a buying committee is forming, and that pattern only exists if something has been watching every night for a quarter.
A like is worth 1, a comment 3, a repost 5, and every one of them halves in value every 30 days. Anything older than six months drops out entirely. Because it is recomputed nightly rather than when you happen to open a page, an engager who has gone quiet cools off with nobody doing anything about it.
Every engager graded A, B, C or off ICP against the profile you defined once. New people captured in the night are graded before you see them.
4 or more people from one account engaging across 3 or more posts in a quarter get flagged and rolled up by company. Nobody has to run the query.
Warm accounts roll onto a pipeline board with deal stage and influenced revenue attached, and a 30 day momentum report you can take straight into a pipeline review. The point of the agents is that this board is current when you open it, not when you remember to refresh it.
This is the one artifact the system produces without being asked. It pulls both halves together, the partners on one side and the accounts on the other, and it arrives whether or not anybody logged in that week.
Partners needing attention is everyone Slipping or Off contract, most urgent first. Partners to amplify is the honored ones with the strongest 30 day engagement, worth doubling down on. Top warm accounts is the buying committees, most people first. Accounts re-engaging is the one nobody else is looking for: a customer you already sold, or one you already lost, that has quietly started engaging again.
If a workspace has nothing in any of the four, the brief does not send. Silence is a signal too, and it is better than a weekly email nobody reads.
The same brief in Slack is not built yet. It is on the board below.
If something is going to run in your account four times a day without asking, you are entitled to a flat list of its permissions before you buy it. Here is ours, and the third column is the longest on purpose.
Scope you set, on day one, in one screen.
All of it inside your workspace, all of it openable.
Not a setting. Not a plan tier. Not built.
Rosa is a copilot, not an autopilot. The line is architectural, not a policy we could quietly move.
The nightly run does not only produce a number, it stores the raw inputs beside it: posts in the window, posts the cadence expected, this partner's average engagement, and the strongest partner's, which is what it is measured against.
Open any engager and the score comes apart into the events that made it: the date, whether it was a like, a comment or a repost, its weight, and how much the 30 day half life has already taken off it.
A health endpoint reports the last sweep, the last rescore, how many voices were swept and how many posts were ingested in the past 24 hours, and flips itself to stale if an agent missed its window. Aggregates only, no personal data.
Every paid call the system makes, to the data provider and to the model, is logged through a single choke point with its tokens and its duration. An agent you cannot audit is just a background process with better marketing.
Illustrative figures in both panels. The formulas, the weights and the 30 day half life are the real ones.
The same agents feed all three. Marketing reads what the presence produced, Revenue Ops reads the scored accounts, Sales reads the shortlist. Same data, two acts, one system.
Stop defending your budget with follower counts. A real number on which voices, topics and posts are producing pipeline, refreshed overnight rather than pulled together the day before the board meeting.
See how Marketing uses Resonue → Revenue act For Revenue OpsEvery engager scored by heat and ICP fit every night, buying committees flagged before sales notices, and an influenced pipeline number you can defend on a board deck.
See how Revenue Ops uses Resonue → Revenue act For SalesA ranked list of who engaged with your own posts, a drafted opener pulled from the exact post they reacted to, and a nudge to post again before the reach dries up. You still send it.
See how Sales uses Resonue →Everything else on this page is built and running right now, every agent on it included. This is what the team is working on next, sorted by where the work actually stands, so you never have to guess which is which. Roadmap, not a commitment: order and scope move as we learn.
The brief already assembles itself and already lands in the workspace inbox every Monday. What is not built is the second channel: the same brief posted into Slack.
An account you already lost or already sold that quietly starts engaging again is in Monday's brief today. What is coming is surfacing it on the pipeline board too, not only in the email.
That is the product you would be buying today. Everything else on this board is what comes after it.
Heat score and ICP grade written straight onto the lead and contact records your reps already live in.
Bring in the account list you are already chasing and see what share of it your content has genuinely reached. Today that arithmetic happens in your spreadsheet.
Grading is per engager today. Next: the account level read, plus what each account actually engages with.
A retainer field on the contract, so the worth it math stops living in your head and starts living in the product.
Decide inside the brief which voices carry which message, instead of deciding it in a separate doc.
Cadence, topic remit, exclusivity and expiry, agreed and tracked in Resonue. Today a contract is cadence and dates.
Their own login: what they owe you this month, what to talk about, how their posts are doing.
Draft scoring learned from your own best posts rather than from rules and a rubric.
Reach, movement against a voice's own baseline, brand fit and contract, weighted deliberately. The weighting is still an open product decision.
Pay content partners against the targets they hit, without leaving the console. Furthest out of anything here.
Resonue is in early access. We're working with our first design partners now.
No, and it never will be. Rosa drafts, scores and explains. She has never had a publish button, and there is no path in the product that puts a post out or sends a message to a prospect. Resonue is intelligence and piloting, not marketing automation. A person always decides what goes out.
Yes. The six hourly sweep, the nightly rescore at 03:00 UTC and the Monday 08:00 UTC brief are scheduled jobs running per workspace in production today. A failed run retries twice more, a minute apart, and a health endpoint marks the system stale if a job misses its window. The app is where you go to look and to decide, not where the work happens.
Every score is arithmetic you can open, not a verdict you have to accept. The nightly run stores the raw inputs behind each partner score, a heat score comes apart into the individual events that produced it, and an ICP grade shows you which dimensions matched and which did not. If a grade looks wrong to you, the usual cause is the ICP definition, which you wrote and can rewrite. Saving a new one regrades your engagers against it.
Not today. There is no public API and no CRM connection in the product right now, and the integrations page says Soon rather than pretending otherwise. CRM sync for HubSpot and Salesforce is on the board above, designed and queued, not shipped. The one HTTP surface that does exist is the health endpoint above, and all it will tell you is whether the jobs ran.
No. Resonue connects to your company's LinkedIn presence, your company page, the employees who opt in, and outside voices who mention you, the same way an analytics or social management tool would. Nothing is read off private profiles, and nothing is published on your behalf. The permission ledger above is the full list.
No. Starter covers your company page, your employees, and outside voices already talking about you, and it's the only paid tool that channel needs. If you do run a partner program, Studio includes the whole cockpit: contracts, compliance, and per-partner pipeline.
The sweeps start the day you connect your voices. The first rescore is that night. The first brief is the Monday after, in your inbox, whether or not anyone opened the app. Structure the resonance. Turn it into revenue.