The SDR tools
Ten thousand people, one template
AI SDR tools send the same message to a list and burn your domain doing it. When you ask why a particular person was picked, there is no answer to read.
For funded startups doing their own outreach
Outreach agents you can actually see inside.
One engine finds the reason to reach out, drafts in your voice, sends with tracking and learns from every reply. You see every run, you own every instruction, and you keep everything if you ever leave.
No credit card. Starters are free and yours.
PROpportunity found
Feature: "The quiet consolidation of warehouse robotics"
National business title, published 41 minutes ago
91
Fit score
Why this person, right now
J. OkaforSenior reporter, supply chain and logistics
- Has covered three stories in your category in the last fortnight.
- This piece asks a question your February deployment data answers.
- Never pitched by you before, so nothing to apologise for.
Drafted in your voice
The number behind the consolidation story
Hi J,
Your piece this morning asks who actually has the deployment data. We do, across 40 sites, and the pattern is the opposite of what the analysts told you.
Happy to send the raw figures, or put our head of operations on a call today if it is useful before you file again.
Approve and sendRequest changesReject
Founders' work featured in
The situation
You have a story. You have a list. You do not have the hours.
The PR cycle
Feast, then silence
A launch, a scramble, a week of pitching, then nothing until the next one. Meanwhile the journalist writing about your category today moves on tomorrow.
The tools
You leave with nothing
Every tool you try keeps your prompts, your data and everything it learned about your market inside its own walls. Cancel, and you are back where you started.
The engine
One engine. Sales, PR and search, in your voice.
The same five steps run behind every agent on the platform. Change the job and the inputs change; the loop does not.
01
Listen
Agents watch news, funding, hiring, podcasts, social and AI search all day for reasons to reach out.
02
Score
Every opportunity gets a fit score against your brand, your proof points and your goals. The reasoning is on the card.
03
Draft
The message is written in your voice from your knowledge base, not from a template with a name slotted in.
04
Approve and send
Nothing goes out without a human click. Sends leave from your own Gmail or Outlook, with open, click and reply tracking.
05
Learn
Replies, meetings and coverage feed back into the agent. Each version is better than the last, and the changelog says why.
Sales
Prospects who just raised, hired, launched or posted about the problem you solve.
View the starterPR
Journalists writing about your category right now, with the angle and the draft attached.
View the starterOwn it
Your instructions. Your data. Your outputs. Forever.
Everything else on this page is table stakes eventually. This is the part nobody else offers you.
01
Read every agent
Every agent has instructions you can open, edit and version. Starters are fully open. Rented agents show you exactly what they do, what they take in and what they produce, with a changelog for every version.
02
Fork anything
Tell the workbench what you want, and it drafts the fork, asks the two questions it cannot answer for you, and saves the result as version one of your agent.
03
See every run
Every run is logged: what it looked at, what it scored, what it wrote and what happened next. Managers get one view across the whole team.
04
Leave with everything
Cancel and you export every instruction you wrote, every fork, every output and every result, as files you can actually use. Our rented cores stop. Everything you made stays yours.
Our ownership policy
Your instructions, your data, your outputs and your forks are yours, forever. Our rented cores are ours, while you subscribe. Starters are free and yours.
The catalogue
Every agent has a starter. Rent the tuned one when you are ready.
A starter is a fully readable, fully yours version of each agent, good enough to run today. A rented agent is the same job with our tuned core underneath: verified contact data, fit scoring, outcome learning and tracking. Managed-only agents are ones we run for customers but have not productised yet.
Service line
Showing 28 of 28 agents
Lead Research
SalesResearches a small number of high-quality leads at a time against your product and your ideal customer profile, and proposes refinements to that profile as it learns.
Managed only
Contact Finder
SalesFinds and verifies the email address for every lead in your CRM, checks the ones you already have, and marks the ones that will bounce before you send.
Managed only
LinkedIn Outreach
SalesRuns paced LinkedIn outreach from your own account, with human-approved sends only and account-safety limits built in.
Managed only
Network Miner
SalesWalks the connections you already have, screens them against your ideal customer profile and your competitor list, and files the ones worth a conversation.
Managed only
Sales Strategist
SalesReads what your outbound is actually doing, treats your own edits to the copy as ground truth, and sets the next set of experiments.
Managed only
Sales Copywriter
SalesWrites the message variants your strategist asked for, and captures what changed every time a human edits a draft.
Managed only
Sales Analyst
SalesCalls winners on running experiments only once there is enough data to justify it, with guards against noise and uneven splits.
Managed only
PR Proactive Discovery
PRHunts the open web for stories your feeds missed, derived from your own positioning, and files the new ones fully enriched.
Managed only
PR Opportunity Enrichment
PRTakes each story in your queue, works out the angle, finds the journalist behind it and drafts the pitch.
Managed only
PR Expert Vetting
PRA senior strategist pass before anything reaches you: kills the weak pitches, rewrites the rest to a sendable standard, and resolves the recipient.
Managed only
PR Pitch Correction
PRSecond pass over vetted pitches: re-reads the source article, removes anything unverified and repairs the copy.
Managed only
Journalist Email Finder
PRFinds the missing email address for every journalist you have a drafted pitch for, and remembers the patterns that worked.
Managed only
Podcast Deep Research
PRDeep research on the shows worth your time: the audience, the booking route, the application process and a drafted approach.
Managed only
PR Director, Daily
PRSpots what is decaying today, fresh journalist matches, high-scoring stories, replies waiting, and names the single highest-value move.
Managed only
PR Director, Weekly
PRJudges the week against your real PR goal, diagnoses the biggest bottleneck and issues ranked instructions to the agents below it.
Managed only
PR Director, Monthly
PRMeasures the month against the last one across coverage and AI search, grades it, and writes the narrative behind the numbers.
Managed only
PR Director, Quarterly
PRThrows out the assumption that the current strategy is right, re-derives the objective from first principles and names the big bets.
Managed only
GEO Audit
GEOBuilds the questions your buyers actually ask, runs them on the answer engines in fresh sessions, and reports how often you are named and cited.
Managed only
GEO Content
GEOTurns the audit's fix list into drafted, evidence-backed pages and posts, ready for you to approve.
Managed only
Competitor News and Hiring
CIUWatches your rivals' announcements, hires, partnerships and pricing moves, and reads each one for what it means for you.
Managed only
Competitor Socials
CIUReads what your competitors post and what lands, weighted so the rivals that matter most get the most attention.
Managed only
Brand and Competitor Mentions
CIUSweeps LinkedIn, X, Instagram, Threads, Bluesky and Reddit for mentions of you and of the companies you are measured against.
Managed only
Competitor Triage
CIUReviews every newly detected competitor against your own context and decides which ones are genuinely yours to track.
Managed only
Competitor Vetting
CIUResearches each candidate competitor properly before it enters your set, and merges the duplicates.
Managed only
Competitive Strategist
CIUReads everything the competitive agents produced this week and turns it into a position: what changed, and what you should do about it.
Managed only
Executive Brief
CIUThe weekly one-page read for whoever has to make the call, drawn from every source the fleet touched.
Managed only
Email Experimentation Lead
GrowthOwns outbound email end to end: segments, sequences, test design, measurement and deliverability, proposing rather than sending.
Managed only
Traffic Data Analyst
GrowthKeeps your funnel numbers honest: validates the tracking, reconciles platform figures against your own, and retracts anything it got wrong.
Managed only
Every agent here runs today. The pill says how you can get it right now. Ask us about anything marked managed only.
Where it runs
Runs in our cloud. Reachable from yours.
Gmail and Outlook
Sends and tracking run through your own accounts, so replies land where you already read them and your reply history stays in one place.
MCP access
Run, list, approve and export from your own Claude or ChatGPT. Force a manual run whenever you like, without opening the workbench.
Bring your own keys
Run agents on your own model API keys if you would rather pay the labs directly and see the spend on your own bill.
Export on demand
Everything, as files, any time. Not only when you leave, and not as a support ticket you have to chase.
Managed onboarding
Set up properly in two weeks. Then it is yours to run.
We build your knowledge base, tune the voice, connect your mail, pick the right agents and run the first campaigns with you. At the end you can run it yourself, keep us on for monthly tune-ups, or hand the whole thing to us.
Setup sprint
Priced per scope
Two weeks, fixed scope
Knowledge base built, voice tuned, mail connected, agents chosen, first campaigns run with you. You watch every step in the workbench.
Monthly tune-up
£599
Pro plan, plus tune-up
We review what the outcomes actually say, propose new agent versions, apply the ones you approve, and report on what changed.
Fully managed
£2,999
per brand, per month
We plan it, run it and report on it. You still approve every send, and every run is still in your workbench to read.
Pricing
One brand, unlimited seats, three ways in.
Starter
£199
per brand, per month
All starters, one rented agent, approve and send, export.
For a founder doing it alone.
- Every starter agent, readable and yours
- One rented agent of your choosing
- Human approval on every send
- Full export whenever you want it
You own your instructions, data and outputs on every plan.
Pro
Most chosen£599
per brand, per month
All rented agents across Sales, PR and GEO, forks, team oversight, MCP access, your own model keys.
For a growth team.
- Every rented agent across Sales, PR and GEO
- Fork and version any agent you like
- One run view across the whole team
- MCP access and bring your own model keys
You own your instructions, data and outputs on every plan.
Managed
£2,999
per brand, per month
Pro plus the setup sprint and monthly tune-ups, or the whole thing run for you.
For a team that wants outcomes without owning the process.
- Everything in Pro
- Two-week setup sprint
- Monthly tune-ups with new agent versions applied
- Or fully managed: we plan, run and report
You own your instructions, data and outputs on every plan.
Compare honestly
How this compares
Where a category is mixed, or where we have not checked every tool in it ourselves, the mark says partly rather than no.
| Capability | Big Ears | AI SDR tools | AI PR tools | Doing it yourself |
|---|---|---|---|---|
| Finds timely reasons to reach out | Yes | Partly | Partly | Partly |
| Writes in your voice from your knowledge | Yes | Partly | Partly | Yes |
| You can read the agent's instructions | Yes | Partly | Partly | Yes |
| You can fork and version agents | Yes | Partly | Partly | Yes |
| Human approves every send | Yes | Partly | Partly | Yes |
| Sends through your own mail | Yes | Yes | Partly | Yes |
| Learns from replies and coverage | Yes | Partly | Partly | Partly |
| Export everything if you leave | Yes | Partly | Partly | Yes |
| Time to first send | Same day | Days | Days to weeks | Weeks |
- Yes
- Partly, or varies by tool
- No
Proof
What it looks like when it lands.
Within 10 minutes of using it, I did 10 emails and I got a lead booked.
You helped me find different opportunities that I hadn't known about. The pitch suggestions take angles I wouldn't have thought of.
Last 30 days
692
messages a human read, approved and sent from Big Ears across every customer, with 38 replies tracked back against them.
Security and data
Built for people who read the security page.
Separated by row-level security
Every workspace is isolated in the database itself, not by application code that could forget. One client's content never reaches another client's queries.
Your data trains nobody's model
We do not use your data to train a model, ours or anyone else's, and we do not use one client's material to serve another.
Model calls go to Anthropic
Or to whichever lab issued the API keys you brought, if you would rather pay them directly and see the spend on your own bill.
Hosted in the EU
Our database runs in Supabase eu-west-2. Deletion on request, export on demand, and a written record of both.
Nothing sends itself
Every outbound message waits for a human approval. There is no mode in which an agent decides to email someone on its own.
Full run logs
What each agent looked at, what it scored, what it wrote and what happened next, kept and readable rather than summarised away.
Questions
The things startups ask us.
What is the difference between a starter and a rented agent?
A starter is the whole agent, written out, free, and yours to keep and change. It does the job well enough to run today. A rented agent does the same job with our tuned core underneath: verified contact data, fit scoring against outcomes we have already learned from, and the tracking that feeds the learning loop. You rent the core monthly and stop renting whenever you like.
Can I see the prompt behind a rented agent?
You see the instructions layer, the inputs, the outputs and the changelog for every version. The core is ours while you rent it. Everything you write on top of it is yours, and stays yours after you stop.
What happens if I cancel?
You export every instruction you wrote, every fork you made, every output and every result, as files you can open in something else. Our rented cores stop working. Nothing you made is held back, and you do not have to ask a support agent to release it.
Will this burn my domain?
No, because it is not built to send at volume. Every message waits for a human approval, sends leave from your own mailbox rather than a shared pool, per-mailbox daily caps are enforced by the system rather than by policy, and new mailboxes ramp up gradually. If you want to blast ten thousand people, this is the wrong product.
Do I need someone technical?
No. Connecting a mailbox is an OAuth click, and editing an agent is editing text. The technical options, your own model keys and access from your own assistant, are there for teams that want them and ignorable for teams that do not.
Can I use my own API keys?
Yes. Run the agents on your own model API keys if you would rather pay the labs directly. Your spend then shows up on your own bill, at your own negotiated rate.
Can I run it from Claude or ChatGPT?
Yes, over MCP. You can list agents, force a run, review what came back, approve a send and export from inside your own assistant, without opening the workbench.
Do you work with agencies?
Yes. Agencies run Big Ears for their clients, and several already do. If you already have an agency you are happy with, they can run this for you rather than instead of you.
Start with a starter. Rent the rest when it earns it.
Run a free starter agent today on your own list. If it never earns its rent, you have still kept everything it wrote.
No credit card. Starters are free and yours.