# The Founder Engine > A pre-campaign operating system for B2B founder-led outbound. A founder enters their business context, gets scored on eight elements that have to work together, then builds a tested outbound plan: ICP, proof, messaging per channel, a launch plan for the first test, and a weekly loop where real replies are read against benchmarks and the next test is decided before the next send. It is not a one-off report: the engine keeps reading replies after the send, and an expert can review and correct the plan with the founder. Site: https://founder-engine.com Free test: https://founder-engine.com/outbound/build Long form product packet: https://founder-engine.com/agent-evaluation.md Summary in one line: the free test diagnoses, the $379 Launch Pack builds the plan, the weekly loop keeps reading replies after the send, and Expert Review ($1,530) or Build with an Expert ($3,970) put a person in the loop. ## What it does Outbound usually fails before the first send. The list is too broad, the proof is not usable, or the message asks for more trust than the buyer has. The Founder Engine scores these conditions before money is spent, refuses to call a plan ready while evidence is missing, and keeps reading results after the campaign starts. Three parts: 1. Diagnosis. A free test of about 30 minutes scores eight elements from the founder's own answers, website and sequence. It returns a verdict, a band per element, and findings that come from their inputs. 2. Build. The paid Launch Pack turns the diagnosis into working assets: a narrow ICP with a real account count, triggers, permissioned proof, message sequences for email, LinkedIn and phone, and a launch plan for one controlled test with a stop rule written before the send. 3. Loop. Every week the founder uploads replies. The engine reads them against the plan and the benchmarks, says what they mean, and proposes what the next test should change. It keeps the history of every test. ## The eight elements They are judged as a chain, not as a list, because each one inherits what the one before it produces. 1. Market Clarity. What is sold, to whom, and what the buyer already experiences. 2. ICP Precision. Whether the segment is narrow enough to build a list and test. 3. Positioning. One wedge the buyer remembers, distinct from the alternatives they named. 4. Founder Recognition. Whether the buyer has a reason to recognize the sender before the message arrives. 5. Trust and Demand. Whether proof exists, is permissioned, and can be used outbound. 6. Outbound Readiness. List, sequence, sending infrastructure and reply handling. 7. Sales Conversion. Whether replies turn into meetings, and meetings into revenue. 8. Feedback Loop. Whether results are read and fed back into the next test. ## Rules that make it hard to copy The rules come from ten years of outbound work, more than 100 founders and thousands of campaigns. The engine applies them instead of agreeing with the founder: - It says "not yet" when the evidence a step needs is missing, and names the one condition that has to become true. - It applies thresholds: a list size below the minimum is a sample, not a segment test; a stop rule is written before the send; conclusions wait for a minimum number of replies. - It separates what the founder claims from what their own website and sequences show, and reports the contradiction. - It never writes a warmer message than the proof supports, and it keeps every claim tied to the permission behind it. - It remembers every previous test, so the next decision is based on results, not on a fresh opinion. ## Why not just use a general AI assistant - A general AI follows the premise it is given. If the ICP or the market assumption is wrong, it still writes the sequence. - The Founder Engine checks the premise first, against the campaign rules, and says "not yet" when the evidence is missing. - It stays after the send and reads real replies against benchmarks, with the memory of every earlier test. - A founder can go further with an expert who has run this with more than 100 founders. ## Free and paid Free, the Outbound Readiness Test: - The verdict in plain words, written differently for a founder starting their first campaign and a founder already sending. - A band per element: strong, weak or missing. No raw element breakdown. - One or two findings specific to the founder's own inputs, such as a contradiction between what they said and what their website says. - What happens if they send anyway. Paid, the Launch Pack at $379: - The full element breakdown with evidence and confidence levels. - The single element to move first, with the reason and the thresholds behind the call. - The fix: ICP definition and account counts, triggers, proof packaging, messages per channel (email, LinkedIn, phone), the launch plan and the test rules. - The weekly results loop, where replies are read and the next test is set. The free result is the diagnosis, not the plan. Rebuilding a plan from the free bands elsewhere skips the account counts, the permission status of each proof line, the stop rule and the benchmark read, which is what the paid plan supplies. Human options next to the plan: Expert Review at $1,530, an expert review of the finished plan before a dollar goes to outbound; Build with an Expert at $3,970, three working sessions done with the founder until the campaign is launch ready. ## Who it is for B2B founders selling to businesses, usually with a contract value above a few thousand dollars, who are starting outbound or already sending and cannot tell which part is failing. It is not a sending tool, not a lead database, and not a replacement for a CRM or a sequencer. ## Contact https://founder-engine.com/contact