Manifesto
Five principles for measuring human reps and AI agents the same way.
Sales is a human craft. What every sales leader wants is reps who can see their own gaps and fix them without being told. Most end up doing it by hand: hours of call reviews, each ending in a judgment call the next reviewer might make differently.
We built Big Sister to take the hours out of that. It scores every customer conversation against one written standard, for reps and AI agents alike, with the evidence attached. The rep sees it first. So that when you do sit down with a rep, you are both looking at the same evidence, and the conversation starts at what to fix.
A growing share of those conversations will be with software. The standard you hold a rep to is the one you will need for an agent. Without it, both go in front of buyers untested.
We start at the first real conversation.
However you generate demand is your business. Human SDRs, AI SDRs, inbound, partners, referrals: run it with whatever works. Our work starts at the first real conversation with a real buyer and ends when the deal is won or lost. Discovery, demo, negotiation, follow-up, and the relationship that holds it together. That is where revenue is decided, and it is the only part we work on.
We build AI around the humans, not instead of them.
We are not replacing salespeople with agents, and we do not believe anyone should hand the whole sales process to software. The leaders we work with want help picking the right AI for their team.
We think the best setup is a human and an AI working one deal together, each doing what the other cannot. Our agents exist to make a salesperson more productive and more autonomous. They take CRM entry and call prep off the salesperson’s plate, so the human gets their time back for the conversation and the relationship.
You decide what to hand over: the follow-up at 11pm, the inbound call on Saturday, the first-pass qualification. You keep the relationship. We keep the quality.
Where the method comes from
Since 2009 we have worked inside sales teams as consultants, building sales processes, hiring and training closers, and watching what actually moves a deal. What we learned from that work is written down as the Big Sister Method.
Playbooks go stale because nothing forces them to update. The method is an asset your company owns: the skills that define good selling, the playbook tuned to your product and your buyers, and what your best conversations have in common. New hires and new agents start from it on day one. And it learns from your results, not ours. Your CRM outcomes feed back into the method, stage by stage, and it tightens around what actually moved your conversion rates.
Your CRM outcomes feed back into the method, stage by stage.
Reps and agents are held to the same standard.
Each conversation is graded against one written standard, the Sales Score Standard: a set of sales skills, each graded against a rubric with reference examples, rolled into one Sales Score per rep and per team. The Sales Score exists to make people better. It is built for coaching and training. If you use it in performance reviews, the rep has already seen every piece of evidence and can dispute it. The overall score is a summary. Coach from the skill scores underneath it. A human rep and an AI agent are graded on the same rubric, so you can see which parts of the work an agent handles well enough to take off a rep’s plate. We built the method to score both from day one.
That includes the agents we build ourselves. Same rubric, same evidence, same appeal. Before an agent talks to your customers, it passes the standard your reps are held to. If we ever held our own agents to a softer test than anyone else’s, the standard would be worthless, and so would we.
Before an agent talks to your customers, it passes the standard your reps are held to.
You get a score, and the moment in the call that earned it.
A score you cannot check is an opinion with a decimal point. Every score cites the moment in the conversation that earned it. Any rep can dispute any score, and the dispute goes to a trained human referee, not to their manager and not back to the model. A review that just rubber-stamps the model’s score does not count.
That is also what makes improvement possible. A closer who scored low on a skill on Tuesday can replay that exact moment on Wednesday, against the same buyer, as many times as it takes. We call it Second Chance.
And we hold ourselves to that test too. When trained experts score the same conversations, they disagree with each other far more than you would expect. That is not a knock on them. It is what happens when there is no shared standard underneath the judgment. We measure our scorer against those experts and publish the agreement, including the skills where a human still wins, and we say so first.
Your CRM has the outcome and your recorder has the call. We score the call against a written standard and handle disputes about the score.
Send us your recorded customer conversations.
From Fireflies, Zoom, Meet or your CRM. Within days, each rep sees where they are strong and what to work on, with the moment in the call behind every score. You see what they see.
Scoring a team? Tell us about it and we onboard you personally.
What happens next
- 01Send us recordings. Whatever you record with now, no new tooling to install.
- 02We score them against the standard, skill by skill, and cite the moment that earned each score.
- 03You get a Sales Score per rep. Each rep can take a score they disagree with to a trained human referee.