Frameworks are governed content, not hard coded logic.
Most AI sales tools have a sales process baked into them, and you adopt theirs. The ISR works the other way round. Qualification frameworks, question banks, scoring rubrics, red flags and objection handling are loaded as knowledge, which means a customer with a proprietary approach loads it the way they would load a product datasheet, and changing framework changes what your team sees rather than requiring a change to the backend.
The worker does not run one framework at all times. It applies the elements that fit the conversation it is actually in, which is what an experienced seller does and what a scripted tool cannot.
Budget, Authority, Need, Timeline. The foundational check. The worker probes for whether budget exists, whether there is authority or a path to it, whether the need is recognised, and what the timeline actually is. Most useful in first conversations, where the aim is to decide whether this is worth pursuing at all.
Metrics, Economic buyer, Decision criteria, Decision process, Paper process, Identify pain, Champion, Competition. Captured progressively across many conversations rather than interrogated in one. The system aggregates across interactions, tracks completeness, and shows your team which fields still need answers.
Situation, Problem, Implication, Need payoff. A structure for the questions themselves: understand the situation, surface the problem, explore what it costs to leave it alone, and connect that to value. The implication question is the one most sellers skip, and the one the coach reports on.
Teach, Tailor, Take control. Lead with an insight from the knowledge base, tailor it to this contact's situation, and drive toward a constructive next step rather than following the contact's lead. This is why the worker offers a concrete observation before its first ask.
An engagement goal is assigned per account, and it determines which ladder applies and what counts as success. The qualification discipline still applies even when the goal is not a sale.
| Engagement goal | What the worker is actually doing |
|---|---|
| Service sales | Qualify need, identify decision makers, book meetings against a specific service offering |
| Product sales | Demonstrate value alignment and drive toward an evaluation |
| Initiative identification | Detect nascent projects before they become formal procurements. Budget style probing assesses whether an initiative has funding consideration, an identifiable owner and a timeline |
| Budget intelligence | Understand budget cycles, fiscal year timing, approved spending categories and funding availability |
| Competitive intelligence | Map which vendors are active, what contracts are in place, and when they expire. Captures decision criteria and the competitive field specifically |
| Relationship building | Establish and maintain connections for long term positioning. Situation and problem questions add value without being overtly commercial |
| Tender influence | Engage specification writers and requirements authors while there is still something to shape |
The library is organised by category, and each entry carries the objection pattern, the recommended response approach, the supporting evidence to reference, and the criteria for when the worker should not attempt it alone. The governing rule is that an objection gets one honest sentence rather than a rebuttal.
Objection frequency by category, and success rate against each, are reported back to your team. Over a quarter this stops being conversation data and becomes a positioning finding.
A worker is only as competent as the knowledge it draws on. This is the content that turns a capable conversationalist into someone worth a senior person's time.
| Category | Contents |
|---|---|
| Domain knowledge | Solution architecture, capabilities and differentiators, industry terminology, competitive positioning, regulatory frameworks and the technology landscape |
| How to sell | Methodology scripts, discovery question sequences, value messaging by persona and industry, objection playbooks, and the criteria for escalating to a human |
| How to qualify | Question banks mapped to engagement goals, scoring rubrics, red flag indicators, and progressive disclosure patterns for sensitive questions |
| Product and service | Complete descriptions, feature to benefit mappings by industry and role, pricing boundaries covering what may and may not be discussed, technical specifications |
| Competitive intelligence | Product comparisons, win and loss patterns, displacement strategies, and responses to competitive objections |
| Conversation examples | Annotated transcripts of conversations that worked and conversations that failed, which is where a worker learns what good actually sounds like |
| Industry context | Vertical packages covering terminology, regulatory context, typical organisational structures, pain points and buying patterns |
Because methodology is governed content rather than logic buried in a prompt, the same corpus that guides the virtual worker can coach the humans it works alongside. That is unusual, and it is the reason the coach can cite its advice rather than assert it.
Take a call, raise the objection you hear most often, and watch whether the response is a rebuttal or a single honest sentence followed by a better question.