Our methodology
How we analyze relationships
Make Sense of Us uses a structured, evidence-based framework to examine relationship communications. Every finding is tied to something specific in what you share — not assumptions, not generalizations, not what we expect to find.
Analysis grounded in evidence, not opinion
Most relationship tools ask you to rate your feelings or answer questionnaires. We do something different: we examine what's actually present in your communications — the patterns, the gaps, the contradictions, and the changes over time.
Our framework draws on established research in communication studies, attachment theory, and relationship psychology. But we don't apply labels or diagnoses. We describe what the evidence shows, note where it's incomplete, and present counterexamples alongside supporting findings.
The result is a report that reflects your specific situation — not a generic template filled in with your names.
The six dimensions we analyze
Every analysis — regardless of relationship type or service — examines the same six dimensions. Together they give a complete picture of how a relationship is functioning.
Communication
Clarity, consistency, and tone across the relationship
We examine how clearly each person expresses themselves, how consistently they communicate, and what the overall tone of exchanges reveals. We look at directness versus ambiguity, how conflict is handled in writing, whether communication patterns are stable or erratic, and what's left unsaid as much as what's said.
What we look for
- Directness and clarity of expression
- Consistency of tone across contexts
- How conflict and tension appear in writing
- Patterns of avoidance or deflection
- What topics are absent from the record
Effort
Investment, reciprocity, and follow-through from each person
Effort is one of the most revealing dimensions in any relationship. We examine who initiates contact, who follows through on commitments, how each person responds to the other's needs, and whether the investment is roughly balanced or significantly one-sided. We note both the presence and absence of effort signals.
What we look for
- Who initiates contact and how often
- Follow-through on stated intentions
- Responsiveness to the other person's needs
- Balance of investment between both people
- Changes in effort levels over time
Trust
Honesty, reliability, and consistency between words and actions
Trust is examined through the consistency between what people say and what the record shows they do. We look for mixed signals, contradictions between stated intentions and behavior, patterns of reliability or unreliability, and how each person handles situations where honesty is difficult. We don't conclude that someone is lying — we describe what the evidence does and doesn't support.
What we look for
- Consistency between stated intentions and behavior
- Mixed signals and contradictions in the record
- Patterns of reliability or unreliability
- How difficult truths are handled
- What the evidence supports versus what it leaves open
Boundaries
How limits are expressed, respected, and tested
We examine whether boundaries are stated clearly, how the other person responds when they are, and whether limits are consistently respected or regularly tested. This includes both explicit boundaries (stated directly) and implicit ones (patterns of discomfort or withdrawal that suggest a limit is being approached). We note when boundaries appear absent as well as when they're present.
What we look for
- Whether limits are stated clearly or implicitly
- How the other person responds to expressed limits
- Patterns of boundary testing or pressure
- Consistency of boundary enforcement
- Areas where limits appear absent or unclear
Connection
Warmth, genuine interest, and emotional presence
Connection is the dimension most people feel intuitively but find hardest to articulate. We look for signals of genuine interest in the other person, warmth and affection in communication, whether interactions feel mutual or transactional, and how each person responds to the other's emotional disclosures. We distinguish between performed warmth and patterns that suggest genuine engagement.
What we look for
- Signals of genuine interest in the other person
- Warmth and affection in communication
- Whether interactions feel mutual or transactional
- Responses to emotional disclosures
- Presence or absence of shared meaning and humor
Change
How patterns are shifting — and in which direction
The Change dimension is what separates a snapshot from a picture of a relationship's trajectory. We examine whether patterns across the other five dimensions are stable, improving, or deteriorating — and whether changes appear to be responses to specific events or part of a longer drift. For Monitor services, this dimension is tracked across multiple check-in periods.
What we look for
- Whether patterns are stable, improving, or deteriorating
- Whether changes follow specific events or reflect gradual drift
- Which dimensions are changing and which are stable
- Whether both people are changing in the same direction
- What the trajectory suggests about where the relationship is heading
How our AI processes your data
Our analysis pipeline is built around a structured AI system trained to apply the six-dimension framework consistently across different relationship types and communication styles.
Ingestion and normalization
Your uploaded communications — whether text exports, screenshots, or written summaries — are processed into a normalized format. Timestamps, speaker attribution, and message context are preserved. Screenshots are processed through OCR to extract text content.
Dimension-by-dimension analysis
The normalized record is analyzed against each of the six dimensions in sequence. For each dimension, the system identifies supporting evidence, counterexamples, and gaps — noting where the record is incomplete or where a finding is uncertain rather than forcing a conclusion.
Confidence scoring
Each finding is assigned a confidence level based on the strength and consistency of the evidence. High-confidence findings are supported by multiple clear signals. Low-confidence findings are flagged as tentative and explained. We don't present uncertain findings as established facts.
Relationship-type calibration
The framework is calibrated for each relationship type. What constitutes a healthy effort balance in a co-parenting arrangement differs from what it looks like in a romantic partnership. The system applies type-specific context to avoid applying the wrong standard to the right evidence.
Report generation
Findings are structured into a readable report with clear sections, specific evidence citations, and practical observations. The report is written to be understood without a therapist present — direct, honest, and free of jargon.
How this differs from other relationship tools
Most relationship tools fall into one of three categories. We don't fit any of them.
Their approach
Ask you to rate your feelings or answer standardized questions. Results reflect your self-perception, not what's actually happening in the relationship.
Our approach
We analyze the communications themselves — not your interpretation of them. The evidence is in the record, not in how you feel about it.
Their approach
Provide exercises, prompts, and guided conversations. Useful for people who want to work on a relationship together — but require both people to participate.
Our approach
We work from one person's perspective and their available record. You don't need the other person's participation or agreement to get a clear picture.
Their approach
Respond to what you tell them with advice, validation, or suggestions. Output quality depends entirely on how you describe the situation — and tends toward confirmation of what you already believe.
Our approach
We analyze the primary source material — the actual communications — rather than your description of them. Findings are tied to specific evidence, not to what you've told us to expect.
What our methodology cannot do
Honest methodology includes honest limits. There are things our analysis is not designed to determine, and we don't pretend otherwise.
We cannot determine intent. We describe what patterns appear in the record — not why someone behaved that way.
We cannot verify facts. If communications contain false statements, we analyze what's present — we can't fact-check claims against external reality.
We cannot predict outcomes. We describe trajectory based on current patterns — not what will happen next.
We cannot replace therapy. Pattern analysis is not clinical assessment, and our reports are not a substitute for professional mental health support.
We cannot provide legal conclusions. Our analysis is not suitable as evidence in legal proceedings and should not be used as such.
We cannot analyze what isn't in the record. Gaps in the uploaded material are gaps in the analysis — we note them, but we can't fill them.
See the methodology in practice
Choose a service, upload your communications, and receive a report built on this framework — delivered privately to your portal.