M
Our Approach

How we turn
questions into
decisions.

Research methodology at North Vector Systems is not a black box. We believe clients should understand how we reach our conclusions — because that understanding is part of what gives them the confidence to act on them.

P.1
Triangulation over single-source reliance
No finding goes into a report supported by only one interview or one data source.
P.2
Disconfirmation as standard practice
We actively look for evidence that contradicts our emerging hypothesis before we commit to it.
P.3
Confidence is explicit, not implied
Every significant finding carries an explicit confidence rating. We don't mask uncertainty in confident language.
North Vector Systems Research Architecture
01
Phase One
Decision Mapping
Identify the exact decision, the information that would change it, and the constraints on quality and speed.
1–3 days
02
Phase Two
Intelligence Architecture
Design the research methodology — methods, sample, instruments, and analytical framework — before any fieldwork begins.
3–5 days
03
Phase Three
Primary Fieldwork
Structured interviews, surveys, and observation with real market participants. Source documentation throughout.
1–3 weeks
04
Phase Four
Analysis & Synthesis
Triangulation, disconfirmation testing, confidence scoring. The narrative is built last — after the evidence is complete.
5–8 days
05
Phase Five
Decision Delivery
A working session, not a readout. Evidence is presented, challenged, and connected to the original decision.
1–2 days
Primary Research Methods

Six tools we use.
Chosen for the question, not habit.

In-Depth Interviews

60–90 minute semi-structured conversations with decision-makers, buyers, operators, and experts. Designed to surface the reasoning behind behaviour, not just the behaviour itself.

15–60 interviews per engagement
Topic guide + probing protocol
Full transcript + synthesis

Quantitative Surveys

Statistically valid sample surveys across defined target populations. Includes conjoint analysis, Van Westendorp pricing, Gabor-Granger, MaxDiff, and standard Likert-scale instruments.

n=200 to 2,000+
Online and CATI modes
Full crosstab data delivered

Expert Panel Research

Structured conversations with former senior executives, ex-regulators, supply chain specialists, and technical domain experts. Best for questions where buyer behaviour alone doesn't tell the whole story.

40,000+ specialist network
Activated within 48 hours
Screened for relevance

Ethnographic Observation

In-context research — in-home visits, store walkthroughs, workplace shadowing. Used when stated behaviour consistently diverges from actual behaviour. Available in 12 markets.

In-home and in-context
Video and observational coding
Consumer packaged goods, healthcare

Focus Groups

Moderated group discussions of 6–10 participants. Used selectively — for concept testing, messaging development, and category exploration where social dynamics add insight rather than bias.

In-person and online
Professional moderation
Observation and recording

Desk Research & Data Synthesis

Systematic secondary research across regulatory filings, financial disclosures, academic literature, trade publications, and news databases. Used to establish baseline — never as the sole source of a finding.

Proprietary database access
Source documentation standard
Never used as primary evidence
Analysis Protocol

How raw data becomes
a decision.

The gap between data collection and useful insight is where most research firms disappoint. North Vector Systems's analysis protocol is a structured process — not an analyst's intuition working alone in a document.

Every engagement follows the same five-step analysis sequence. It takes longer than the alternative. The findings are significantly more defensible as a result.

1
Evidence Inventory
All primary data — interview transcripts, survey results, observation notes — is tagged, coded, and organized by research question. No analysis begins until this is complete.
Output: Evidence database
2
Pattern Identification
Analysts work independently to identify themes and patterns across the evidence base, then cross-check against each other. Disagreements are flagged as questions — not resolved by seniority.
Output: Preliminary findings map
3
Disconfirmation Testing
For each emerging finding, we actively search for evidence that contradicts it. If contradictory evidence exists and is strong, the finding is modified or removed. If it's weak, it's documented as a caveat.
Output: Stress-tested findings
4
Confidence Scoring
Each finding is assigned a confidence score (1–5) based on source count, source independence, and corroboration across methods. Low-confidence findings are clearly labelled — never buried in confident language.
Output: Scored findings matrix
5
Decision Synthesis
Findings are organised around the original decision — not around research categories. The final narrative connects intelligence directly to the recommendation the leadership team needs to make.
Output: Decision-ready report
Quality Standards

The numbers
we hold ourselves to.

3+
Minimum sources per finding
No finding enters a report supported by a single source. Three minimum. Major findings require five or more.
72hr
Internal peer review window
Every draft report is reviewed by a second analyst who was not involved in fieldwork before it goes to the client.
100%
Source documentation rate
Every claim in every report is traceable to a primary source in the evidence appendix. No exceptions.
0
Reports delivered without a delivery session
We never drop a report in an email. Every engagement closes with a live working session where findings are tested.
Research Ethics

Standards we
don't negotiate.

Informed consent — always
Every research participant is told they are participating in research, by whom, and for what general purpose. No covert observation. No deceptive framing.
Conflict of interest declaration
We check for conflicts before every engagement. If we've worked for a direct competitor in the past 18 months on a related question, we disclose it — and sometimes decline the work.
Respondent data protection
All respondent data is anonymised in reports. Personal information is held for the minimum necessary period and never shared with clients in identifiable form without explicit consent.
No fabrication of evidence
If the evidence doesn't support a finding the client wanted, we say so directly. We have walked away from engagements rather than adjust a finding to match a preferred narrative.
ESOMAR and MRS compliance
All primary research is conducted in accordance with ESOMAR International Code of Research Ethics. Our UK operations are MRS member registered.
Methodology you can
defend in the boardroom.
Start a Brief →