TL;DR
There’s no one-size-fits-all research model. Use DIY for speed and iteration, Managed for rigor and high-stakes decisions, and Hybrid when you need both. This guide shows how teams actually blend the three—and how Mirats structures projects to deliver speed and certainty.
Why this choice matters
Budgets are tighter, timelines are shorter, and leadership expects defensible decisions. Teams that match the operating model to the business question ship insights faster without sacrificing quality. At Mirats Insights, we work across all three—Managed Research, DIY via Mirats Quanto, and Hybrid programs. Below is our practical, step-by-step playbook.
The three models at a glance
DIY (Self-Serve)
- Best for: rapid concept checks, weekly creative/UX iterations, small- to mid-stakes decisions, internal pilots.
- Strengths: speed, cost control, transparency, in-house learning.
- Risks: survey design gaps, sampling errors, weaker QC if you’re newer to research.
- Mirats fit: Mirats Quanto DIY with validated templates (CSAT/NPS, ad tests, concept screens) and programmatic targeting.
Managed (Full-Service)
- Best for: brand strategy, pricing & segmentation, trackers, claim support, board-level or regulated work.
- Strengths: methodology depth, compliance, advanced analytics, bulletproof deliverables.
- Risks: longer cycles; higher cost if scope isn’t tuned.
- Mirats fit: End-to-end via Mirats Insights—decision framework → instrument design → field & QC → analytics (SEM/CFA/Conjoint) → exec-ready outputs.
Hybrid (Co-Pilot)
- Best for: teams with smart analysts but bandwidth gaps; programs that mix quick sprints with one big readout.
- Strengths: your control + our expertise; faster cycles with lower risk; flexible costs.
- Risks: unclear RACI slows things—governance is essential.
- Mirats fit: You own what you want (e.g., scripting, first-pass cuts); we provide panel, QC and advanced analytics; shared dashboards keep everyone aligned.
A simple decision flow
What’s at stake?
- High → choose Managed or Hybrid.
- Medium/Low → start DIY; escalate to Hybrid if confidence gaps appear.
How mature is the brief?
- Crisp hypothesis + metrics → DIY/Hybrid.
- Ambiguous scope → start Managed to frame the study.
Method complexity?
- CSAT/NPS, concept screens, A/B creatives → DIY/Hybrid.
- Pricing, segmentation, longitudinal trackers → Managed/Hybrid.
Compliance & QC needs?
- Regulated/brand-critical → Managed/Hybrid with stricter governance.
Real-world mappings
- Creative/Ad iterations (weekly): DIY for headlines, thumbnails, hooks; quarterly Managed deep-dive for message architecture & drivers.
- Product-market fit probe: Start Hybrid—Mirats designs the instrument & panel; your team runs DIY iterations; Mirats validates the final readout.
- Pricing study: Managed (Conjoint/Gabor-Granger). After the big read, a Hybrid handover lets you run a lightweight internal pricing tracker.
- Brand health tracking: Managed setup (method + baseline + weights) → Hybrid upkeep (you run monthly cuts; Mirats audits quarterly).
What “good” looks like in each model
DIY Success Checklist
- Validated templates (CSAT/NPS/U&A/ad tests)
- Targeting & quotas locked before launch (avoid “anyone” samples)
- Pilot with 10–20 completes
- QA: speeders, straight-liners, red-herring checks, device fingerprinting, geo/IP controls, reCAPTCHA
- Pre-defined decision thresholds (e.g., “If PI < 60%, iterate creative”)
- Assumptions documented (incidence, LOI) in the deck
Managed Success Checklist
- Decision Brief: business question → KPIs → must-have analyses
- Governance: GDPR/CCPA/DPDPA, ISO-aligned processes, AES-256 at rest, TLS in transit
- Up-front analytics plan: driver models, segment stability, claim wording tests
- Executive outputs: one-page narrative, “so-what” cascade, prioritized actions, analyst data pack
Hybrid Success Checklist
- RACI for sampling, scripting, field QC, analytics, delivery
- Shared workspace & dashboards (roles/permissions)
- “Escalation to Managed” rule when stakes/complexity rise
- Cadence: weekly ops stand-up; monthly synthesis; quarterly method audit
How Mirats implements this
Platforms & Panels
Programmatic reach across 60+ countries with stringent quality controls: fraud detection, device fingerprinting, geo/IP whitelisting, attention gates, and deduplication.
Methods
CSAT/NPS, U&A, brand tracking, concept/product tests, ad/communication evaluation, pricing (Van Westendorp, Gabor-Granger, Conjoint), segmentation.
Audits & Mystery Shopping (Owlit)
Compliance and CX measurement for retail/service networks with photo/audio evidence and senior-auditor scoring.
Delivery
Executive story decks, analyst workbooks, and interactive dashboards—with optional readouts for leadership and a secure data room for analysts.
Security & Compliance
ISO 9001 & ISO/IEC 27001 aligned processes, privacy-by-design, encryption at rest and in transit, role-based access and audit trails.
Roles & RACI (typical Hybrid setup)
- Client Insights/PM: R (brief, priorities, stakeholder mgmt.), A (final decisions)
- Client Analyst: R (first-pass cuts), C (instrument edits)
- Mirats Method Lead: A/R (design, analytics plan, QC gates)
- Mirats Field Ops: R (sampling, quotas, fraud controls, LOI)
- Mirats Data Science: R (drivers, segmentation, pricing)
- Mirats Engagement Lead: A (timeline, risks, change control)
SLAs & Cadences
- DIY: same-day scripting; pilot in 24–48h; toplines in 24h post-field
- Managed: design 3–7 days; field per incidence/LOI; analytics 5–10 days; exec deck on T-1
- Hybrid: weekly ops call (30 min), monthly synthesis (60–90 min), quarterly method audit
QC & Data Integrity (our non-negotiables)
- Before field: logic checks, device/OS guardrails, time-to-complete bounds
- During field: deduping (fingerprint + IP + cookies), geo fencing, attention gates
- After field: speeder/straight-liner trims, pattern checks, toxicity/AI-spam filters, quota reconciliation
- Analytics QA: sanity checks, variance diagnostics, weighting audits
Scope knobs (how we tune cost & time)
- Sample size & splits (by segment/market)
- Questionnaire length (LOI targets)
- Depth of analytics (from simple cuts → drivers → conjoint/segmentation)
- QC strictness (baseline vs enhanced)
- Reporting depth (topline vs full story + workshop)
Pitfalls to avoid
- Treating DIY like Managed: small samples + weak QC + big decisions = risk
- Over-engineering low-stakes tests: burns time and budget
- Skipping respondent quality controls: a few farms can wreck your data
- No success criteria: define green/amber/red thresholds before field
Quick selector (bookmark this)
- Need it this week + medium stakes? DIY template on Mirats Quanto → pilot → go/no-go
- Board-level decision or claim support? Managed with Mirats Insights
- Multiple sprints + one big readout? Hybrid: DIY for sprints; Mirats for method/QC/final story
- Not sure? Start Hybrid—keep speed, add guardrails, de-risk the decision
Onboarding: what we need from you
- Decision brief (business question, KPIs, must-haves)
- Audience definition (targeting, quotas, exclusions)
- Success thresholds (what “good” means)
- Constraints (timeline, markets, compliance flags)
- Preferred deliverables (C-suite deck, data room, workshop)
FAQs
Can we start in DIY and “upgrade” mid-flight?
Yes. Hybrid lets us assume design/QC/analytics ownership when stakes rise.
How do you protect against low-quality responses?
Multi-layer controls: fingerprinting, geo/IP checks, attention tests, deduping, speeder/straight-liner trims, AI-spam filters, manual audits.
We need exec-ready outputs—do we still get raw data?
Yes. You get the story deck and analyst workbooks/exports.
Can you run global studies?
Yes—programmatic reach across 60+ countries with consistent QC and governance.
Ready to choose?
If you know the question but not the path, start Hybrid—we’ll shape the design, you keep the speed, and leadership gets the certainty. If you already have a crisp hypothesis and need quick reads, go DIY on Mirats Quanto. If the decision is high-stakes, run Managed for end-to-end rigor and defensibility.
