Methods and Models for Interview Design and Field Research

Good interviews are engineered, not improvised

Lesson 1 established why interviews exist in the RelayOps learning plan. This lesson supplies methods and models: interview guides, question types, sampling logic, note-taking systems, and shadow protocols you can reuse every week. RelayOps is building dispatch software for 80-200 technician HVAC/plumbing firms; every method here targets the same-day rebalance job.

The Mom Test principles (adapted for B2B)

Rob Fitzpatrick's Mom Test rules adapt well to B2B discovery:

1. Talk about their life, not your idea.
2. Ask about past specifics, not future hypotheticals.
3. Listen more than you talk; avoid pitching.

B2B adaptation: add economic buyer questions and workflow artifacts requests. RelayOps banned questions like "Would you pay $X for AI dispatch?" Replace with "What did you spend last month on overtime and scheduling tools?"

Interview guide structure

RelayOps uses a 45-minute dispatcher guide:

COO guide swaps job story for budget narrative: last funding decision for operations software, thresholds, kill metrics for vendors.

Guides are printed with required questions (must ask) and optional probes (if time).

Question types and anti-patterns

Leading questions contaminate evidence. Avoid "Don't you hate spreadsheets?" Ask "How do you track skills today?"

Sampling and saturation

RelayOps targets thematic saturation: new interviews repeat prior patterns without new obstacles. Rule of thumb: 12-20 interviews per segment hypothesis, then shadow subset. Saturation signals:

• Last 5 interviews add no new obstacle categories
• Pain ranking stabilizes
• Verbatims rhyme across firms

Oversampling feels productive but delays synthesis. Undersampling produces false certainty.

Note-taking and coding model

Within 24 hours, interviewers complete a note template:

• Segment tags (firmographics)
• Job story summary
• Obstacle tags (skill visibility, tabs, phone, parts)
• Evidence strength (1-5)
• Verbatim quotes (2)
• Assumption impact (supports / challenges which)
• Follow-ups (shadow, IT call, data request)

Coding means tagging obstacles so Friday synthesis can count frequencies. RelayOps uses a shared tag list; new tags require team approval to avoid synonym sprawl.

Shadow and contextual inquiry methods

Contextual inquiry combines observation with think-aloud questions during live work. RelayOps shadow rules:

• One researcher observes, one optionally asks quiet clarifiers
• Time three critical events minimum
• Screenshot with permission; no customer phone numbers in images
• Debrief dispatcher 5 minutes after peak passes

Remote versus on-site field research

Post-pandemic, much discovery is remote screen share. Trade-offs:

RelayOps does first shadow remote; on-site for pilot customers within driving distance to catch environmental stress.

Part A: Required opening

"Thanks for making time. We're researching how dispatch works on busy days. We're not selling today. May I take notes? No recording unless you prefer."

Part B: Core story prompt

"Pick the most chaotic day in the last two weeks. Walk me minute by minute from when you knew the board was in trouble."

Part C: Coding output (BluePeak)

Part D: Synthesis check

After 10 coded interviews, tab_switching appears in 9/10 → prioritize CRM sync wedge. Check: tag frequency drives roadmap ✓

Probing spend without anchoring price

Spend probes ask about past behavior: invoices, overtime totals, module fees, headcount. They avoid hypothetical pricing: "Would you pay $2,800?" Hypotheticals produce polite yes and false WTP signals. RelayOps spend probe script: "What did you pay last year for scheduling-related tools and overtime premiums on peak weeks?"

Spend probes must separate categories. Buyers lump software, services, and labor. RelayOps asks line by line. A COO who cites $40k overtime and $12k module fees gives budget hook vocabulary for ROI conversations later.

Anchoring risk appears when founders mention their intended price early. Interview guide places spend section before any RelayOps pricing exists. B2B buyers anchor on first number heard; discovery protects neutrality.

Saturation and when to stop interviewing

Saturation means new interviews repeat prior obstacle categories without adding falsifiers or mechanisms. RelayOps rule: when last five qualified Segment A interviews add no new obstacle tags and pain rank stabilizes, shift hours to shadows and pilots.

Stopping early creates false certainty; stopping late delays synthesis and burns founder calendar. Saturation is not a fixed n=20 formula. Electrical explore may need only eight interviews to see split signal; Segment A rebalance needed nineteen before confidence.

Document saturation decision in synthesis memo: "Stopped Segment A discovery interviews at 22 because last five lacked new tags; reallocate 6 hours to IT workshops."

Part A: Six COO spend responses

Part B: Budget hook summary

5/6 cite overtime or modules → WTP signals strong for H1.

Part C: Vague responses

COO 3 and 6 scheduled for follow-up with artifact request.

Part D: Managerial read

Spend coding upgrades opportunity evidence toward level 1-2 mix. Check: 5/6 = 83% hook rate ✓

Dispatcher says "I'd love your app." No spend data yet.

1. Classify evidence level of "I'd love your app."
2. Write three follow-up spend probes (past behavior).
3. Does this quote affect A1 COO budget assumption?
4. Saturation: can you stop after this quote if n=18?

Solution

1. Level 1 opinion/intent without behavior.

2. Probes: "What tools did your company invoice for dispatch last year?" "How much overtime last heat week vs plan?" "What headcount did you add to cope with scheduling breaks?"

3. No impact on A1 until COO-level spend confirmed; dispatcher enthusiasm supports A2 only weakly.

4. Cannot stop on this quote; saturation requires tag stability, not single intent statement.

Check: probes convert intent toward WTP evidence ✓

Tag governance and synonym control

RelayOps tag governance requires approval for new obstacle tags. Synonym sprawl ("tabs", "tab_switch", "context_switch") breaks frequency math. Weekly tag review merges synonyms and updates guide.

Tag list versioned in Notion with change log. Interviewers on old guides corrected within one week.

Debrief protocol after shadows

Five-minute dispatcher debrief after shadow captures emotions and political context invisible in timing sheet. Debrief quotes feed emotional job insights without recording peak chaos.

Guide version control

RelayOps treats guide version control as operational discipline for mid-market HVAC and plumbing dispatch discovery, not a one-time workshop topic. Founders document decisions in the opportunity decision memo and segment strategy memo so Maya Chen and Jordan Okonkwo align daily calendar choices with beachhead rules.

In practice, guide version control connects to measurable leading indicators: qualified interviews, shadow medians, assumption register statuses, and pilot telemetry. When indicators diverge from thesis language, the team runs a forcing function review within five business days rather than waiting for quarter-end board meetings.

Corporate innovation teams can mirror the same discipline: name owners, dates, falsifiers, and budget hooks before scaling a discovery squad. Without guide version control, ventures default to activity metrics (meetings held) instead of learning metrics (assumptions supported or falsified).

Remote shadow limitations

RelayOps uses remote shadow limitations in weekly synthesis and monthly validation committee reviews. Customer success, sales, and engineering read the same RelayOps anchor facts: Segment A 80-250 technician residential-heavy HVAC and plumbing, same-day rebalance job, Core pricing near $2,800 per month, COO overtime trigger near 8 percent.

Remote shadow limitations prevents drift after competitive shocks such as ServiceSuite QuickReroute. Advantage pillar narratives update while Problem and Segment pillars remain stable unless new evidence crosses kill thresholds written in Unit 1 and Unit 6.

Operators should be able to explain remote shadow limitations to a dispatcher, a COO, and a seed investor without changing the core claim. That tri-audience test is the ENT 401 standard for applied validation work.

Part A: Baseline

Beachhead Segment A; 9 paid logos Month 9; cold OT -4%; warm OT -9%.

Part B: Intervention

Apply lesson concept to cold cohort playbook for next 30 days.

Part C: Expected movement

Cold OT toward -7%; DAU toward 68%; services toward 28 hours per logo.

Part D: Managerial read

Link intervention to validation pillar grades. Check: metrics named ✓

RelayOps cold pipeline 22 opportunities; 6 in contract; IT median 52 days.

1. Name two leading indicators for next 30 days.
2. Which Unit 3 assumption register rows move?
3. Write one falsifier sentence.
4. Continue, pivot, or kill if cold OT stays -4% at Day 90?

Solution

1. Cold OT median and IT median days on new pipeline. 2. A2 adoption and A3 integration rows. 3. If cold OT median above -4% at Day 90 with ritual shipped, pivot packaging or segment narrow. 4. Conditional continue until Day 90; pivot if falsifier hits.

Check: falsifier linked to pillar ✓

RelayOps applied review: connecting this lesson to validation

Every ENT 401 lesson supports the same Month 9 validation decision for RelayOps, the B2B SaaS dispatch and scheduling venture serving mid-market HVAC and plumbing firms with 80 to 250 technicians. Maya Chen and Jordan Okonkwo founded RelayOps after operating dispatch at Summit Climate. Their beachhead job is same-day crew rebalance under absenteeism and demand spikes, sold to COOs on overtime reduction near an 8 percent trigger, with Core pricing near $2,800 per month and CRM read integration in phase one.

This subsection ties lesson concepts to pillars investors grade: Problem, Segment, Solution, Economics, Market, and Advantage. Problem and Segment stay strong when qualified operations leaders rank rebalance pain in top three weekly pains and spend on overtime or scheduling modules. Solution weakens when cold cohort dispatcher daily active use sits near 63 percent while warm cohorts reach 76 percent. Economics weakens when customer acquisition cost payback stretches past 20 months and services hours per logo exceed 28. Advantage weakens when ServiceSuite QuickReroute bundles free reroute features that narrow speed-based differentiation.

Operators should translate every abstract framework in this lesson into calendar events, owners, and falsifiers. Founders should write what would change their mind before the next board meeting. Investors should ask for cold cohort tables, not blended averages. Learners should practice explaining RelayOps decisions to three audiences without changing the underlying evidence chain from Units 1 through 6.

Corporate innovators can map the same structure: opportunity memo, segment rules, interview instruments, insight portfolio, sizing brief, validation scorecard. The vocabulary changes by industry; the sequence does not. Selection before segmentation, segmentation before instrument design, instruments before synthesis, synthesis before sizing honesty, sizing before continue or pivot or kill.

Managerial stakes when this lesson is misunderstood

Teams that skip this lesson's discipline usually show predictable failure signatures within two quarters. Sales promises outrun evidence. Engineering builds features no economic buyer funds. Services teams drown in custom integration work. Marketing speaks at category level while dispatchers live at Tuesday morning chaos level. Finance models heroic TAM instead of obtainable SOM tied to account executive productivity.

RelayOps guards against those signatures with written memos, assumption registers, insight portfolios, and Month 12 thresholds. A lesson is not academic when it prevents a $195,000 monthly burn company from raising seed extension on warm cohort fiction. A lesson is not academic when it helps a corporate squad kill an innovation theater project before a seven-figure build.

Re-read the worked examples and practice problems with this validation lens. Each exercise should produce a decision, an owner, and a metric. If an answer only restates theory, revise until a RelayOps operator could execute it Monday morning in Phoenix or Dallas metros where reference density strategy concentrates learning and word-of-mouth among HVAC and plumbing operations leaders.

Study integration checklist for ENT 401 learners

Before moving to the next lesson, confirm you can: (1) state RelayOps beachhead in one sentence with inclusion and exclusion rules; (2) name the core job in situation-motivation-outcome form; (3) cite at least one falsifier with an instrument; (4) identify which validation pillar your lesson topic affects most; (5) describe what warm versus cold cohort split would do to your conclusion if ignored.

If any item is difficult, return to the worked example and practice problem sections. ENT 401 is cumulative by design. Unit 5 sizing fails when Unit 2 segment definition is vague. Unit 6 validation fails when Unit 3 assumption thresholds are missing. Unit 4 synthesis fails when Unit 1 evidence strength hierarchy is ignored.

RelayOps remains the anchor venture so you can see those links across 24 lessons without resetting context. The depth bar from the lesson authoring guide requires prose that teaches, not bullets that index. This integration subsection is intentionally repetitive on anchor facts because repetition builds fluency beginners need before running real discovery programs.

Draft a 30-minute IT discovery guide for RelayOps integration risk (R3). Include: 4 required questions, 2 anti-patterns to avoid, note template fields, saturation signal for IT calls.

Solution

Required questions:

1. "Walk through last software rollout that touched dispatch; what broke?"
2. "What security review steps are mandatory for a mobile SaaS app?"
3. "Typical calendar time from vendor selection to dispatcher login?"
4. "What integration pattern do you prefer: read-only CRM API, webhook, CSV?"

Anti-patterns: "Our API is easy." / "Would IT block this?"

Note fields: review steps list, days estimate, blockers named, CRM API access level, assumption impact R3.

Saturation: Last 3 IT calls repeat same review steps and timeline band without new blockers.

Key takeaways

• Mom Test rules plus B2B buyer and artifact probes structure interviews.
• Guides with required questions keep teams consistent.
• Code notes within 24 hours; count obstacle tags for synthesis.
• Shadows need timing and debrief discipline.
• RelayOps prioritizes tab_switching signal from coded interviews.

1. Build a 45-minute guide for your primary job executor with five required questions.
2. Create a tag list of ten obstacles you expect; run one coded interview.
3. Continue to Lesson 3: Evidence, Metrics and Assumptions in Interview Design and Field Research.