How to use this guide
No textbook chapter required. For each term, you’ll find four things: meaning → decision → common trap → practical example. Examples are illustrative, not client claims. Jump to the chapter that matches your current challenge.
The bigger picture: one product, twenty connected decisions
Picture a lending platform where applicants keep abandoning onboarding. The team could rush to redesign screens. Or it could start by asking what applicants are actually trying to get done. That is the customer’s job (JTBD), builds a small test (MVP), tests feasibility where needed (POC) and checks whether a repeatable segment actually values it (PMF). It then ranks choices (RICE), protects essential scope (MoSCoW), documents rules (PRD) and defines outcomes (OKR). After release, it follows real experience (UX), meaningful customer value (NSM) and commercial health (CAC, LTV, MRR, ARR, NRR, ROI). You don’t need to recite all twenty in a meeting. You need to know which question to ask next.
CHAPTER 01 ◈ DISCOVER
Discovery & validation
Before investing in delivery, decide whose problem matters, what evidence would change your mind, and how little you can build to learn.
1. MVP, Minimum Viable Product
An MVP is a small but usable version of a product built to learn something important from real users. Small is not the goal; useful evidence is.
→ THE DECISION
Define the hypothesis, target customer, success signal and stop condition before listing features.
! COMMON TRAP
Removing ten features doesn’t automatically make something an MVP. Without a clear question to test, you’ve just built a smaller product.
→ IN PRACTICE
For a lending onboarding journey, test whether eligible applicants can complete the essential steps before adding optional dashboards.
2. PMF, Product-Market Fit
Product-market fit is what you start to see when the same kind of customer chooses your solution, gets value and keeps coming back. One enthusiastic demo isn’t enough.
→ THE DECISION
Which segment returns, pays, refers or expands without unsustainable incentives?
! COMMON TRAP
Five people saying ‘love the idea’ feels good. It still doesn’t tell you whether they’ll use it next month, or pay for it.
→ IN PRACTICE
Track cohort retention, repeat usage and willingness to pay for the same target segment.
3. JTBD, Jobs to Be Done
JTBD starts with a deceptively simple question: what is the customer really trying to get done? Their proposed feature may be only one possible answer.
→ THE DECISION
What situation triggers the job, what alternatives do people use, and what outcome matters to them?
! COMMON TRAP
Confusing a suggested feature with the underlying job can produce elegant but irrelevant software.
→ IN PRACTICE
An applicant’s job may be ‘know whether I qualify without wasting a day,’ not ‘complete an eight-page application.’
4. POC, Proof of Concept
Before committing to a full build, a POC checks whether a difficult technical or operational idea can work under defined conditions.
→ THE DECISION
What uncertainty must be resolved before funding a production system?
! COMMON TRAP
A demo that works on Friday afternoon is encouraging. It isn’t proof customers need the solution, or that it will survive Monday morning operations.
→ IN PRACTICE
Test whether a document-extraction component handles varied documents before building the complete workflow.
5. TAM, Total Addressable Market
TAM is the full theoretical revenue opportunity for a defined market. Think of it as the size of the whole playing field, not your sales forecast.
→ THE DECISION
What market boundaries, pricing unit and realistic serviceable subset underpin the estimate?
! COMMON TRAP
A slide saying ‘$10 billion market’ is not a customer acquisition plan. Ask who you can actually reach and serve.
→ IN PRACTICE
Separate all possible lenders from the actual lender segment you can reach, sell to and support.
CHAPTER 02 → DECIDE
Prioritization & delivery
Once opportunities are visible, turn them into explicit trade-offs and delivery decisions that engineering, commercial teams and leadership can defend.
6. RICE, Reach, Impact, Confidence, Effort
RICE helps compare possible bets using reach, impact, confidence and effort. It’s a conversation aid, not a machine that makes the decision for you.
→ THE DECISION
Where did the estimates come from? If reach and impact are guesses, the neat-looking score is still a guess.
! COMMON TRAP
A high score is not permission to ignore mandatory compliance work, strategy or dependencies.
→ IN PRACTICE
Score two onboarding improvements with the same time window, then examine the uncertainty behind the ranking.
7. MoSCoW, Must Have, Should Have, Could Have, Won’t Have (this time)
MoSCoW puts requirements into Must, Should, Could and Won’t Have (this time). Its real value is making the ‘not now’ decisions visible.
→ THE DECISION
Which capabilities are necessary for the agreed outcome and which can be deferred without breaking it?
! COMMON TRAP
If everything gets stamped ‘Must Have’, the exercise has become a wishlist with four labels.
→ IN PRACTICE
For an initial loan application, eligibility and consent may be essential while personalization is deferred.
8. WSJF, Weighted Shortest Job First
WSJF helps teams weigh the cost of waiting against the size of the job. It’s useful when timing matters as much as potential value.
→ THE DECISION
What is the cost of waiting, including time sensitivity, risk reduction and opportunity enablement?
! COMMON TRAP
Treating uncertain cost-of-delay estimates as exact financial predictions creates false confidence.
→ IN PRACTICE
A regulatory deadline may outrank a useful dashboard even when the dashboard is easier to build.
9. OKR, Objectives and Key Results
An OKR pairs a direction worth pursuing with measurable signs that you’re getting there. ‘Ship the dashboard’ is a task; ‘reduce time to a useful answer’ is an outcome.
→ THE DECISION
Which user or business outcome should improve, rather than which activity should the team complete?
! COMMON TRAP
‘Ship five features’ measures output; it does not show whether customers benefit.
→ IN PRACTICE
Objective: reduce onboarding friction. Key result: increase verified completion rate from a measured baseline.
10. PRD, Product Requirements Document
A PRD is the document that helps people build the same product they discussed. It records the problem, expected behaviour, business rules and what ‘done’ means.
→ THE DECISION
Are acceptance conditions, exceptions, dependencies and ownership clear enough for implementation and testing?
! COMMON TRAP
A polished Figma screen won’t tell engineering what to do when consent expires or a verification fails.
→ IN PRACTICE
For a loan flow, define what happens when income data is missing, consent expires or verification fails.
CHAPTER 03 ◎ DELIVER VALUE
Experience & growth
A product is not successful merely because it functions. These terms help explain whether customers understand, complete and return to the experience.
11. UX, User Experience
UX is what happens from the moment someone tries to use the product until they finish, or give up. It includes clarity, trust, errors and recovery, not just colours and buttons.
→ THE DECISION
At which moment does the customer lose understanding, confidence or the ability to complete the job?
! COMMON TRAP
Equating UX with visual polish ignores broken workflows and unclear consequences.
→ IN PRACTICE
Observe the complete application path, including rejected documents and resuming an interrupted session.
12. NSM, North Star Metric
The North Star Metric is the one measure that best reflects customers receiving value over time. For many products, raw sign-ups won’t tell that story.
→ THE DECISION
Does the chosen metric represent meaningful value delivered and plausibly lead to sustainable business outcomes?
! COMMON TRAP
Choosing registrations as a North Star can reward acquisition without successful usage.
→ IN PRACTICE
A workflow product might measure successfully completed customer jobs, alongside guardrails for quality and cost.
13. PLG, Product-Led Growth
In PLG, the product does some of the growth work: people can experience value, adopt it and often expand usage without a sales conversation at every step.
→ THE DECISION
Can users discover and realize value quickly enough to support self-service growth?
! COMMON TRAP
A free trial without a path to value is not a growth strategy.
→ IN PRACTICE
Map time-to-first-value, successful activation and what users must experience before upgrade.
14. ICP, Ideal Customer Profile
An ICP describes the companies you’re genuinely best equipped to serve, not every company that could theoretically buy the product.
→ THE DECISION
Which buyer attributes predict urgency, ability to adopt and commercially viable delivery?
! COMMON TRAP
‘We serve SMEs’ is a starting point, not an ICP. Two businesses of the same size may have completely different buying problems.
→ IN PRACTICE
A lending infrastructure provider may focus on a narrower class of lenders with a specific workflow gap.
15. CAC, Customer Acquisition Cost
CAC asks what you spent to win a new customer. The answer only means something if everyone agrees which costs and time period count.
→ THE DECISION
Which channel, cohort and cost boundary produced the number, and can its economics be repeated?
! COMMON TRAP
Comparing blended CAC to a single-channel result or excluding sales costs misleads decisions.
→ IN PRACTICE
Measure acquisition spend and appropriate sales/marketing costs against new paying customers by cohort.
CHAPTER 04 ◆ MEASURE
Commercial performance
These are business measures, not trophies. Each becomes useful only when the underlying accounting, time window and customer behavior are understood.
16. LTV, Customer Lifetime Value
LTV estimates what a customer relationship is worth over time. The important distinction is whether you’re talking about revenue or what remains after serving the customer.
→ THE DECISION
Is expected customer value sufficient after serving, supporting and retaining that customer?
! COMMON TRAP
A revenue-only LTV compared with fully loaded CAC can overstate attractiveness.
→ IN PRACTICE
Model contribution margin, retention, discounts and service cost, then stress-test the assumptions.
17. MRR, Monthly Recurring Revenue
MRR puts recurring subscription revenue on a monthly basis, so teams can see what is building, shrinking or disappearing.
→ THE DECISION
What portion of revenue is predictable and how is it changing through additions, expansion, contraction and churn?
! COMMON TRAP
Annual invoices, services income and inconsistent recognition can distort the headline.
→ IN PRACTICE
Track opening MRR plus new, expansion, contraction and churn MRR to explain movement.
18. ARR, Annual Recurring Revenue
ARR expresses recurring subscription or contract revenue on an annual basis. Useful for the trend; not a substitute for revenue recognition or cash flow.
→ THE DECISION
Does contracted recurring revenue translate into collected cash and healthy retention?
! COMMON TRAP
ARR is not the same as realized revenue, cash in bank or an audited forecast.
→ IN PRACTICE
Use ARR alongside renewals, collections, implementation capacity and contract quality.
19. NRR, Net Revenue Retention
NRR looks at the revenue you kept from existing customers after upgrades, downgrades and cancellations, without counting brand-new customers.
→ THE DECISION
Are current customers growing in value enough to offset losses within the same cohort?
! COMMON TRAP
Strong NRR can mask weak new-customer acquisition; weak NRR needs cohort diagnosis.
→ IN PRACTICE
Calculate (opening cohort revenue + expansion − contraction − churn) ÷ opening cohort revenue.
20. ROI, Return on Investment
ROI asks a plain question: compared with what we spent, what did we actually gain? The difficult part is measuring the real benefit fairly.
→ THE DECISION
What is the credible counterfactual, and have we counted implementation, ongoing cost and uncertainty?
! COMMON TRAP
Counting only gross benefits or claiming an unmeasured causal effect makes ROI a sales slogan.
→ IN PRACTICE
Compare avoided processing hours and incremental outcomes against build, integration, training and operating costs.
THE DECISION LAB / WORKED EXAMPLE
Should we build the next feature?
Situation: Imagine a B2B onboarding product with 100 new accounts this month. Only 38 finish verification. Sales wants a new reporting dashboard. Customer success says applicants are getting stuck uploading documents. Engineering has capacity for one meaningful improvement. What would you do first? These figures are hypothetical and do not represent a Parikh Advisory client.
- Frame the job (JTBD): Customers need to complete verification with confidence and minimal rework.
- Set the result (OKR/NSM): Increase verified completions while protecting fraud controls and support capacity. Define the event and cohort before measuring.
- Discover the cause (UX): Observe sessions and support cases. Is failure due to missing explanation, low-quality capture, a technical defect or ineligible users?
- Test the smallest intervention (MVP/POC): Prototype clearer instructions, or run a narrow extraction feasibility test if technical accuracy is uncertain.
- Rank the alternatives (RICE/WSJF): Compare evidence quality, likely affected users, benefit, effort and cost of delay. Keep compliance obligations outside the popularity contest.
- Record the choice (PRD/MoSCoW): Specify valid documents, exception handling, recovery paths and what is excluded from this release.
- Inspect the economics (CAC/LTV/ROI): If more customers activate, assess whether lower support cost or improved contribution justifies the investment.
Decision rule: Don’t fund the next feature until you’ve identified the bottleneck you can actually substantiate. If observations show that the customers are actually ineligible, interface polish will not fix the root cause. That is why a decision framework is more valuable than an acronym list.
Five questions for your next product meeting
- Which customer’s job are we improving, and how do we know?
- What is the weakest assumption behind the proposed investment?
- What is the smallest test capable of challenging that assumption?
- Which outcome and guardrail will tell us whether it worked?
- What evidence would make us stop, defer or change the plan?
Where Parikh Advisory fits
A real example: In the PlantDecoder engagement, the application could work technically while questions about launch readiness, pricing and acquisition sequencing still needed decisions. A working product and a market-ready business are not the same thing.
This is the kind of reasoning behind Product Audit and the Strategy Blueprint: understanding the problem, mapping the journey and sequencing decisions before funding unnecessary execution. For a deeper look at backlog trade-offs, read How to Prioritize Product Features.
Have a product decision with real commercial consequences? Discuss the decision with Parikh Advisory. Logic Before Code.
Further reading & methodology
The definitions above reflect common product-management usage, but organizations vary in how they calculate commercial metrics and apply frameworks. State your formulas and measurement windows explicitly.