Comparing Product, Service, Platform, and Marketplace Models

Why archetypes matter

Real companies are messy. Amazon sells products, runs a marketplace, operates a cloud platform, and bundles services like Prime. Still, most businesses cluster into four archetypes that imply different scale drivers, risk profiles, and metrics. Clarity about archetype prevents strategic cosplay: calling yourself a "platform" when economics behave like a product company leads to wrong hiring, wrong valuation expectations, and wrong growth tactics.

From Lesson 1, you built a full canvas. From Lessons 2 and 3, you tied revenue models and unit economics. From Lesson 4, you tested industry structure fit. This capstone lesson compares product, service, platform, and marketplace models so you can label honestly, metric correctly, and hybridize deliberately.

The managerial stakes are concrete. A services firm that prices like SaaS without productized delivery will miss margin. A product firm that opens APIs and declares "platform" may still carry product COGS and churn while investors expect ecosystem multiples. A marketplace that subsidizes both sides without liquidity discipline burns cash faster than unit dashboards show.

Four archetypes at a glance

Hybrids are normal. Chaos arrives when one team runs three archetypes without separated metrics and owners.

Product models

Product businesses package value into repeatable artifacts: phones, cereal, shrink-wrapped software, standardized widgets. Economics feature upfront R&D (research and development) or design cost, then marginal COGS per unit, plus distribution and channel costs. Scale helps when variable cost is low and brand or IP blocks commoditization.

Strategic levers include brand, patents, manufacturing scale, channel control, and roadmap pace. Failure modes include feature parity competition, race to the bottom on price, and inventory obsolescence (consumer electronics, fashion).

Modern software often hybridizes: the artifact is code, but delivery is continuous (updates, cloud hosting). Microsoft Office moved from license product toward subscription product-plus-service. The archetype label shifts when capture and delivery change even if the artifact looks similar.

Managers track gross margin by SKU (stock keeping unit, a single product variant), inventory turns, return rates, and channel mix. Unit economics from Lesson 3 apply per unit sold or per subscriber if subscription capture wraps the product.

Distribution is often the hidden governor of product scale. A consumer packaged goods firm with 55% gross margin can still fail if slotting fees and retail programs consume contribution. A hardware startup with strong BOM (bill of materials, component cost stack) margin can bleed cash in warranty returns and reverse logistics. Product archetype thinking must include reverse supply chain costs when guarantees are generous.

Inventory risk deserves explicit mention. Ordering 50,000 units of last year's model ties cash that cannot fund R&D. Product firms in fashion and electronics live or die by markdown cadence. Software packaged as perpetual licenses carried upgrade revenue risk; subscription capture shifted that risk toward retention and continuous delivery cost (hosting, support), changing the archetype economics even when the artifact is still "software."

Service models

Service businesses sell applied expertise: consulting engagements, law, implementation, agencies, healthcare procedures tied to clinician time. Revenue often ties to billable hours, project milestones, or outcome fees. Talent is COGS; utilization (billable ÷ available hours) drives profit.

Strategic levers include reputation, methodology, partner leverage, and narrow specialization. Failure modes include quality variance, key-person dependency, and utilization dips in downturns when clients defer projects.

Scaling paths: productize methodologies (playbooks, training), build software that multiplies expert output, tier leverage with juniors supervised by partners, or focus a domain so repetition lowers delivery cost.

Service models fit high buyer power when the job is bespoke (Lesson 4). They struggle when substitutes commoditize the job (self-serve software eats lightweight consulting).

Quality control is the scaling bottleneck. Two consultants from the same firm can deliver divergent client outcomes if methodology is implicit. Productized services document scopes, templates, and acceptance criteria so margin is repeatable. Utilization without quality yields churn: clients do not renew because the last project felt unlike the first.

Partner leverage models (junior staff billed at blended rates supervised by seniors) increase contribution per partner hour but create training and liability obligations. Professional services firms are portfolio managers of talent: wrong hiring shows up in utilization dips before it shows up in marketing metrics.

Platform models

Platform businesses provide infrastructure others build on: operating systems, cloud APIs, payment rails, app stores. Economics feature high fixed platform investment and low marginal cost to serve another developer or tenant if capacity scales digitally.

Strategic levers include developer ecosystem, complementary innovation, and switching costs via integrations. Failure modes include launching platform before critical mass (not enough complements), mishandling openness vs control (too closed kills ecosystem; too open kills capture).

Cold start tactics subsidize one side (free SDK, revenue share guarantees) until complements create value for the other side. Platform metrics include active developers, API call volume, attach rate of third-party apps, and net revenue retention on platform fees.

Honesty test: if third-party complements do not materially drive customer value, you likely have a product with integrations, not a platform (see worked example below).

Governance choices define platform economics. Apple iOS balances tight control (quality, security) against developer friction. Open platforms grow faster early but suffer spam and quality dilution. API pricing (per call vs revenue share vs free tier) is capture architecture from Lesson 2. Underpricing API access can subsidize ecosystem growth; overpricing drives developers to multi-home on competing platforms, weakening moat.

Platform operators must measure complement quality, not only complement count. Ten thousand low-usage plugins are weaker than fifty plugins embedded in customer daily workflows. Lesson 4 entrant and substitute forces often attack platforms by peeling off the most popular complements (e.g., a hyperscaler launching a native substitute to a popular third-party tool).

Marketplace models

Marketplace businesses match two sides who struggle to find each other efficiently: guests and hosts, drivers and riders, freelancers and clients. Revenue is usually GMV × take rate minus subsidies and trust/safety costs.

Strategic levers include liquidity density (matches per hour), curation, payments, insurance, search quality, and repeat rates. Failure modes include multi-tenanting (users hop among competing marketplaces), disintermediation after first match, and subsidy wars.

Metrics: time to match, repeat purchase rate, contribution per trip or order (Lesson 3 QuickBite-style), subsidy burn, and geographic liquidity heatmaps.

Marketplaces face harsh industry structure when rivalry subsidizes both sides (Lesson 4). Model fit requires discipline on which side to subsidize and when to raise take rate.

Liquidity is local. A marketplace can be dominant in one city and irrelevant in another because density of supply and demand differs. Expansion is not copy-paste; unit economics must be recomputed per geography. Multi-tenanting (drivers on both Uber and Lyft, hosts on multiple OTAs) limits take-rate power: if you raise rates, sides leak to competitors.

Trust and safety costs are variable COGS easy to underestimate: insurance, fraud review, support for failed matches. Lesson 3 contribution per order must include these, not only payment processing. Marketplaces that win long term often own payment flow and reputation systems that make off-platform disintermediation costly.

Choosing and hybridizing deliberately

Decision guide (first-cut):

1. Does the customer need custom application of expertise each time? Lean service.
2. Can value be packaged and replicated at scale? Lean product.
3. Do third parties build complements that multiply your value? Lean platform.
4. Do you match two sides with trust, payment, and discovery needs? Lean marketplace.

Hybrids work when each layer has a clear owner and P&L logic. Shopify combines platform (merchant tools) with payments (marketplace-like financial services). Tesla combines product (vehicles) with service (software updates, repair network). Amazon layers retail product, third-party marketplace, and AWS platform.

Integration without separation creates metric soup: one leadership team celebrates GMV while another needs gross margin per SKU while engineering treats everything as platform uptime.

Amazon as deliberate hybrid (how layers coexist)

Amazon retail (product and marketplace) shares logistics infrastructure but separates seller services P&L from first-party inventory risk. AWS (platform) has its own operating metrics and capital allocation. Prime (service/bundle) increases purchase frequency on retail and ties customer habit. Each layer has different force exposure: retail faces rivalry and buyer power; AWS faces platform competition and developer substitutes; Prime is retention architecture across layers.

Most firms are not Amazon. The lesson is not "be a conglomerate." The lesson is: when you add a layer, assign owner, metric, and force map for that layer. A B2B software firm adding a services arm to boost ARR should not merge services utilization with software NRR in one blended dashboard, or leaders will subsidize failing services with software gross margin without noticing until audits.

Transition paths between archetypes

Companies evolve archetypes intentionally or accidentally. Services firms productize IP into software (service → product). Product firms open APIs and partner marketplaces (product → platform). Product firms add two-sided marketplaces for suppliers (product → marketplace). Each transition changes unit of analysis from billable hour to seat to API call to transaction.

Transitions fail when capture lags delivery: you build marketplace liquidity before contribution per order works; you sell platform narrative before developers ship complements. Use Lessons 2 and 3 thresholds (contribution positive, LTV/CAC workable) as gates before rebranding archetype.

RelayPay sells payment APIs to e-commerce merchants (fictional). Leadership calls it a "platform." Facts:

• 98% of revenue from RelayPay's own fraud scoring and checkout SDK used by merchants.
• 12 third-party plugins (loyalty, tax) exist; combined <8% of merchant value perception surveys cite plugins as reason to stay.
• Marginal COGS: payment processing pass-through + cloud; gross margin 38%.
• Sales cycle 90 days enterprise; churn 14% annual logos.

Part A: Archetype diagnosis

Primary economics resemble product ( proprietary SDK + fraud stack) with integrations. Not a platform in economics sense: complements do not drive marginal value or switching costs materially.

Part B: Metric map if mislabeled vs corrected

Part C: Numeric merchant unit (annual)

Average merchant net revenue to RelayPay: $4,800/year.

Variable COGS 62% of revenue (processing-heavy): $2,976.

Contribution $1,824/year.

CAC $3,200 → payback > 1 year on contribution alone ✓ borderline.

If labeled platform, board might tolerate long payback for "network effects" that are not present. Relabeling triggers product GTM fixes: reduce CAC via partnerships, raise price on fraud value, improve retention.

Part D: Managerial read

Honest archetype labeling saves capital misallocation. Lesson 1 canvas value proposition is fraud + checkout reliability, not ecosystem breadth.

HarborHost matches vacation renters and property owners (fictional Airbnb-like, single city launch).

Part A: Cold start economics

Launch city: 200 listings, weak demand side.

Subsidy: $40 off first booking for guests; owner onboarding free photography worth $150.

Guest CAC analog via promos: $40 per first booking guest.

Take rate 12% on $220 average booking = $26.40 revenue per booking.

Variable trust/safety + support $8 per booking.

Contribution $18.40 per booking.

One booking per guest → LTV/CAC 0.46× (destructive) until repeat.

Part B: Liquidity improvement

After 6 months: repeat guests 35% within year; avg 2.8 bookings per active guest/year.

LTV_contribution ≈ 2.8 × $18.40 = $51.52 vs CAC $40 → 1.29× (still weak; need 3×+ for efficient scale).

Owner side: 200 listings, 55% booked monthly avg 1.4 nights → liquidity improving.

Part C: Model lever

Introduce Superhost subscription for owners: $29/month for priority placement + insurance bundle. 60 owners adopt.

Subscription contribution ~$25/month after variable → $1,500/month high-margin layer.

Hybrid capture reduces pure take-rate pressure (Lesson 2 revenue architecture).

Check: 60 × $25 = $1,500 ✓

Part D: Board questions

1. Which side is subsidized and until what liquidity threshold?
2. Contribution per trip vs corporate fixed ops burn?
3. Multi-tenanting rate vs unique inventory?

Marketplace archetype demands these metrics weekly, not quarterly product roadmap reviews.

Metric cheat sheet by archetype (what to watch weekly)

Product: gross margin by SKU, inventory days, return rate, attach rate of warranties or consumables.

Service: utilization by role, realization rate (collected vs billed), repeat client %, revenue per partner.

Platform: active developers or tenants, API/error rates, third-party revenue share, integration attach driving retention.

Marketplace: liquidity (time to match), repeat transaction rate, contribution per transaction after subsidies, supply exclusivity where legal.

Using another archetype's hero metric creates false confidence. A product company reporting GMV because a sales leader came from a marketplace is a warning sign. A services firm reporting logo count without utilization is another.

Closing integration across Unit 2

Lessons 1 through 5 form a chain: canvas describes logic, revenue and cost structures translate logic into P&L shape, unit economics tests incremental soundness, industry structure bounds the pool, archetypes clarify scale and metrics. When you evaluate any company, walk the chain in order before recommending strategy. Skipping steps produces fashionable answers ("be a platform") without numeric or structural support.

Before your next strategy offsite, ask each leader to write the archetype label, the top industry force, and one unit economic metric on a single index card. Misalignment in the room surfaces faster than slide decks with overlapping buzzwords. Alignment does not guarantee success, but it prevents teams from optimizing conflicting goals.

When advising founders, push for one primary archetype metric in the monthly investor update, not ten vanity metrics. Discipline in measurement follows discipline in model design. Hybrids can add a secondary metric, but the primary must reflect where most contribution dollars originate today, not where the pitch deck hopes to be in three years.

Archetype clarity also guides talent strategy: marketplace operators need city managers and liquidity analysts; platform operators need developer relations and partner engineers; services operators need utilization coaches and methodology designers; product operators need SKU rationalization and supply chain leaders. Hiring "like Netflix" without Netflix economics is a common failure mode.

Classify each company (primary archetype; allow hybrid label). Name one metric that matters most and one structural risk.

1. LexCorp Legal (200 attorneys, client matters billed hourly + fixed fee projects)
2. NanoChip (sells IoT sensors for factories, hardware + firmware updates included)
3. OpenRail (cloud API; 400 third-party extensions; revenue share on extension sales)
4. CityRide (connects commuters with corporate van pools; 15% take rate)

For LexCorp, compute simplified monthly profit: 180 billable attorneys averaging 140 billable hours/month at $320/hour; fully loaded attorney cost $22,000/month each; overhead $1.2M/month.

Solution

1. LexCorp: Primary service (hybrid if fixed-fee productized playbooks). Metric: utilization and realization rate. Risk: key partner departures, demand cyclicality.

2. NanoChip: Primary product (hybrid service via firmware updates). Metric: gross margin per SKU, inventory turns. Risk: commoditization, supply chain disruption.

3. OpenRail: Primary platform (extensions drive value). Metric: active extensions, revenue share per developer, API attach. Risk: cold start if extensions weak; openness vs capture tradeoff.

4. CityRide: Primary marketplace. Metric: liquidity (time to match), contribution per ride, repeat rate. Risk: multi-tenanting, subsidy wars.

LexCorp profit math:

Revenue = 180 × 140 × $320 = $8,064,000/month.

Attorney cost = 180 × $22,000 = $3,960,000.

Contribution before overhead = $4,104,000.

Operating profit = $4,104,000 − $1,200,000 = $2,904,000/month.

Utilization check: 140 billable hours on ~160 available ≈ 87.5% (strong).

Check: $8.064M − $3.96M − $1.2M = $2.904M ✓

Your startup sells analytics software with a public API and 6 integration partners. 95% of customers buy for dashboards; partners optional. Churn 18%; gross margin 72%; sales-led.

1. Primary archetype?
2. One strategic mistake if you fundraise as a "platform play."
3. Write one sentence on how to evolve toward genuine platform economics without rebranding prematurely.

Solution

1. Primary product (enterprise SaaS analytics) with integrations; not platform-dominant.

2. Mistake: spend heavily on developer relations and rev-share before complements drive retention; accept platform valuation multiples implying ecosystem moat that churn and sales cycle contradict; underinvest in core dashboard ROI proof.

3. Evolution sentence: deepen partner-built modules that increase switching costs (embedded workflows in partner CRMs) and measure attach rate where partner features correlate with lower churn before claiming platform economics.

Explain why: platform status is earned when third parties create marginal value customers pay for, not when an API exists.

Key takeaways

• Four archetypes imply different scale drivers, metrics, and risks.
• Honest labeling prevents wrong capital and hiring decisions.
• Hybrids are common; separate P&L and metric ownership per layer.
• "Platform" and "marketplace" require ecosystem or liquidity proof, not branding.
• Connect archetype choice to Lessons 1–4: canvas, unit economics, industry fit.

1. Classify your organization (allow hybrid); list one metric you should track weekly but do not.
2. Identify one layer in a hybrid competitor that subsidizes another layer; guess the strategic reason.
3. Return to the unit page for assessments, or continue to Unit 3: Managerial Problem Solving.