Your first 100 customers are not a generic channel target. First define what counts as a customer, who buys, who uses, what proof they require and how your business delivers value. Then choose a narrow market and run founder-owned experiments that fit the model: direct sales for a complex B2B product, liquidity work for a marketplace, product and margin proof for ecommerce, activation loops for a consumer app, or trust-building for a local service. Scale only after the message, fulfillment and retention evidence repeat.
Direct answer
Start with one business model, one buyer, one painful job and one explicit customer definition. Design the smallest acquisition motion that produces both a customer decision and useful learning; do not treat a signup, follower, download or lead as a customer unless the operating model says so.
What “first 100 customers” really means
The number 100 is an editorial operating target, not a sourced benchmark, success threshold or guarantee. The denominator changes by model: a B2B SaaS account may contain many users; a marketplace has supply and demand participants; ecommerce has orders and purchasers; an app may have users before subscribers; an agency has contracted clients. Freeze the unit before reporting progress.
| Business model | Count as customer | Do not count automatically | Proof required | Measurement limit |
|---|---|---|---|---|
| B2B SaaS | Paying account under the defined contract | Invited user or demo attendee | Buyer, account, paid state and activation | One account may contain many users |
| Marketplace | Completed transaction-side account under the chosen unit | Listing, registration or unmatched request | Supply, demand, match and completion | Two sides need separate denominators |
| Ecommerce | Distinct purchaser with a valid order | Email subscriber or abandoned cart | Order, refund state and purchaser identity | Orders are not unique customers |
| Consumer app | Subscriber or other declared commercial unit | Install or account creation | Activation, paid state and eligibility | Users and customers may differ |
| Local service or agency | Contracted client or completed paid engagement | Inquiry, quote or meeting | Scope, acceptance and payment state | Projects and clients are different units |
Visual slot: Add a real diagram here: business-model-to-first-customer map with buyer, user, account, transaction and payment definitions. Replace this block with the final visual later.
Why business model changes the acquisition playbook
Teece’s business-model research connects value creation, delivery and capture. That is the reason acquisition motion cannot be separated from the model: the founder must reach the actor who can adopt, pay or complete the transaction, then deliver the promised value at a support and margin structure the company can sustain. The paper does not prescribe a channel or a 100-customer formula.
| Model | Primary acquisition motion | Proof before expansion | Operational bottleneck | Premature move |
|---|---|---|---|---|
| B2B SaaS | Founder-led account outreach and assisted proof | Repeated problem, buyer, activation and handoff evidence | Long evaluation and multiple stakeholders | Hiring sales before the founder understands objections |
| Marketplace | Concentrated supply-demand sequencing | Matches complete in one narrow market | Local/category liquidity | Expanding geography before repeat matches |
| Ecommerce | Product proof, contribution logic and repeat intent | Valid orders, returns context and margin inputs | Inventory, fulfillment and trust | Buying broad traffic before offer proof |
| Consumer app | Activation path and purposeful sharing loop | Eligible users repeat the core action | Weak first value or habit cadence | Optimizing installs instead of value |
| Local service or agency | Founder network, proof and referral density | Scoped delivery and client-fit pattern | Trust and fulfillment capacity | Lead volume beyond delivery capacity |
Model rule
Choose the motion that lets the correct buyer experience and evaluate the model’s value path. A channel is ready only when it can be tied to a defined customer, proof artifact, owner and fulfillment plan.
Do things that do not scale before you buy scale
Paul Graham’s original essay argues that early founders often recruit users manually and learn by unusually attentive service. YC’s essential startup advice likewise places direct customer contact before premature scale. Treat both as founder-education guidance, not evidence that any anecdote will repeat or that manual work guarantees growth.
| Founder-led work | Learning captured | Keep manual while | Systemize when | Do not infer |
|---|---|---|---|---|
| Prospect research | Language, role and trigger | The buyer pattern is unresolved | The segment and exclusion rules repeat | List size predicts demand |
| Founder outreach | Reply reasons and objections | Message changes after conversations | A stable message works across comparable accounts | A reply-rate benchmark |
| Assisted onboarding | Steps, friction and first value | The product path is still being learned | The same steps can be documented and delegated | Help caused retention |
| Manual matching | Marketplace constraints and failed matches | Supply-demand sequence is unclear | Match rules repeat in a narrow market | Liquidity exists everywhere |
| Founder delivery review | Promise and fulfillment gap | The offer is being shaped | Quality checks and ownership are stable | Automation is the next priority |
Visual slot: Add a real worksheet here: founder-led outreach tracker with target reason, message version, objection, proof requested, next action and learning owner. Replace this block with the final visual later.
Unscalable-work boundary
Manual work is useful when it exposes the buyer, problem, proof and delivery path. It becomes dangerous when hidden custom work is mistaken for a repeatable product or when founder effort is omitted from cost and capacity decisions.
Choose a narrow beachhead before choosing channels
MIT Sloan’s Disciplined Entrepreneurship material starts with the customer and recommends focusing scarce resources on one market segment; MIT’s beachhead explanation emphasizes similarity in buying, value expectations and word-of-mouth connection. Use those criteria to narrow the test, not to claim dominance or guaranteed adoption.
| Beachhead criterion | Evidence to collect | Good-enough decision | Reject when | Next action |
|---|---|---|---|---|
| Problem intensity | Recent costly or urgent workaround | Named job is repeatedly described | Problem is hypothetical or low priority | Recruit more problem interviews |
| Buyer reachability | Founder has a lawful path to the buyer | Comparable buyers accept conversations | Only mass reach is available | Narrow role, geography or community |
| Shared buying path | Similar stakeholders and review steps | Proof format can repeat | Every account needs a different sale | Split the segment |
| Delivery fit | Team can fulfill the complete promise | One bounded service/product path works | Custom work overwhelms capacity | Reduce scope or change segment |
| Peer connection | Buyers observe or refer within a network | Learning can travel inside the segment | No common context exists | Choose a denser subgroup |
Visual slot: Add a real worksheet here: beachhead market selection with problem intensity, buyer reachability, shared buying path, proof format, delivery fit and rejection rule. Replace this block with the final visual later.
Beachhead test
A beachhead is narrow enough for repeated learning and complete fulfillment, yet connected enough to reveal an adjacent path. “Small businesses” or “everyone who needs productivity” is not a testable first segment.
B2B SaaS: founder-led sales, proof and handoff
Steve Blank’s customer-development material frames early work as testing business-model hypotheses through conversations and experiments. For B2B SaaS, the founder should map problem owner, buyer, user, security or procurement reviewer, first-value event and handoff rule before delegating sales. Conversations are evidence inputs, not guaranteed pipeline.
| B2B stage | Founder action | Proof artifact | Handoff evidence | Stop rule |
|---|---|---|---|---|
| Discovery | Interview recent behavior and current workaround | Problem record with context | Consistent role and urgency pattern | Polite interest without a real process |
| Qualification | Confirm buyer, user, trigger and consequence | Account qualification note | Defined next decision and owner | No access to decision process |
| Proof | Run a bounded product or service proof | Agreed success observation and limitations | Buyer reviews the evidence | Proof becomes unpaid custom consulting |
| Close | Document scope, price and obligations | Signed order or verified paid state | Commercial acceptance | Verbal enthusiasm only |
| Onboarding | Observe first value and support burden | Milestone and issue log | Repeatable activation and ownership | Founder work remains invisible |
B2B decision
Do not hire a sales team to discover the message the founder has not learned. Handoff begins when qualification, proof, objections, next steps and delivery responsibilities can be taught and reviewed.
Marketplace: supply, demand and liquidity sequencing
A marketplace has at least two acquisition problems and a match problem between them. Choose one constrained geography, category, time window or use case; identify which side is harder to recruit or retain; seed enough of the constrained side to test completed matches; then recruit the other side. Registration volume does not prove liquidity.
| Marketplace condition | Sequence | Measure | Founder work | Do not claim |
|---|---|---|---|---|
| Supply is scarce | Recruit and verify bounded supply first | Available qualified supply and response | Manual onboarding and availability check | Listings guarantee demand |
| Demand is scarce | Recruit specific demand around ready supply | Qualified requests and completed matches | Manual routing and follow-up | Traffic guarantees transactions |
| Quality varies | Verify criteria before opening volume | Rejected, accepted and completed reasons | Manual review and education | More choice is always better |
| Cold start is local | Concentrate one place/category | Time to match and failed-match reason | Concierge matching | One market proves another |
| Repeat is unclear | Observe post-transaction need | Eligible repeat behavior by side | Interview both sides | First transaction proves retention |
Visual slot: Add a real diagram here: marketplace supply-demand sequencing with constrained side, readiness rule, match event, completion event, failure reason and expansion gate. Replace this block with the final visual later.
Marketplace boundary
Count both sides and completed matches separately. A first-customer plan fails when it reports registrations while hiding unavailable supply, unqualified demand, failed matches or manual subsidies.
Ecommerce: product proof, margin and repeat purchase
For ecommerce, the first-customer motion connects product promise, audience, order economics, fulfillment and return learning. Use small, traceable tests of one product-market message; preserve landed product cost, fulfillment, payment, service and return inputs; and distinguish a first order from repeat purchase. Do not borrow a conversion or repeat-rate benchmark.
Ecommerce decision
Test whether a specific buyer accepts a specific product promise and whether the complete order can be fulfilled within an owned cost model. Broad paid traffic cannot repair weak proof or unknown fulfillment economics.
Visual slot: Add a real diagram here: ecommerce proof chain from message and product evidence to order, fulfillment, support, return state and repeat eligibility. Replace this block with the final visual later.
Consumer app: activation, habit and referral loops
A consumer app may reach many users before it reaches customers. Define the commercial unit separately from the activated user, specify the first-value event and natural return cadence, and log who shares, who receives and whether the recipient activates. A referral is a traceable path, not a viral promise; installs alone do not establish value.
Consumer decision
Choose one audience and one repeated job. Measure eligible activation and purposeful return before spending to enlarge the install base, and treat sharing as a product behavior that needs its own denominator.
Visual slot: Add a real worksheet here: consumer activation and referral loop with eligible user, first value, return cadence, sender, recipient, activation and broken-loop reason. Replace this block with the final visual later.
Local service or agency: trust, proof and referral density
A local service or agency sells trust, scoped delivery and capacity. The founder can start with a narrow service, buyer situation and geography or professional network; show verifiable work without inventing results; document qualification and scope; and ask for introductions only after a real delivery experience. Lead volume that exceeds fulfillment capacity is not progress.
| Trust layer | Evidence | Founder action | Scale condition | Risk |
|---|---|---|---|---|
| Problem fit | Recent buyer situation | Ask about trigger and current alternative | Qualification pattern repeats | Generic networking |
| Capability | Real process or approved work sample | Explain scope and limits | Proof can be reused accurately | Fabricated case result |
| Commercial clarity | Written scope and acceptance | Separate inquiry, quote and client | Pricing and exclusions repeat | Free discovery becomes delivery |
| Delivery | Owner, capacity and quality review | Track rework and client questions | Service standard is teachable | Volume harms quality |
| Referral density | Introductions within a defined network | Request contextual introduction | Source and fit can be traced | Referral assumed to be free |
Service boundary
A meeting is not a client and a referral is not a contract. Count the commercial unit only when scope and acceptance match the customer definition.
Measurement workflow without fake acquisition benchmarks
Shepherd and Gruber’s academic treatment of the Lean Startup framework connects hypothesis formulation, testing and learning; Sarasvathy’s effectuation research distinguishes prediction from control under uncertainty. Use these as context for bounded experiments and controllable founder actions, not as universal channel rules. The publisher page for Traction, The Mom Test’s first-party book site and The Lean Startup’s official site are workbook-approved practitioner references, not benchmark datasets.
first_customer_focus = problem_intensity × buyer_reachability × proof_strength × founder_access. problem_intensity is evidence of a consequential current job; buyer_reachability is a lawful direct path to the decision participant; proof_strength is the quality of verifiable value evidence; founder_access is the founder’s ability to learn from the customer path.
channel_learning = qualified_conversations × objection_quality × activation_signal × repeatability. qualified_conversations are interactions meeting the declared fit rule; objection_quality is specificity about buying and value barriers; activation_signal is the defined first-value observation; repeatability is evidence that comparable cases follow a similar path.
scale_readiness = repeatable_message × conversion_evidence × fulfillment_capacity × retention_signal. repeatable_message is stable problem and proof language; conversion_evidence is the startup’s own mature decision-path evidence; fulfillment_capacity is deliverable volume at defined quality; retention_signal is eligible repeated value under a stated window.
paid_test_caution = unclear_positioning × weak_proof × low_activation × unknown_unit_economics. unclear_positioning is unresolved audience and promise; weak_proof is insufficient verifiable evidence; low_activation is failure to reach the named first value; unknown_unit_economics is missing full acquisition and delivery inputs. These are editorial planning formulas, not guaranteed acquisition models or universal benchmarks.
| Formula input | Company evidence | Decision use | Review cadence | Misuse |
|---|---|---|---|---|
| problem_intensity | Interviews and current-workaround records | Choose the narrow problem | Every segment revision | Scoring opinions as facts |
| buyer_reachability | Founder contact-path log | Choose reachable beachhead | Weekly experiment review | Treating list size as access |
| qualified_conversations | Documented fit criteria | Compare message learning | After complete test batch | Using raw replies |
| conversion_evidence | Mature customer decisions with denominators | Assess repeatability | After observation window | Borrowing a benchmark |
| fulfillment_capacity | Owner, time, cost and quality log | Set the scale ceiling | Before volume change | Ignoring founder labor |
| retention_signal | Eligible repeated value | Check post-acquisition fit | By mature cohort | Using any return visit |
Visual slot: Add a real worksheet here: channel learning and scale-readiness scorecard with definitions, company evidence, owner, window, uncertainty and stop rule. Replace this block with the final visual later.
Measurement rule
A channel “works” only for a stated segment, offer, proof, customer definition, observation window and fulfillment capacity. Report learning and uncertainty before efficiency claims.
Weak acquisition plan versus stronger acquisition plan
| Layer | Weak acquisition plan | Stronger acquisition plan | Evidence created | Remaining limit |
|---|---|---|---|---|
| Target | Small businesses | Operations leaders at one defined B2B workflow and trigger | Beachhead record | Adjacent markets remain untested |
| Customer | Any signup | Paying account that reaches the named first value | Customer and activation contract | Later retention is unknown |
| Channel | Use SEO, social, outbound and ads | Run one founder-led motion against one hypothesis | Message and objection log | Results are context-specific |
| Scale | Spend when leads appear | Expand after message, decision path and delivery repeat | Scale-gate review | Causality may still require testing |
| Success | Reach 100 quickly | Build a traceable path from first 10 to next comparable accounts | Cohort and change log | The target is not a guarantee |
Stronger plan
A stronger plan states who counts, why the founder can reach them, which proof matters, what unscalable work is allowed, how fulfillment is protected and what evidence blocks scale.
Detailed operational example
Assume a hypothetical B2B SaaS helps independent logistics brokers collect carrier documents before onboarding. The buyer is the operations lead; users are coordinators; the beachhead is small brokerages that still chase documents through email. This is an illustrative planning case with no real customer, conversion, CAC, revenue or retention data.
| Example element | First 10 customer plan | Path from 10 to 100 | Weak plan | Stronger plan |
|---|---|---|---|---|
| Reach | Founder introductions and carefully researched accounts | Repeat the proven account profile through founder-led outbound and selected partners | Buy a broad lead list | Keep one role, trigger and exclusion rule |
| Unscalable work | Founder maps the existing document flow and assists setup | Turn repeated setup into a documented handoff | Hide manual setup | Log founder time and automation candidates |
| Proof | Show a bounded workflow using customer-approved inputs | Reuse only accurate proof for comparable accounts | Promise time savings | Show process completion and limitations |
| Experiment | One message and proof hypothesis per batch | Change one acquisition variable at a time | Run many channels together | Version message, segment and proof |
| Scale gate | Named buyer, paid state and first-value observation | Message, decision path, fulfillment and repeat value recur | Scale after replies | Scale only after mature evidence |
| Measurement item | Definition | Owner | Decision | Do not claim |
|---|---|---|---|---|
| Qualified conversation | Target account confirms current process and decision role | Founder | Continue or revise beachhead | Reply rate is universal |
| Customer | Paid account under written scope | Finance operations | Count toward operating target | Every user is a customer |
| Activation | Account completes the first verified carrier-document workflow | Product | Repair setup or proof | Activation guarantees retention |
| Objection | Specific barrier with stage and stakeholder | Founder | Change message, proof or segment | One objection represents market |
| Fulfillment | Setup and support delivered within declared capacity | Operations | Set scale ceiling | More demand is always better |
| Repeat value | Eligible account repeats the core workflow in its normal cadence | Product | Review post-acquisition fit | Early repeat predicts revenue |
Visual slot: Add a real roadmap here: first 10 to first 100 with customer definition, founder tasks, message versions, evidence gates, fulfillment capacity and change dates. Replace this block with the final visual later.
Audit the first-customer motion before adding channels
Bring the customer definition, beachhead, founder outreach, proof and fulfillment constraints into one scoped review. The audit produces a test sequence and evidence gaps without promising the first 100 customers, growth, revenue, conversion, lower CAC or faster scale.
What to remove, rewrite or keep audit-only
| Claim or artifact | Remove | Rewrite safely | Keep audit-only | Evidence needed |
|---|---|---|---|---|
| “This channel gets the first 100” | Universal channel claim | State segment, motion and test boundary | Founder anecdote | First-party experiment with definitions |
| “Our CAC is good” | Unsourced benchmark comparison | Report owned inputs and maturity | Incomplete cost estimate | Full acquisition and fulfillment cost |
| “Marketplace liquidity achieved” | Signup-based conclusion | Report matched and completed units by side | Early registration total | Defined match and completion evidence |
| “Referrals will scale” | Unmeasured assumption | Track source, fit and completed customer path | Informal praise | Traceable introduction cohorts |
| “Paid ads validate demand” | Causal or product-fit claim | Use a bounded message test with caution | Platform-attributed result | Appropriate experiment and activation evidence |
Common first-customer mistakes
Common mistakes are counting users as customers, selecting channels before the model, targeting an unbounded market, outsourcing founder learning, hiding custom delivery, running several experiments at once, scaling from replies rather than mature customer decisions, ignoring retention and fulfillment, and copying famous startup anecdotes as instructions. Another mistake is declaring a channel successful without its segment, denominator, window and exclusions.
Mistake check
If the plan cannot name buyer, customer unit, proof, founder action, fulfillment owner, observation window and stop rule, it is not ready for channel scale.
FAQ
Is there one best channel for getting the first 100 customers?
No. The useful channel depends on who buys, who uses, how value is delivered, what proof is required and whether fulfillment can support the customer. Founder-led outreach may fit a complex B2B sale; concentrated matching may fit a marketplace; trust and referrals may fit a local service. Test one bounded motion against your own customer definition instead of adopting a universal channel ranking.
Should I use paid ads before product-market fit?
Paid tests can be used cautiously for a narrow message or audience question, but they do not prove product-market fit and should not outrun weak activation, proof or unit economics. Define the hypothesis, budget boundary, eligible visitor, customer event and stop rule first. Keep platform-attributed results separate from causal conclusions, and do not scale spend merely because clicks, leads or early signups appear.
Do things that do not scale still matter?
Yes, when manual founder work reveals the buyer, problem, proof, onboarding and delivery path. Direct recruitment, assisted setup or manual marketplace matching can produce learning that automation would hide. Record founder time and custom work, however. The goal is not permanent heroics: systemize only after repeated cases show what should remain personal, what can be standardized and what should not be offered.
How do I define a customer in the first 100?
Tie the definition to the business model and commercial unit. A B2B SaaS customer may be a paid account, not each user; ecommerce may count distinct valid purchasers, not orders; a marketplace may need separate side and completed-match measures; an agency may count contracted clients. Document payment or acceptance state, eligibility, refund or cancellation handling and deduplication before reporting the operating target.
What should a marketplace do first: supply or demand?
There is no universal sequence. Identify the constrained side in one narrow geography, category, time window or use case. If qualified supply is unavailable, recruit and verify it before driving demand; if demand is uncertain but supply is ready, test specific buyer requests. Track readiness, match, completion and failure reasons separately. A registration count on either side does not by itself demonstrate marketplace liquidity.
When should I scale the channel?
Scale only after comparable customers repeatedly understand the message, complete the decision path, receive the promised value and fit available fulfillment capacity. Use mature observation windows and include founder labor, support and delivery constraints. A channel may still require controlled testing for causal claims. If the segment, proof, activation or unit economics changes materially, treat the next expansion as a new experiment rather than a continuation.
What I would test first
I would first test whether one reachable buyer group recognizes one consequential problem and will take a concrete next step after seeing one accurate proof artifact. I would record objections and observe the complete customer and fulfillment path. Only after that pattern repeats would I test a less manual motion or paid channel, with the original founder-led path retained as a comparison and learning source.
First test
Select one customer definition, beachhead, message, proof artifact and founder-owned acquisition motion. Predefine the batch, decision event, observation window, fulfillment ceiling and reasons to stop.
Turn channel ideas into a first-customer evidence plan
A startup acquisition review can map the business model, beachhead, founder work, proof and scale gates into a bounded experiment plan. The deliverable is decision support, not a promise of customers, growth, revenue, conversion, lower CAC, investor interest or faster scale.