Geek, Professional, B2B, Consumer: Markets as Evidence Environments

Every new-product discussion hits the same two questions. B2B or consumer first? Then the familiar line appears:

Geek → Professional → B2B → Consumer

It fails all the time: products go straight from professionals to consumers, or win a huge consumer base and then retreat to professional scenarios. These are not four levels of maturity but four evidence environments.

Before choosing the next stop, ask:

What unknown is most likely to kill this product right now, and which market can expose it at the lowest cost with the least distortion?

Market entry is about the best learning environment, not the biggest customer base.

Technology maturity doesn’t tell you whom to serve first

NASA’s Technology Readiness Level describes how a technology moves from proof-of-concept to operational use; it can’t prescribe a market sequence. NASA: Technology Readiness Levels

Squeeze technical readiness, user value, workflow, delivery economics and scale into one line and you get three misjudgments:

Because the technology works, assume users will switch from their existing solution;
Because someone is willing to try, assume the business model replicates;
Because the data looks good in a small market, assume the evidence transfers to a big one.

NSF’s I-Corps treats the road from lab to market as another set of hypotheses. NSF: About I-Corps

Find the biggest unknown first

A new product usually faces five unknowns at once.

Uncertainty	The question to answer	What goes wrong if you get it wrong
Value	Do users care about this outcome, and will they change behavior or pay?	Novelty without a reason to come back
Workflow	Which step does the product enter; how do people, AI and devices divide the work?	Strong capability, no completed task
Technology & system	Do performance, reliability, safety, battery and cost hold up in real use?	Demo works; real environments fail
Delivery & economics	Can sales, deployment, service and unit economics be replicated?	Profitable projects; losses at scale
Scale	Can acquisition, onboarding, retention, brand and support absorb many users?	Traffic surges; returns and churn climb

Priority is not about which experiment is easiest. Three tests: does the wrong judgment kill the product? How weak is the evidence? Is it about to trigger large, hard-to-reverse spending? Validate the high-impact, weakly evidenced unknown about to trigger irreversible spending.

Geeks: find the capability boundary, don’t prove low friction

Geek is not an age, income or occupation label, but people willing to pay configuration, learning and fault-tolerance costs.

Best at answering:

Unexpected uses in the real world
Performance and compatibility boundaries
Extensibility and integration
A community willing to build

Geeks underestimate configuration, charging and maintenance, and overestimate customizability and novelty — a product they hand-calibrate may lose the average user at setup.

Professionals: prove the workflow, not complexity

Professionals are people whose income, output or career results depend on a task.

Best at answering:

Which step it replaces or strengthens
The quality threshold that makes results useful
Which improvement users pay for
Exceptions that must stay under human control
Repeated use and showable output

Professionals bring their skills, equipment and complex processes into the product. Batch operations, deep parameters and dedicated integrations may be real needs in a professional market, and not belong in the consumer version.

Domain B2B: prove delivery economics, don’t let customization masquerade as product

Domain B2B puts the product inside real organizations — budgets, permissions, procurement, deployment and accountability. It tests whether results count as revenue, cost, efficiency or risk; whether the system runs reliably across many people, permissions and long cycles; what sales, training, deployment and support cost; and whether contracts renew, expand and replicate.

But a customer paying is not a market existing. A big contract may demand dedicated processes, proprietary interfaces, an on-site team and special hardware; the project has revenue, the product doesn’t.

B2B pilots need customization guardrails:

A customer-specific requirement must state which reusable capability it is testing;
Pricing must include implementation, support and opportunity cost, not just contract value;
Success criteria must include a replicable deployment, not a single acceptance;
One lighthouse customer cannot substitute for repeated evidence from several similar customers.
Consumer: prove the product stands alone, don’t let traffic hide weak value

The consumer market strips away the protection that tolerant users provide: whether users can onboard themselves, perceive a result quickly, come back after the novelty fades, and whether channel, returns and support costs are affordable — all exposed.

One viral moment, a price cut, or a platform recommendation can produce huge registrations without repeated value; as numbers rise, teams ignore segmented retention, organic use, returns and support cost.

Watch whether users return without reminders and whether word-of-mouth survives without the team.

The same person can play different market roles

The same person can play different roles: a photographer buying a prototype camera acts as a Geek; buying a mature system, as a Professional. The classification doesn’t run on company size, price point or user count — only on what this group is validating right now.

Market role	Best at proving	Primary evidence	Biggest distortion
Geek	Technical possibility, unexpected uses, ecosystem interest	Prototype use, modifications, issues, community contributions	Overestimates fault tolerance and novelty
Professional	Workflow, quality threshold, willingness to pay	Repeated tasks, time and quality gains, output	Pushes the product toward over-complexity
Domain B2B	ROI, deployment, service, unit economics	Contracts, renewals, expansion, implementation cost	Treats customized revenue as a replicable product
Consumer	Onboarding, retention, channel, scale	Activation, segmented retention, returns, support cost	Lets traffic mask a weak pain point

Market roles are not labels. They are evidence responsibilities.

Write down how you will leave before you enter

Too many pilots end up as “we’re still learning,” because nobody wrote down what would change the decision. Every learning market should have five gates.

Stage gate	Answer in advance
Entry	What state must product, safety, supply and support reach before entering
Success	Which behavioral or economic outcomes would support the core hypothesis
Failure	What result would overturn the current hypothesis
Exit	What evidence means this market has done its job
No-go	Which safety, compliance, cash or reputation risks are unacceptable

“Great feedback,” “the customer wants to keep talking,” and “the numbers are still growing” are not stage gates — criteria need time, sample, scenario and behavior bounds.

Stage gates also constrain product expansion: features added during a pilot should serve the hypothesis under test, or the product absorbs demands from all four markets while the unknowns never shrink.

Market paths must allow returns

Geek → Professional → Consumer for exploring technology, then proving workflow, then validating low-friction scale.

Professional → Domain B2B → Consumer for confirming a high-value task, then validating organizational delivery, then simplifying complex capability for the mass market.

Domain B2B → adjacent B2B is fully legitimate: if the value lives in an industry process, don’t force a consumer story.

Sometimes you see Consumer → Professional → Consumer: mass-market testing exposes an unclear task, the team retreats to professionals to redefine it, then returns to scale validation.

Returning is not failure — it means the current market cannot answer the next question. Weak consumer retention may be a broken core task; a stalled geek community may mean the exploration is finished.

Each return needs a written answer: what unknown will be killed again, and what has to change before re-entry.

Three products shouldn’t take the same path

Three illustrative scenarios, not real companies.

Enterprise AI. Model capability is already usable; the biggest unknown is whether it produces ROI inside real permissions and processes. Domain B2B is the right learning market, but pilots should center on one reusable task and limit proprietary interfaces and on-site support. The exit condition is several similar customers deploying and renewing the same solution — not one big contract.

AI creation tools. The biggest unknown is which step AI should take over and who judges quality. Professional users supply high-frequency tasks and strict feedback; once workflow and quality thresholds stabilize, Consumer tests onboarding, expression and retention. The professional control panel should not be carried into the consumer version unvalidated.

Smart wearable hardware. While sensors, compute and the control chain are unstable, Geeks expose the technical boundary. But if the long-term goal is daily wear, comfort, social acceptance and natural reuse must be re-validated by the target consumers — geeks hand-calibrating and charging constantly is not consumer evidence.

Three products, no shared order — only a shared method: let the market best suited to the job kill the biggest unknown, and watch how it distorts the product.

The next stop should be the market that produces more critical evidence. Before leaving, state four things: the lethal unknown, who can expose it best, how those users will skew the product, and what evidence means stop.

Market size defines the distant imagination; the next critical piece of evidence decides where you go now.