The SaaS Distribution Course #22: How Supademo Stacked Free Work, Search, and Sharing to Reach $1M ARR

TL;DR

Supademo did not find one magic growth channel. Co-founder Joseph Lee says the company moved from roughly $100,000 to $1 million in annual recurring revenue during 2024 with a team that had grown to ten people by the time of his May 2025 interview. The distribution system began with work that could not scale: he built complete interactive demos for strangers on Reddit and posted the finished work in public. That reportedly produced a couple thousand signups.

The next stages reused the same product utility at larger scale. Supademo embedded demos into search pages, ungated existing product capabilities as free tools, and let customer-created demos travel through websites, documentation, onboarding, and sales material. Lee reported that free-tool pages generated 50% or more of traffic and converted roughly 11–12% of visitors into signups. A later Supademo playbook says product sharing became more than one-third of traffic.

Those numbers are founder- or company-reported and unaudited. No reviewed source attributes ARR to a single channel. Copy the sequence: manual value → public proof → searchable utility → frictionless use → product sharing → measured channel expansion. Do not copy the revenue attribution that the evidence cannot supply.

What you will build

You will build a progressive-proof distribution system. Each stage produces the evidence, language, or exposure required by the next one. The system does not automate an unproven message. It converts manual customer work into reusable acquisition surfaces, then measures whether those surfaces attract the same people who can become retained customers.

Your outputs are:

• a narrow audience and painful demonstration job;
• a public value-first offer with a delivery template;
• an evidence ledger from the first 20 completed jobs;
• three search-page types that embed real product utility;
• one ungated free tool built from an existing product capability;
• a product-sharing loop with attribution and quality safeguards;
• a channel scorecard and a rule for adding or stopping the next channel.

Use this system if

• your product can create a useful artifact before the buyer commits;
• the finished artifact can be shown publicly without exposing private data;
• buyers already search for adjacent tasks, comparisons, or examples;
• normal product usage creates something that gets shared, embedded, or forwarded;
• you can personally deliver the first 20 examples and learn from every one.

Do not use it if

• the free work replaces a high-cost service with no path to product usage;
• search volume belongs to an audience that will never need the paid workflow;
• sharing reveals customer data or creates spam;
• you need a new product and support operation just to maintain the free tool;
• you cannot define activation, qualified signup, retention, and channel cost.

Verified case snapshot

Stage	Evidence	Limit
Company growth	About $100K to $1M ARR in 2024, with double-digit monthly growth	Founder-reported and unaudited
Manual acquisition	Free demos posted publicly reportedly produced a couple thousand signups	No analytics export or exact window
Product-led search	Hundreds of embedded demos and programmatic pages around adjacent tasks	Exact page and revenue contribution undisclosed
Free tools	50%+ of traffic and roughly 11–12% visitor-to-signup conversion	Founder-reported; attribution and cohort definitions absent
Product sharing	Later company playbook says more than one-third of traffic	Later period and company-reported
Independent signal	G2 ranked Supademo number five among its 2025 fastest-growing products	Ranking validates market momentum, not ARR
Stopped channel	Outbound “completely flopped” and the lean team stopped forcing it	No volume, duration, spend, or response rate

The model: turn one useful job into several distribution surfaces

Supademo's product makes software understandable. The same capability could therefore do several jobs. A custom demo helped one founder. The public reply helped every person reading the thread. An embedded tutorial answered a search query. An ungated editor delivered value before signup. A customer-shared demo introduced the product to a new viewer while completing the customer's own work.

This is not simple repurposing. Each surface lowers a different cost. Manual work lowers the buyer's effort. Public proof lowers uncertainty. Search lowers discovery cost. An ungated tool lowers time-to-value. Product sharing lowers the effort required from the existing user to create the next exposure. Measurement lowers the cost of deciding what to build next.

The system breaks when a stage does not feed the next one. Free work without captured language becomes consulting. SEO without embedded utility becomes generic content. A free tool for a broad audience creates vanity traffic. Sharing without attribution hides whether viewers are qualified. More channels without stop rules consume the small team's attention.

Step 1: choose a job you can complete before asking

Supademo's first ICP was the busy founder. These people needed to show what their software did but often lacked the time to create a clear demo. Lee's Reddit offer removed almost all effort: post a website, and he would sign up, learn the workflow, build the interactive demo, and return an editable result. Buying was optional.

Write a value-first offer with five fields: For [narrow person], I will produce [finished artifact] from [small input] within [time]. You can use it without buying. I may show the result publicly only with permission. The artifact must stand alone. A “free audit” that ends with vague recommendations is usually a disguised sales call.

Deliver 20 manually. Record the input, time spent, pain language, existing workaround, visible reaction, product action, signup, activation, and seven-day return. Pass condition: at least 12 people accept delivery, eight use or share the result, and five complete your activation event. Stop condition: recipients like the work but never touch the product. These are planning thresholds, not Supademo results.

Step 2: publish the proof where the audience already watches

The Reddit tactic worked because the completed demos appeared in the thread. One delivery served one founder, but also showed every reader what the product could create. The product was the means of delivering visible value rather than the subject of a launch announcement.

Choose one community where people already post the raw input or ask for the job. Follow its rules. Ask permission before exposing a result. Return the artifact in the original conversation and add a short note describing what you changed. Do not gate the result, automate comments, or fabricate urgency.

Measure completed jobs, public views, profile visits, qualified signups, activation, and seven-day retention. Separate the person who requested the work from observers who discovered it. Pass condition: each completed job creates at least one additional qualified product session and the requester still receives enough value to recommend the process. Stop condition: visibility rises while activation quality falls.

Step 3: convert repeated questions into product-led search pages

Supademo studied Zapier's programmatic search model, but it did not only publish pages about “interactive demos.” It built tutorials around adjacent software tasks, product-demo searches, and comparisons where the intended audience already had demand. The page contained an interactive Supademo, so the visitor experienced the product while getting the answer.

Create three page queues from the manual evidence:

1. Task pages: how the buyer completes a narrow job in a familiar product.
2. Example pages: an ungated, self-serve demonstration of a product or workflow.
3. Comparison pages: honest alternatives for buyers already evaluating solutions.

Every page must satisfy the query without signup, demonstrate one native capability, and offer one relevant next action. Track ranking separately from useful behavior. Pass condition: after an appropriate search-learning period, qualified visitors interact with the embedded product and activate at a rate worth the ongoing content cost. Stop condition: pages attract large traffic from people outside the paid ICP.

Step 4: ungate an existing capability, not a separate business

Supademo made its product experience accessible before signup, then exposed existing modules as free tools. A screenshot editor and hotspot capability could answer adjacent searches without a new product architecture. Lee said these pages produced at least half of traffic and roughly 11–12% of visitors became signups.

Inventory your product's smallest independently useful actions. Score each on ICP overlap, immediate output, compute cost, privacy risk, support burden, and bridge to the paid job. Choose one with high audience overlap and low incremental maintenance. Put the useful action first. Ask for signup only when the visitor needs to save, repeat, collaborate, or continue into the core workflow.

Pass condition: the tool reaches a useful output in under two minutes, at least 8% of qualified visitors continue to signup, and activated users retain no worse than another organic cohort. Stop condition: usage is expensive, anonymous abuse grows, or signups do not reach the core job. The 8% threshold is an operator gate, not a universal benchmark.

Step 5: make successful usage create the next exposure

A Supademo leaves the creator's account. It can be linked, embedded in a website, added to support documentation, or used in onboarding and sales. The recipient experiences the artifact and can discover the product through normal use. Supademo's later company playbook says this usage-based referral loop generated more than one-third of traffic.

Map where your product crosses an account boundary: share, embed, invite, publish, export, forward, present, or collaborate. Add attribution only where it does not damage the customer's output. Preserve private and white-label options. The user should never have to advertise you as a separate task.

Pass condition: a meaningful share event occurs after activation, viewers receive value before a prompt, and viewer-to-activation quality can be compared with other channels. Stop condition: the loop depends on an intrusive watermark, public data leakage, or invitations sent without clear user intent.

Step 6: add one channel only after defining the stop rule

Lee described moving from unscalable work toward one predictable channel, then adding another in a later month. Supademo eventually reported three to five channels working consistently. The exact mix and economics were not disclosed, but the sequencing rule is clear: do not run ten ambiguous experiments at once.

Give each experiment an owner, audience, hypothesis, leading signal, customer-quality metric, maximum time, maximum spend, and binary decision. Keep learning time appropriate to the channel. Community delivery can show signal in days. Search can require months. A product loop requires enough activated users to create shares.

Supademo's outbound did not meet the team's bar. Lee said inbox noise was high and strong outbound required focused work across data, messaging, deliverability, timing, and people. The team preferred to stop rather than execute poorly. Pass condition: the new channel produces retained customers at an acceptable fully loaded cost. Stop condition: it consumes the owner required to keep the proven channel healthy.

What failed and what the success story hides

“Build a great product and people will come” failed first. Supademo spent time on product detail before predictable distribution existed. Broad founder adoption also had an economic limit: small organizations were price-sensitive, used the tool less habitually, and offered less expansion. Supademo moved its strongest focus toward organizations above roughly 50 employees while preserving self-serve access.

The numbers can create false certainty. A traffic share is not revenue attribution. A signup is not activation. G2's official 2025 ranking independently confirms that Supademo had strong market momentum, but the ranking is based on review and market-presence criteria. It does not audit ARR or prove which channel caused growth.

Search pages require editorial maintenance. Free tools create compute, support, and abuse costs. Public community work can become spam. Product sharing can conflict with privacy and white-label needs. The system works only when customer quality and contribution economics survive after the vanity metrics rise.

Your 30-day implementation plan

1. Days 1–3: choose one buyer, one painful job, and one activation event.
2. Days 4–10: deliver ten finished artifacts manually and log every outcome.
3. Days 11–14: deliver ten more publicly, with permission, in one community.
4. Days 15–17: extract the repeated tasks, words, objections, and alternatives.
5. Days 18–21: publish one task page, one example page, and one comparison page.
6. Days 22–24: ungate one existing product capability with a clear paid bridge.
7. Days 25–27: instrument share events, viewer sessions, activation, and retention.
8. Days 28–30: compare cohorts, keep one channel, repair one, and stop the rest.

The day-30 output is not 100 blog posts. It is 20 completed jobs, an evidence ledger, three useful search pages, one small free tool, one measured sharing event, and one decision about the next channel.

Implement it in Lead Scorer

Lead Scorer can run the research, qualification, drafting, and learning loop around this system. It cannot build the free tool, promise customer results, post in communities automatically, or send outreach without human review.

Start with icp-offer-context. Store the narrow buyer, painful job, required raw input, finished artifact, product activation event, disqualifiers, privacy constraints, approved proof, and maximum manual delivery time. Keep requesters, public observers, search visitors, and product viewers in separate lists because their intent and evidence are different.

Use signal-research-dossier for people who publicly describe the problem. Require two dated signals, such as a current role and a recent post, launch, help request, or product update. Use icp-scoring-rubric before paid enrichment. Score problem evidence 30 points, role fit 20, visible raw input 15, product-workflow overlap 15, timing 10, and permission-safe delivery 10. Exclude below 70/100.

Find people in [community or public source] who currently own [job]. Require a dated problem signal and a current-role source. Separate people asking for help from observers. Skip anyone whose use case requires private data or whose company cannot use the paid workflow.

Enrich only qualified people with lead-enrichment-pipeline and stop at every credit gate. Use ai-authored-campaign only when a private follow-up is appropriate. The message should deliver or ask permission to deliver one useful artifact. Run outreach-qa-audit and require 85/100 before the founder reviews each draft. Nothing is activated or sent by the agent.

Draft one message from verified evidence. Offer to create [artifact] from [small input] within [time]. Make purchase optional. Do not infer urgency, private data, results, or consent. Keep the draft queued for human review.

Use Content Studio for the evidence ledger: original problem, delivered artifact, reaction, objection, activation, and follow-up result. Turn repeated questions into useful pages. With reply-triage, separate interest, timing, wrong person, objection, and rejection. When a public artifact earns visible engagement, capture those people as a separate signal audience and score them before any contact discovery.

The operating view should connect signal → qualified person → approved artifact → public proof → search or viewer session → activation → retained account. The pass condition is not total traffic. It is evidence that each new surface preserves or improves customer quality.

Operator checklist

• The free output is genuinely useful without a purchase.
• The first 20 deliveries are manual and logged.
• Public proof is shared only with permission.
• Search pages embed product utility and fully answer the query.
• The free tool reuses an existing capability and attracts the paid ICP.
• Signup, activation, retention, and revenue are measured separately.
• Product sharing happens during normal work and respects privacy.
• Every channel has an owner, time box, cost ceiling, and stop rule.
• One proven channel stays healthy before another is added.
• All agent-written messages remain queued for human approval.

Sources and evidence limits

The main source is Joseph Lee's May 2025 SaaS Podcast interview and full transcript, audited across all 49,318 effective transcript characters. The $100K to $1M ARR movement, double-digit monthly growth, team, user and company counts, Reddit signups, free-tool traffic, and signup conversion are founder-reported and unaudited.

Supademo's product-led SEO playbook documents comparison pages, embedded tutorials, free tools, expected effort, and a reported 20%+ visitor-to-signup rate on comparison pages. Its later growth playbook describes the sequence from search to founder-led community work to product sharing. Both are company-authored sources and should not be treated as independent audits.

An original 2023 Reddit post preserves the early activity, including outreach to more than 200 startups. A later Practical Founders interview corroborates the continued product-led and enterprise motion from another publisher, but its revenue data still comes from the founder.

No reviewed source discloses channel CAC, payback, gross margin, activation definition, paid conversion from free tools, retention by channel, programmatic-content cost, or revenue attribution. Search rankings, Reddit rules, and AI discovery continue to change. Use the worksheet and decision gates to test the mechanism with your own economics.

Frequently asked questions

Did free tools alone create $1M ARR?

No. They were a major traffic source inside a broader system. The sources do not isolate their revenue contribution.

Is this a content strategy or a product-led strategy?

Both. The page captures demand, but the embedded or ungated product action delivers the answer. Distribution improves because the content and product share the same job.

What should a two-person SaaS copy first?

Deliver ten finished artifacts to one narrow audience. Do not start with programmatic SEO. The manual work supplies the language and use cases that make later pages useful.

When should you add outbound?

Only when the buyer is identifiable, the proof is strong, and someone owns data quality, deliverability, messaging, follow-up, and measurement. Supademo's experience is a warning against treating outbound as a tool purchase rather than an operating system.