Blog Profitable Founder
Guide

Upwork Automation: How Lancer Hit $10K MRR With Zero Ads

Ivan Nedelkovski built Lancer, an Upwork automation agent, to $10K MRR in about 3 months with zero ads. Two affiliates drove the growth. Full playbook.

Ivan Nedelkovski got his SaaS to $10,000 a month about three months after launch.

He spent $0 on ads, and he had no audience to launch to (his X account is basically dormant).

The product is Upwork automation: an AI agent called Lancer that finds jobs on Upwork, qualifies them, and writes personalized proposals while you're doing literally anything else.

But the product is only half the story.

And most of the growth came from two people. Two affiliates, recruited by hand.

Ivan broke down the whole thing on Starter Story, and I watched it twice because the distribution playbook is that good. Here's the full breakdown, with the real numbers.

The Idea Was Hiding Inside His Agency

Ivan is a builder from Macedonia. For five years he ran MB Masters, a software development agency that did over seven figures and grew to almost 20 full-time employees.

The agency's main client acquisition channel was Upwork. He calls it a very underrated source of quality leads, and the numbers back him up: around 200,000 jobs get posted on the platform every month.

Here's the catch. Those jobs drop around the clock, and each one takes about 10 minutes to qualify and answer with a personalized proposal.

So you're either glued to the job feed at 3am, or you're losing deals to whoever replied first.

Ivan's take in the episode: "This seems like a perfect case where an agent would outperform a human in this task by like 10 times or even more."

So he built one. Over a weekend. For internal use only.

It crushed for the agency. That weekend hack is the whole origin story.

The Beta: 3 Five-Figure Clients in 2 Weeks

Before launching anything publicly, Ivan invited a few friends who also ran agencies to test the tool.

Between them, they closed three five-figure clients within two weeks of using it.

That was the signal. He spent the next three months turning the weekend hack into an MVP and launched it as a standalone product: Lancer.

The pricing is deliberately premium for a SaaS:

  • Pay as you go: $79 for 30 proposals, then $2 per extra proposal
  • Light: $300 for 250 proposals, then $1.50 per extra
  • Unlimited: $500 a month (the original launch offer)

Three or four months after launch, Lancer was at $10K MRR.

Which brings us to the part I actually want you to steal.

The Connector Strategy: Growth With Zero Ads

Ivan's ideal customer turned out to be agencies and freelancers sending a high volume of proposals on Upwork. They self-onboard easily and they barely churn, because the tool pays for itself with one closed deal.

The obvious playbook would be to chase those people directly. Cold email at scale, paid ads, hundreds of demo calls for a $200-a-month product.

Ivan skipped all of it.

Instead, he looked one layer above his customers and asked: who already has access to hundreds of them, plus their trust?

He calls these people connectors. You probably know the mechanic as affiliate marketing for SaaS, but the targeting is what makes it different. Instead of recruiting random people with newsletters, he went after the exact humans his buyers already pay for advice.

For Lancer, that meant Upwork coaches.

Think about what an Upwork coach's business looks like:

  • They sign between 5 and 20 new paying clients every month
  • Those clients pay them $600 to $1,000+ to get better at winning Upwork jobs
  • They have strong Upwork profiles themselves, with constant inbound

Every single one of those coaching clients is a perfect Lancer customer. So instead of hundreds of sales calls, convincing one coach becomes a single high-ticket deal that keeps delivering customers every month.

Two Affiliates Drove Almost All the Growth

This is the stat that got me. Most of Lancer's growth came from just two Upwork coaches.

Coach number one came through the beta. One of Ivan's test users had worked with the coach before, so he made the intro. Ivan demoed the product, the coach was blown away, and he's referred every client since.

Coach number two took more nerve. Ivan found him on LinkedIn, cold messaged him, and straight up offered him $1,000 just to jump on a call. He paid the guy to become a user.

(Would you spend $1,000 to open a channel that sends you customers every month? Most founders wouldn't, which is exactly why it worked.)

The commission structure is simple:

  • 30% lifetime commission if the coach closes the client and onboards them onto Lancer
  • 20% lifetime if they just make the referral

Quick napkin math from me: one coach who moves 10 clients onto the $300 plan generates $3,000 in MRR. The coach collects $900 of that every month, for as long as those clients stay. Because the commission runs for the lifetime of the customer, the coach has a reason to keep referring month after month.

Ivan's 5-Step Connector Playbook

Pat asked Ivan what he'd do if he had to start over with a brand new SaaS. His answer is a playbook you can run this week.

Step 1: Define your ICP from real paying users

You can't guess your ideal customer before people pay you. Look for the ones who onboard with almost no friction and stick around. For Lancer, that was agencies doing high proposal volume, not solo freelancers.

Step 2: Find the layer above them

Ask one question: who has a wide network of my ideal customers AND their trust? Coaches, consultants, agency owners, community operators. People your buyers already pay.

Step 3: Pitch them like a high-ticket deal

This is not a mass email. Research each connector, reference their work, record a long personalized Loom if you have to. Closing one affiliate is worth more than closing one subscriber, so you can afford the effort.

Step 4: Work out terms that respect their network

Ivan says 20% to 30% lifetime commission is the standard range. Adjust for the size of their network, their social following, and any existing deals with competitors. Sometimes that means cash up front, like the $1,000 call.

Step 5: Track everything and pay monthly

Once a connector is live, you get inbound Stripe sales and booked calls without touching anything. Lancer runs payouts through Tolt, an affiliate marketing tool built for SaaS, so commissions and monthly payments are automated.

The Stack Behind a $10K/Month AI Agent

Nothing exotic here, which is kind of the point.

  • TypeScript on front end and back end, with Next.js and Node.js
  • Cursor with Opus 4.5 for writing most of the code ("we barely touch any actual programming language")
  • OpenRouter for the LLM APIs
  • Hetzner and GCP for hosting
  • Elasticsearch for querying the 200K monthly jobs
  • Proxy providers for connecting safely to Upwork accounts
  • Tolt for affiliate tracking and payouts

The LLMs do exactly two jobs inside Lancer: qualify whether a job is worth bidding on, and write the proposal.

A weekend prototype, then three months of MVP work on the same stack his agency already used. The tech was never the hard part. Getting it in front of buyers was, and the connector strategy solved that.

The Bigger Opportunity: Agents on Top of Platforms

Pat's closing reflection in the episode is worth sitting with. Lancer is one example of a pattern: AI agents that automate tedious, repetitive work that lives on top of an existing platform.

Upwork today. But the same shape exists on Fiverr, LinkedIn, Poshmark, Pinterest. Starter Story even interviewed someone doing it on Photoshop.

The formula Ivan describes: find a platform with a huge user base, find the workflow everyone hates, and build the agent that does it 10x faster. Your ICP is already public, because they live on the platform.

I covered a different flavor of this in the solo AI agent business story, where Nick gets to $5K a month automating unglamorous back-office work. Same lesson both times: the money sits in tedious workflows that already have a crowd around them.

And it kept working after the cameras stopped. Indie Hackers later reported that Ivan sold his agency to go all in, with Lancer passing $20K MRR.

His advice for anyone still waiting: he delayed building products for years to optimize for income and security, and he now thinks that was probably the wrong call. The skills you build shipping a product compound harder than a salary, even a six-figure one.

FAQ

What is Upwork automation?

Upwork automation means using software or AI agents to handle the repetitive parts of winning work on Upwork: monitoring the ~200,000 jobs posted each month, qualifying which ones fit you, and drafting personalized proposals. Lancer, the AI agent in this story, does all three and charges between $79 pay-as-you-go and $500 a month unlimited.

How did Lancer get its first customers without ads?

Through connectors. Ivan recruited two Upwork coaches, each with 5 to 20 new paying clients a month, as affiliates on 20% to 30% lifetime commissions. One came from a beta user's intro. The other came from a cold LinkedIn message and a $1,000 offer just to take a call. Those two coaches drove most of the growth to $10K MRR.

What commission should a SaaS pay affiliates?

Ivan's standard range is 20% to 30% lifetime: 30% when the affiliate closes and onboards the customer, 20% for a plain referral. For connectors with unusually large or engaged networks, he suggests negotiating case by case, including upfront payments on top of the recurring commission.

Do you need an audience to grow a SaaS?

Lancer says no. Ivan had no following and ran zero paid ads. He borrowed trust from people who already had his customers' attention. Building an audience can take years; identifying the five coaches or consultants your buyers already listen to can take an afternoon.

Steal Playbooks Like This Every Week

I interview bootstrapped founders doing $100K to $10M a year on the Profitable Founder Podcast, and I dig for exactly this kind of detail: the two affiliates, the $1,000 cold offer, the 30% lifetime commission.

If Ivan's story saved you a few months of guessing, the podcast will too.

Watch the latest episode

Florian Darroman, founder of Distribb and host of Profitable Founder
About the author

Florian Darroman

Florian Darroman is a French distribution guy based in Bali, founder of Distribb and host of Profitable Founder. He interviews bootstrapped founders making $100K-$10M/year and documents the journey of growing Distribb to $100K MRR.

Experience: affiliate SEO to 6 figures, infoproducts to 7 figures, and built and sold Les Makers for $130K.

Read more in Guide

Keep reading

Building a SaaS toward $100K MRR?

Profitable Founder Club is a mastermind for founders doing $5K–$50K MRR. Bi-weekly calls, monthly Q&As with founders past $100K MRR.

Join the Club