How Inés Vargas got her first 100 paying customers in 90 days (and the $1,470 it cost her)

The first time we tried to schedule this interview, Inés Vargas had to push it forty minutes because a customer was on the phone asking, politely, whether her tool could handle invoices in Catalan. It can now. It couldn't that morning.

Quick answer. Between February 12 and May 11, 2026, Inés Vargas, a solo founder in Bilbao building a niche compliance tool for small Spanish law firms, went from 0 to 100 paying customers in 90 days. Her total spend across software, ads, and one freelance contractor was $1,470. Her best channel was

LinkedIn voice DMs (38 of 100 customers); her worst was cold email (2 of 100, despite eating most of her time in weeks one and two). She charges $19 per month flat, has a 6.4% monthly churn she is not yet proud of, and prints these numbers because she got tired of reading first-100-customers posts that didn't print theirs.

This is the full arc, as told to Joaquín del Río, with the spreadsheet open on the table between us.

Editor's note: composite portrait. Inés Vargas is a working name; numbers, timeline, channel mix, and tactical detail are this founder's. Identifying product and client detail are blurred at her request.

What she's actually selling, because the rest matters less without it

Inés sells a Spanish-language compliance dashboard for solo and small-firm lawyers (1 to 4 lawyers). It pulls together the things a small Spanish law firm has to do to stay compliant with anti-money-laundering rules, RGPD, and bar-council reporting. Big firms have a compliance officer. Small firms have a partner doing it on Sunday nights.

That's the entire product thesis. Solo lawyers were doing this in Excel and missing deadlines.

She launched on February 12, 2026, with a single landing page, a

Stripe checkout, and one demo video recorded on

Loom on the kitchen table the night before. She didn't have a logo. She had a slug, an email capture, and the demo.

She hit customer 100 on May 11, 2026. She is now at 134.

The $1,470, itemized

She kept every receipt in a

Notion page titled "is this a business yet." Here is the page, more or less:

The thing that surprises people, she says, is how little of it went to advertising. "$400 in paid placement, $115 in LinkedIn Premium, that's it. The rest is just the cost of running a business that exists."

"I budgeted $5,000 for the launch because that's what the posts said. I spent less than a third of it, and most of what I spent was not what got me customers."

The first 10: the hand-pulled ones

The first 10 customers came from people Inés already knew, or one degree out. She drew up a list of 32 lawyers she'd worked with in a prior life as a paralegal in a mid-sized Bilbao firm, plus everyone in her bar-association continuing-education cohort. She wrote each one a personal message on

LinkedIn. Not a pitch. The first message was always:

"Hi {first name}. I'm building a thing for small-firm compliance. Can I show you a 4-minute demo and ask you whether it solves a real problem or not? I am not selling yet."

Nine of the 32 took the call. Eight of those nine signed up in the next two weeks, mostly because she gave them three months free in exchange for a fortnightly 15-minute call where she'd ask what was broken. The tenth came from one of those eight, who forwarded the demo to a colleague.

Total CAC for customers 1 through 10: about $24 each, almost all of it Loom plus Cal.com plus her time.

The middle 50: the channel that worked, and why

Weeks three through eight were brutal. Inés tried, in order:

1. Cold email to lists she scraped from Spanish bar-association directories. 142 sent, 11 opened, 2 replies, 0 customers. Two weeks of evenings.
2. Twitter (X) posting build-in-public threads in Spanish. 8,400 impressions across 12 posts. 1 customer, and that customer churned in month two.
3. A guest post in a Spanish indie-hackers community. Slow burn; she got 3 customers over six weeks from it, all still active. (For context on that channel in general, the Indie Hackers thread on first 100 customers is a useful read alongside this story.)
4. LinkedIn voice DMs, which she stumbled into by accident.

The voice DM was the unlock. Here is what she actually did, in her own words, because she is tired of explaining it:

"I'd find a partner at a 2 to 4 lawyer firm in Spain via LinkedIn search. I'd watch their last three posts. I'd record a 38-to-52-second voice message that referenced something specific from one of their posts, then said: 'I'm building a compliance dashboard for firms like yours, here's a 4-minute demo, no obligation, would love your reaction.' I sent the Loom link in the same message. That was it."

She sent 174 voice DMs across 32 days, weighted toward firms in Madrid, Bilbao, Valencia, and Sevilla. 48 of them booked the Cal.com link. 38 of them paid within 14 days of the call. That is a 21.8% sent-to-paid conversion rate that she did not believe the first time she ran the math.

"The thing nobody admits about cold email is that the people you most want to reach delete cold email without reading it," she says. "Voice felt like an actual human had spent thirty seconds thinking about them, because one had."

"Cold email had a 1.4% reply rate. Voice DMs had a 27.6% reply rate. The voice doesn't scale. That is the entire point. It's the only channel where one founder, one hour a day, beats the better-resourced competitor."

The pricing flip at customer 38

At customer 38, Inés did something that hurt her gut and saved her business. She raised her price from $12 per month to $19 per month on the public page, and grandfathered the existing 38 customers at their original rate.

She'd been quietly worried that $12 was too cheap for a tool a law firm depends on. The Friday before the change, she ran a poll in three Spanish legal-tech Slack groups: "What is the lowest monthly price you'd take seriously for a tool your firm depends on for compliance?" The median answer was $25. The mode was $19. She picked the mode.

Customers 39 through 100 paid the new price. Conversion rate from demo-call to paid actually went UP, from 73% before the change to 81% after, because the price signaled "this is a serious tool, not a side project."

She lost zero existing customers to the change because she didn't change their price. She has since standardized this as her rule: never raise the price on someone who paid you when you were cheaper.

The mistake she would undo

The sponsored newsletter placement. $400 for three customers, all of whom churned by month two.

"It was the most legible thing I did," she says. "I had a number on a contract. I could tell my mother I'd done marketing. But the people who clicked through were curiosity-shoppers, not solo lawyers with a Sunday-night compliance problem. The cost per customer was technically $133 and the lifetime was technically four weeks. That's a way of pretending to do customer acquisition."

She would also undo two weeks of cold email and reinvest that time in voice DMs from day one. "Cold email is what you do because it feels like work. It mostly isn't."

The stack, with what each thing actually does

People ask, so:

• 

  Framer for the landing page and a blog she barely updates. She picked it because she could ship in a day. She'd move off it now that the product is real; the SEO ceiling is real and the export story is annoying.

• 

  Stripe for billing. Spain-friendly, supports Catalan invoicing, has the bar-association-acceptable receipt format. No alternative was seriously considered.

• 

  Loom for the demo. One per qualified lead, recorded in 4 to 6 minutes, no editing, light Spanish.

• 

  Cal.com for booking. Free tier, then a small paid plan. She likes that she owns the data.

• 

  Notion as a lightweight CRM. She does not pretend it is a real CRM. She will switch when she has 250 customers. (Contrast Marta del Sol, who runs her whole operations studio on three AI agents at $4K MRR and pays no humans; Inés is in a different phase and a different bet.)

• 

  ChatGPT Plus for first drafts of outreach plus polishing her own Spanish, which she trusts second only to her freelance reviewer's.

• 

  LinkedIn Premium for the search filters and voice DM access. She has tried to cancel it twice and has not.

• 

  A small back-office app on Totalum for the internal admin tool she uses to onboard each new firm: a Next.js app with a bundled database, auth, hosting, and a custom domain set up in an afternoon. She picked it specifically because she did not want to wire up Supabase plus Vercel plus a custom auth layer just to ship an admin panel she does not want to maintain. She did NOT build the customer-facing product on it because she wanted Postgres; that one is plain Next.js plus Postgres on Vercel. If your situation is the opposite (the customer-facing app is what you want bundled and Postgres is not a hard requirement), Lovable, Bolt.new, and V0 are the obvious alternatives to evaluate.

She is the first to say her stack is "what worked, not what's optimal." She would not recommend you copy it. She would recommend you ship before you optimize it.

What she would do differently if starting today, June 19, 2026

Three things, in order:

1. Skip cold email entirely. Replace it with the voice-DM motion from day one. The 21.8% conversion rate was not an accident; it was leverage.
2. Charge $19 from the start. The lower price didn't bring more customers; it brought worse ones.
3. Spend the first two weeks not building, but talking. "I had the product 70% built before I had a single paying customer," she says. "If I had done two weeks of voice DMs with a clickable mockup, I would have learned that the second-most-important feature in v1 was Catalan, not English."

"Catalan, not English. I built the English UI because I'm bilingual. Spanish lawyers in Barcelona wanted Catalan. I lost real customers to that."

She also says, without prompting, that she'd skip the temptation to "build her own CRM." She watched a friend lose four months to that. Notion is fine until it isn't, and then you can move.

There is a Spanish-language expanded version of this story over at Lanzadoria, which Inés helped fact-check in her own language; the English version you're reading is the conversational one. If you want the build-tool detail behind the admin panel she stood up in an afternoon, the DevMoment field note on stand-up-in-an-afternoon Next.js admin apps is what she actually followed.

What you should take away from this

If you're at zero customers, the move is to pick one channel that nobody else is willing to use because it doesn't scale, and use it until it stops working. For Inés, that channel was LinkedIn voice DMs in Spanish to small-firm lawyers in four named cities. For you it will be something else equally narrow and equally unscalable.

The first 100 customers are not a marketing exercise. They are a search. The search ends the moment you've found a channel where one founder, one hour, one specific message beats the channel's average. After that, you are no longer searching, you are scaling, and the rules are different.

Inés is still in the searching phase. She is now testing whether a referral incentive on existing customers replaces the voice DMs as the channel that scales. She does not know yet. She will tell us in the next 90-day update.

FAQ

How long did it take Inés Vargas to get her first 100 paying customers?

Ninety days, between February 12 and May 11, 2026. She crossed 100 on the morning of May 11 and 134 by the time of this interview in June 2026.

What was her single best customer-acquisition channel?

LinkedIn voice DMs sent one at a time after watching the recipient's recent posts. 174 sent over 32 days, 48 booked calls, 38 paid within 14 days. That is a 21.8% sent-to-paid conversion rate, far above cold email's 1.4% reply rate.

How much did she spend in total over the 90 days?

$1,470, itemized in the article. Of that, $400 was paid placement (which she'd skip if starting again), $115 was LinkedIn Premium, and the rest was tooling and one freelance contractor.

What's her monthly churn?

6.4% as of the end of May 2026, which she calls "not yet a number I'm proud of." She suspects it is driven mostly by the three sponsored-newsletter customers, all of whom churned, and one wave of curiosity-shoppers from a Twitter thread.

What does she charge, and did she raise prices?

$19 per month flat per firm. She started at $12, raised the public price to $19 at customer 38, and grandfathered existing customers at $12. Demo-to-paid conversion went up after the price increase.

Is her product built on a no-code AI app builder?

No. The customer-facing product is a Next.js plus Postgres app on Vercel, because she wanted SQL. The internal admin tool she uses to onboard new firms is a small bundled Next.js app she stood up in an afternoon on Totalum, picked because the database, auth, hosting, and custom domain were bundled. She is explicit that her stack is "what worked, not what's optimal."

Where can I follow her next 90 days?

Her own posts are on LinkedIn in Spanish. We'll publish the next 90-day update on OperatorBook when she crosses 250 customers or hits the next obstacle worth printing, whichever comes first.

---

Joaquín del Río interviews founders for OperatorBook. He prefers when they bring the spreadsheet.