GojiberryAI reached $424K MRR in about 13 months, led by cold outreach and Reddit

Week 437 min read

TL;DR

GojiberryAI, an AI tool that finds B2B leads from LinkedIn buying signals, showed $424,368 MRR at TrustMRR's last sync on 14 Aug 2026. This story estimates gross margin on its $99 plan at about 31% to 75% at light use, and covers how the founder grew it.

On this page
MRR at last sync
$424,368
revenue, last 30 days
$407,199
active subscriptions
3,944
all-time revenue
$1.8M
Source: TrustMRR, Stripe-verified, last synced 14 Aug 2026
At a glance
Founded
July 2025
Founders
Pierre-Eliott Lallemant (CEO), Romàn Czerny (CMO), Dylan Teixeira (CTO)
Funding
$500K seed, Y Combinator Spring 2026 batch
Pricing
Pro $99/mo, custom plans for teams of 3 or more

What it is

GojiberryAI sells an AI sales agent for B2B teams. It watches LinkedIn and the web for buying signals (a prospect changing jobs, following a competitor, announcing funding), scores each person against the customer's ideal buyer profile, looks up contact details and sends personalized email and social messages.

TrustMRR's last Stripe sync, on 14 Aug 2026, showed $424,368 in MRR (monthly recurring revenue, the subscription income billed each month), about $5.1M annualized. Sacra estimated rival Clay at $150M ARR in May 2026, so Gojiberry holds a small share of its market.

How fast it grew

  1. Company founded, per TrustMRR.
  2. An AI meeting-notes tool collects 2,000 waitlist signups and no paying customers. The founders rebuild around lead generation, first as a manual service that reaches $7K MRR before any software exists (Czerny, via Profitable Founder).
  3. $33,000 in revenue within 4 months and $50K MRR after 6 months (founder-reported).
  4. The YC launch post reports $112K MRR, 1,000+ paying customers and 44% monthly growth after 9 months.
  5. Czerny says MRR has reached $300K (his reported figure).
  6. Last TrustMRR sync: $424,368 MRR. TrustMRR's summary lists 12% month-over-month growth.

That is about 9 months to roughly $112K MRR by the company's account, and about 13 months to $424K MRR on TrustMRR's last sync.

How it makes money

The Pro plan costs $99 a month: 2 AI agents, up to 1,800 prospects contacted per month, contact lookups across 15+ data providers, a shared inbox, and integrations with HubSpot, Pipedrive, Attio and Claude through MCP, a standard for connecting AI tools to other software. Teams of 3 or more get a custom plan.

The last MRR divided by the subscription count is about $108 per subscription (derived), which fits some customers paying for custom plans.

Margin per subscription

No gross margin is published. YC's company profile describes the company as profitable, per WebTribunal. The table is an estimate for one $99 subscription over a month, dated 19 Oct 2026, built from public prices for the main variable costs. Gojiberry's actual costs are private.

Per subscription, per month Light use (450 prospects) At the 1,800 cap
Revenue (list price) $99 $99
Contact data, one paid lookup per prospect at $0.03 to $0.10 $14 to $45 $54 to $180
AI model calls $3 to $10 $8 to $25
Sending, hosting, support share $5 to $10 $8 to $15
Card fees (assumes 2.9% plus $0.30) $3 $3
Gross profit $31 to $74 -$124 to $26
Gross margin 31% to 75% -125% to 26%

Contact data sets the margin. Published prices run from about $0.01 per verified email at Prospeo and $0.03 at BetterEnrich to $0.07 to $0.37 per verified contact in one developer's waterfall test. LinkedIn-only messages need no email lookup, so one paid lookup per prospect is the conservative case.

Acquisition cost eats the first months. Czerny posted on X that $2,000 a day of Meta ads produced 240 paying customers in 30 days, about $250 per paying customer (derived, paid channel only). At $31 to $74 of monthly gross profit, that pays back in 3.4 to 8.1 months at light use. His lifetime affiliate program pays 30%, or $29.70 a month on a $99 plan, which cuts light-use gross profit to about $1 to $45 on affiliate-sourced customers.

Retention is undisclosed. Czerny told Profitable Founder that free blueprints widened reach and also brought smaller teams and more churn.

How the founders found customers

Per a founder interview summarized by WebTribunal, the founders sold the result before the product existed: a PowerPoint pitch, prospect lists built by hand, and software written after customers paid. Czerny describes an internal test in which outreach to people showing buying signals converted 4 times better than outreach to random Apollo prospects. No sample size or method is published.

Czerny's own channel order: outbound up to $6K MRR, mostly Reddit from $10K to $25K, content and free blueprints to $75K, then partnerships, X, B2B creators, sponsored newsletters and affiliates up to $150K. Other founder-reported tactics include 5 LinkedIn accounts running the product on its own prospects, a TrustMRR ad spot that brought 50 to 100 visitors a day, and 50+ paid LinkedIn creators for an average 1.5x revenue return. A first YC application at about $20K MRR was rejected, and a second at about $1M ARR was accepted.

Did the founder have an audience first?

No pre-launch follower count is public, and the evidence points to little reliance on one. Coco AI, the founders' earlier company, reached about $500K ARR through cold outreach, per Superframeworks, with no paid ads, viral content or Product Hunt launch. Gojiberry began as a manual service sold through about 20,000 cold emails.

The audience came with the product. Czerny reports 11 million Reddit impressions in 4 months. His X following is a later figure: TrustMRR's founder page shows 9,646 followers in an undated snapshot, a creator index lists 19K on 24 Aug 2026 with 2.2K added in 30 days, and TrustMRR's listing pages show 18.6K to 21.8K. That pace suggests most of the following was built while selling Gojiberry (our inference).

He did have a track record: Coco AI sold for seven figures in 2025, and TrustMRR lists four startups by him with $5.25M in verified revenue, including Gojiberry's $1.8M.

Competitors

The five rivals come from the companies WebTribunal places in each part of Gojiberry's workflow, limited to those with a published price or funding figure.

Rival Overlap Entry price Funding or scale Figure type
Clay Signals, data enrichment, email sequencer Launch $185/mo, Growth $495/mo (Mar 2026) $115M Series D at $7.1B (Sep 2026); ARR est. $150M (May 2026) Reported; Sacra estimate
Apollo Contact data, sequences, buying intent From $49/mo billed annually Not compared Third-party listing
Instantly Cold email sending From $37.60/mo billed annually Not compared Third-party listing
Artisan AI sales agent for outbound Free trial; paid price not published $25M Series A (2025) Company-announced
11x AI sales reps Not published $50M Series B (Sep 2024) TechCrunch, citing sources

Gojiberry's $99 plan sits above the Apollo and Instantly entry prices and below Clay's Launch tier, and it bundles parts those tools sell separately.

Assumed stack

Confirmed from public pages: Stripe billing (TrustMRR), a Framer marketing site (its page generator tag), the web app at app.gojiberry.ai, 15+ buying and social signals, and a waterfall of 15+ contact-data providers. A third-party directory of founder lessons lists Next.js, Supabase, Vercel, the OpenAI API and Tailwind CSS without linking a primary source, so treat that list as an assumption.

What an indie developer should know before building one

  • Prospect selection drives the result. The founders' 4x comparison varied the audience, not the message copy.
  • Usage caps protect margin. A full-cap subscription can lose money at higher data prices, so the prospect limit and paid top-ups matter as much as the sticker price.
  • LinkedIn restricts automation. Its help page prohibits unauthorized bots and scrapers, and TechCrunch reported on 7 Jan 2026 that LinkedIn temporarily restricted Artisan's accounts before reinstating them. Gmail also requires sender authentication and unsubscribe links for high-volume senders.
  • Enterprise buyers ask for certifications. A vendor listing says Gojiberry holds no SOC 2 or ISO 27001, so the realistic first customers are founders and small teams.
  • Distribution came first. The CMO runs growth and the CTO builds the product, so a solo developer needs a distribution plan before a build plan.

Product ideas

  • A niche version for one buyer type, such as recruiters or agencies selling to recently funded startups, where contact data is cheaper to source than for a general tool.
  • An account-health and deliverability monitor for AI outbound users, covering LinkedIn restrictions and Gmail sender requirements.

Best for

  • developers with a niche and a distribution plan who can pre-sell outbound results by hand before writing code, as the founders did.

Skip if

  • the plan is to undercut Apollo and Instantly on price, or to offer heavy usage on a flat $99 plan without a data cost cap.

Sources

  1. TrustMRR: GojiberryAI: revenue, subscriptions, sync date
  2. TrustMRR: Romàn's startups: follower snapshot, verified revenue
  3. GojiberryAI: pricing, features, integrations
  4. YC launch post: $112K MRR, customers, growth
  5. YC company profile: founders, Spring 2026 batch
  6. WebTribunal deep dive: founder interview, YC figures, competitor map, platform risks
  7. Profitable Founder: $0 to $50K MRR: pre-selling, outreach, Reddit, affiliates
  8. Profitable Founder: $50K to $300K MRR: channel order, churn, $300K claim
  9. Czerny on X: Meta ads: ad spend, paying customers
  10. Superframeworks case study: rebuild, early revenue, Coco AI method
  11. The Register: indie software to follow: X followers on 24 Aug 2026
  12. BetterEnrich review: per-email price
  13. Best waterfall contact providers: Prospeo price
  14. Waterfall enrichment test: cost per verified contact
  15. Sacra: Clay: ARR estimate, pricing
  16. SaaSRise: Clay Series D: round size
  17. The AI Agent Index: rival prices, certifications
  18. Founder-lessons directory: unconfirmed stack list

Continue reading