HeyMark

Social media management, rebuilt for the age of AI.

Plan what comes next, publish everywhere, and understand what works for your brand.

Driven VC, July 2026

Earning attention turned into an internal operation

Every company that wants to grow has to earn attention, and increasingly that happens through the content it publishes on social. Social media stopped being the place where you run occasional campaigns. It became an operating channel.

LinkedIn Engineering Blog

Posting on LinkedIn is up 41% in the past three years.

March 2026

20M

There are over 20 million videos uploaded daily to YouTube.

YouTube Official Blog, January 2026

Both views and time spent watching original Reels on Facebook approximately doubled in the second half of 2025.

Meta Newsroom, February 2026

Who runs that operation

The community manager is the person who coordinates all of that work. In smaller teams they also execute every one of those tasks themselves.

Community manager

Plan the month

Brief the creative

Get it approved

Publish

Respond

Report

HeyMark is where that operation runs

HeyMark

Community manager

Plan the month

Brief the creative

Get it approved

Publish

Respond

Report

One place where the work, the people and the decisions around every piece of content live together.

Our big ambition

To become the standard way brands manage social media in the age of AI.

And where we begin

Content management. The calendar, the coordination, the approvals and the performance of every piece.

This is what that starting point looks like today

One piece of content carries the visual, the approvals, the iterations, the reasoning behind each decision, and the numbers it produced. It is the densest object in the entire workflow. That is why we start there.

Product demo assets/demo.mp4

These platforms already exist. So why is the job still broken?

They are adding AI to a workflow built for a different era.

This is not a category we need to invent

Willingness to pay for social media management is already proven, from individual operators to the largest enterprises in the world.

Sprinklr $857M

Sprout Social $458M

Hootsuite ~$350M

Buffer $25.6M

Metricool €17M

$0B

annual revenue across these five companies

0K

paying customers at Buffer alone, self-serve

$0B

category spend today, growing ~20% a year

Sprinklr FY2026 and Sprout Social FY2025 from SEC filings. Buffer from its live public metrics dashboard. Metricool as disclosed by team.blue at acquisition. Hootsuite is a third-party estimate and is not audited. Category size and growth from Fortune Business Insights, $32.5B in 2025 to $164.5B in 2034, CAGR 19.7%.

The shift is already visible

Postiz is an open-source scheduler built in public by a single founder. It went from $18K to $152K of monthly revenue between January and June 2026. The inflection lines up with the moment it started selling to AI agents rather than only to people.

Six revenue points published on Postiz's public dashboard. No interpolated months.

Publishing infrastructure alone is not the whole opportunity. But it is clear evidence that customers adopt a new product in this category when it reflects how the work is changing.

A proven category is being rebuilt right now.

We narrowed the problem by building it

From February to June 2026 we built and sold an AI agent that tried to run a company's entire marketing operation: ideas, scripts, images, video, social publishing, paid media, email and comment automation.

$0K

of our own capital invested

0

sales meetings

0

on the waitlist

$0K

MRR, with 17 paying customers

1

Too broad to become exceptional

Every feature we shipped competed against a product dedicated to only that. Covering all of marketing left us good at everything and exceptional at nothing.

2

Fully autonomous content does not earn reach

Users arrived expecting to automate all of their content. It is possible, but the output is weak: it removes the human from the part where they add the most value, the creative one.

3

Too operational to scale as software

To keep the human in the creative loop we put our own people on ideation and editing. Scaling that meant becoming an AI-enabled agency, not a product.

Creation belongs to humans. What a tool can do exceptionally well is help them manage the work around it.

One job. Three operating environments.

The same community manager role runs in three different contexts. The scale and the coordination burden change. The core job does not.

In-house

~1 brand

Works inside a company, running its own brand. Early on this is often the founder or whoever leads marketing.

Freelance

~4 brands

Works independently, running social for several client brands at once.

In an agency

5 to 15 brands

Works inside an agency, handling multiple client accounts in parallel.

The operator is the community manager, but the person approving the spend is often their counterpart: the brand owner, the marketing lead or the agency. That is why we charge per brand, one workspace per brand, not per seat.

We have not yet decided which of these is our first narrow wedge. The pilot picks it with evidence, comparing pain, activation, time to value, retention and willingness to pay.

Start with one community manager. Earn the right to expand.

1

Now

Common core

2

Pilot

Pick the sharpest ICP

3

Depth

Build deeply for that user

4

Breadth

Expand across the workflow

5

Scale

Adjacent users and teams

>1.6M

estimated serviceable brands, Spanish-speaking markets plus US Latino

→

~2,575

paying brands in the 24-month planning case

=

~0.16%

of that serviceable base

Chile is where we run paid acquisition first, because feedback and CAC are cheap there. The product is bilingual and self-serve, so organic distribution starts internationally from launch. Chile is the laboratory, not the limit.

Bottom-up model built from official company counts, filtered by active social operation and ability to pay. The serviceable base is an estimate.

A simple starting model

Self-serve first. The billing unit is the brand.

Pro

$29/ mo

One workspace per brand. Creators and small businesses.

Business

$99/ mo

More capacity for teams and agencies with a heavier operation.

Extra networks

$3/ network

Expands each workspace without changing the per-brand economics.

Paid plans leave a 64% to 80% margin. Today the product runs a 7-day free trial with a card on file, behind a hard paywall, which is what the closed beta uses.

Still open, and deliberately so: the public launch mechanic. Freemium and reverse trial are live options once there is cash to build and operate a free tier. The expansion axis, whether it runs on brands, networks or seats, is the other open question.

Build for the daily user. Let the organization follow.

Designing for the community manager is also what makes the distribution work. One person can start with one brand and get value without an implementation project or an organizational rollout.

1

One CM

Gets immediate value

2

One brand

Builds shared context

3

The team

Joins the workflow

4

More brands

Expand the account

Content

Two lines, product and brand vision. The main engine.

Our own audience

The founders already publish and have organic distribution.

Product loops

Every analysis becomes an insight or report the CM can share with whoever approves the spend.

Paid, in Chile first

Where we can validate the engine at the lowest CAC before scaling it elsewhere.

Adoption starts with individual usefulness and can grow into organizational value. A higher-touch offer for large companies can come later. The opening motion is self-serve.

The metrics that will decide the model

We are not presenting assumptions as results. These are the signals we are instrumenting, and what each one decides.

Which ICP we pick

Activation, the share who connect an account and reach the verdict. Time to value. Whether the analysis resonates. Willingness to pay.

Whether the engine works

Retention at D1, D3 and D7. Whether they use the approval loop or invite the client or team. Publishing reliability with zero silent failures.

Whether it compounds

Conversion, CAC and payback by channel, gross margin at real usage mix, and net revenue retention.

Two honest caveats. First, the pilot today is comped, so it measures value and loop, not conversion. Conversion becomes measurable once the trial and the differentiator are both in users' hands. Second, NRR has to be designed, it does not appear just because a product is self-serve. At $29 and one brand there is no expansion axis yet. Sprout Social, the only public comparable that discloses it, reports 100% blended and falling. Designing our expansion axis is an explicit goal of this stage, not an assumption.