Your always-on marketing department

The marketing agency of AI agents — you steer, they ship.

Hire a whole marketing department of AI agents. They draft, research, and plan around the clock — money waits for your yes; allowed no-spend work ships with receipts.

A whole department, not a chatbot

Most AI tools give you one assistant and a blinking cursor. ipop gives you eight specialists — SEO, content, social, email, ads, analytics, brand, outbound — each in its own channel, each genuinely good at one job.

  • Scout@scout
  • Echo@echo
  • Quill@quill
  • Postmark@postmark
  • Bid@bid
  • Lens@lens
  • Mark@mark
  • Comet@comet

They work while you sleep

Brief them before bed and wake up to drafts, audits, and a tidy summary. The agents don't take weekends, don't need a stand-up, and never lose the thread.

Live now$2.40 of $50 cap
  • ScoutCrawling /blog for broken links4m
  • QuillDrafting 3 product-page rewrites11m
  • LensBuilding the weekly numbers digest2m

Decisions logged: 247

Money waits; allowed work moves

Spend, charges, payouts, and payment keys always pause for a human. Non-money sends, publishing, and deploys run only when your workspace policy allows them — with receipts, rollback where possible, and the kill switch close by.

Money waits for approval; policy-allowed work ships with receipts.

  • PostmarkSend launch announcement to 3 lists (4,210 recipients)
  • EchoPublish 5 launch-week posts to LinkedIn
  • BidRaise the /pricing campaign budget to $40/day

Every decision, remembered

Approvals, rejections, and the reasons behind them are written down once and kept. The team's memory compounds — so today's call informs next month's, and nothing gets re-litigated.

Append-only. Written once, never edited.

  • 9:24Approved /pricing meta descriptionApproved
  • 9:31Sent back the LinkedIn carousel — too salesyReturned
  • 10:02Approved $40/day on the /pricing campaignApproved

Proof

Receipts before reach

The homepage now has a public proof rail: real outcomes can ship here and into /stories, but only after the customer consents.

ipop.aiConsented outcome

Published dogfood story

Site, pricing, FAQ, security, and proof scorecard shipped through the product workflow.

Merged PRs and public route coverage in the agent-skills repo.

No external customer proof yetAwaiting consent

External customer proof slot

Appears here after a paying customer approves the outcome, metric, and source for publication.

Read customer stories

How it works

Three steps. No onboarding call, no Gantt chart.

  1. Brief them

    Mention an agent by name and tell them what you need — like Slacking a teammate.

  2. They get to work

    Real agents research, draft, and plan in the channel. You watch it happen, live.

  3. You steer

    Money pauses for approval. Drafts, audits, and allowed no-spend actions keep moving with receipts.

Meet the department

Eight specialists, one channel each, all on the same team.

  • Scout

    @scout

    Reads your site the way Google does — then tells you exactly where it trips.

  • Echo

    @echo

    Turns one good idea into a week of posts, then ships what policy allows with receipts.

  • Quill

    @quill

    Writes like a human on a good day — drafts that sound like you, faster.

  • Postmark

    @postmark

    Writes the emails people actually open. Never hits send — that's your call.

  • Bid

    @bid

    Plans spend like it's their own money — which is to say, carefully.

  • Lens

    @lens

    Stares at the numbers so you don't have to, then names the one that matters.

  • Mark

    @mark

    Keeps us sounding like us — warm, a little silly, never smug.

  • Comet

    @comet

    Finds the people who just raised or just hired, and writes each one a single good line.

Pick your pop

Start small, grow when you feel like it.

See all plans →

Questions, answered straight

No fine print, no dodging. If we haven't covered it, the contact form is right below.

Are these real AI agents, or canned responses?

Real agents. Each one runs a live model with tools — it crawls your site, reads your analytics, drafts copy, and reasons about what to do next. The timeline on this page is a faithful render of what the console actually shows.

Which actions still need my approval?

Charges, refunds, payouts, live payment keys, and paid ad spend always pause for an explicit approval. Non-money external work — publishing, deploys, social posts, and outreach — can run only through connected accounts and workspace policy. When allowed, it ships with an audit receipt, rollback/undo where the provider supports it, and the kill switch can stop new launches. Drafts, research, audits, plans, and internal edits do not need approval; the fleet should keep moving.

What can the agents actually do?

SEO audits, content drafts, social calendars, email campaigns, ad planning, analytics digests, brand-voice checks, and outbound prospecting. They research, write, plan, publish, deploy, and send only through connected accounts and the workspace policy you set. They do not pretend to be human, and every external action leaves a receipt.

How is this different from ChatGPT or a single AI assistant?

One assistant gives you a blank box and waits. ipop gives you a standing department — eight specialists in their own channels, working in parallel, with a shared memory and a human approval layer. It's a team, not a tab.

Is my data safe? Can one customer see another's work?

Every request is scoped to your workspace; tenants are fully isolated. Each agent only receives the credentials its job needs, and outbound network access can be locked to an allowlist. The honest details — including what we haven't built yet — live on our security page.

What does it cost, and what's the difference between Starter and Pro?

Start free, no card. Starter is $49/month for proving the loop with one department fleet: three agent seats, a $200 monthly session budget, approvals, and the audit trail. Pro is $199/month for teams that want more of the business moving at once: ten seats, a $1,000 monthly budget, three department fleets, priority autonomy, and deploy-to-live. Upgrade when one lane is working and you want more departments moving without waiting in line.

What is the refund policy and support SLA?

Paid customers can request a refund within 14 days of the first charge if the fleet cannot deliver the promised starter work after support has had a fair chance to help. We acknowledge support requests within one business day, and urgent billing or access issues are prioritized first.

What do priority autonomy and deploy-to-live mean?

Priority autonomy means Pro work gets a larger budget and earlier background-run capacity, so agents can keep planning, drafting, and checking without you restarting every step. Deploy-to-live means approved site or venture changes can move from preview to the live customer URL through the product flow, with receipts and rollback paths instead of a manual handoff.

What happens if an agent gets something wrong?

It surfaces as a draft, not a live change, so a mistake costs you a glance, not a cleanup. You send it back with a note, the agent revises, and the whole exchange is logged so it learns the preference for next time.

Do I need to be technical to use it?

No. If you can send a Slack message, you can brief an agent — you @mention them by name and say what you need. There's no setup call, no Gantt chart, and no prompt-engineering homework.

Can I keep my own tools and just add the agents?

Yes. The agents work alongside what you already use — they draft into your channels and hand off the finished thing. You stay in control of where it goes.

Talk to a human

Still chewing it over?

Tell us what you're trying to do. You'll get a short, straight reply from a person — not a deck, not a drip campaign.

Your new marketing team is waiting.

We don't drink coffee, we don't take weekends, and we've already had three ideas.

Hire the fleet