Updated 29 September 2026 · by Amy Wilkinson

AI growth automation

Growth automation that prepares the work. A person approves it.

We automate the daily work behind growth, the research, the drafting, the routing and the follow-up, and keep the decisions with your team. Everything is built in the tools you already use, and nothing goes out without a named person.

Where the line sits

What we automate, and what we never do.

The workAutomatedStays with a person
Finding companiesReading the market daily, researching each company, judging fit against your definitionDeciding who a good customer is
Reaching themFinding the warmest route, drafting the opener in your voice, planning the follow-upSending, from your own account
ContentPlanning the calendar from releases, calls and results, drafting every format with sources, staging it in your CMSWhat you believe, and approving what goes out
SocialPlanning native posts for the brand, founder and team, scheduling them, drafting replies to commentsEach owner approving their own posts
SearchMapping queries, writing page briefs, proposing internal linksApproving every page that goes live
AdsBuilding audiences and creative from messages that already convertSetting the budget and launching
ReportingWhat went out, what came back, what to try next weekChoosing what to change

Why most automation disappoints

Speed without judgment is a faster way to be wrong.

Most teams that tried growth automation remember the same things: a list of people who were never going to buy, messages that sounded like nobody, and a sender reputation that took months to repair.

None of that is the software's fault. It is what happens when nobody decides, in writing, who is worth contacting, what is true enough to say, and who approves. Our builds start there. The system judges fit before it drafts, ties every claim to a source, checks every draft against your own writing, and stops at a person.

The result is less volume and more replies. In our own run, one in ten companies contacted booked a call in the first week. Read how.

The approval model

Three permissions, held by different people.

PrepareThe system researches, drafts and queues. It can do this every morning without anyone lifting a finger.
ApproveA named person on your team reads, edits and approves. Approval is tied to the specific draft; there is no blanket permission.
SendMessages go from your own accounts. Campaigns land paused until someone sets a budget. Every send has a receipt.

What we automate

Six systems, one shared memory.

Questions

Questions about growth automation.

What is growth automation?
Growth automation is using software, and now AI, to do the repeatable work of growth: researching accounts, building lists, drafting outreach and content, routing replies, preparing reports. Done well, it prepares the work and a person decides what goes out. Done badly, it sends a lot of things nobody checked.
Is growth automation the same as marketing automation?
No. Marketing automation, in tools like HubSpot or Mailchimp, sends pre-written messages when a rule fires. Growth automation does the thinking work that happens before a message exists: who to contact, why now, what to say. The two work well together; one prepares, the other delivers.
Will it send anything without my approval?
No. Prepare, approve and send are separate permissions. The system holds the first; your team holds the other two. Every send is logged with the name of the person who approved it.
Which tools does it use?
Your CRM, your inbox, Notion or Google Docs, Slack and your ad accounts, plus the data and AI layer underneath: enrichment and verified contacts from providers like Clay and Hunter, and current AI models for research, judgment and drafting. We keep that layer current so your team does not have to.
What should never be automated?
Sending, spending and deciding what you believe. The system can prepare an email, a campaign or a piece, but a person sends it, sets the budget and owns the opinion.
What does growth automation cost?
Builds start from EUR 6,000, quoted in writing after a paid diagnosis that is credited against the build. Retainers are optional. Running costs for the AI and data layer are usually small next to a salary. How pricing works is on the build page.

Related

Start here

Let's build what runs without you.

Twenty minutes on how growth runs in your company today. If a build is not the answer, we will tell you what is.

See pricing