- Home
- Case studies
- Case
Artificial Workflow: an AI automation agency for consultants at $10K a month
Artificial Workflow is an AI automation agency that designs, builds and runs workflow automations for consultants and other business teams, and was at $10K a month in founder-reported revenue in June 2024 (Starter Story interview).
The short answer to how it got there: it did not start as a product. It started as custom AI consulting, ran a large volume of cold outreach on LinkedIn, noticed it was rebuilding the same solution for different clients, and then packaged that solution into a subscription platform. Growth after that came from four places the founder names repeatedly: LinkedIn outreach, email, SEO-driven tool pages, and referrals.
How it started
Artificial Workflow started in December 2023 and is based in Brisbane, Queensland, Australia. One co-founder, Andrew, came from marketing in the Web3 space. The other, James, spent time as a management consultant at EY, where he watched consultants put long hours into manual data processing, analysis and other repetitive work.
That observation shaped the first version of the business: an AI consulting company selling bespoke work to firms that wanted to integrate AI. Getting that work meant outreach. In the founder's words: "This venture involved a significant amount of cold outreach on LinkedIn, resulting in around 200 sales calls primarily with consulting firms we found on the platform."
Those calls produced the pivot. The team kept building the same technology for multiple clients, and decided to package the core functionality into a SaaS-like platform and sell access by subscription.
The first build was deliberately narrow: a knowledge-work analysis tool that let consultants upload interview transcripts and have the AI analyse them. The team sketched wireframes in Figma, revised them using feedback from the early LinkedIn calls, then hired a very cheap developer from Fiverr to build the MVP. It shipped, "but just barely." Four to five enterprise clients were onboarded before the team discovered the developer had not built async into the app: when one client was using it, everyone else waited. Asked what they would do differently, the answer was to start by finding a technical co-founder who could build the whole MVP independently and later lead the development team.
What it sells and how it charges
On its official website, fetched on 2026-09-30, Artificial Workflow presents itself as an AI automation agency that designs, builds and runs business workflow automations. It describes its work as turning repeated operating work into a managed system: documenting the existing chain of handoffs first, removing the steps a machine can handle, and keeping people on decisions that need context.
The site lays out a four-step pattern for a live system:
- Capture the request — a form, inbox, CRM event or scheduled trigger starts the work.
- Apply rules and AI — the system classifies, drafts, checks or routes the item.
- Keep a human checkpoint — high-risk actions wait for approval before moving.
- Write back and report — every result lands in the right system with a clear record.
The services listed are AI implementation services, AI automation for small business, AI automation consultant, business process automation, and the full AI automation services map. The site also sets expectations low on purpose: a useful first project can be one workflow, and does not need to become a company-wide transformation program.
The website does not publish prices; no pricing page was present when it was fetched. Prices come from the founder's own account in the June 2024 interview: enterprise clients at $299 per month, and paid training and onboarding at a flat fee of $2,500 for four hours of calls plus training materials.
How it got its first customers and what kept growth going
Cold outreach on LinkedIn remained the leading revenue source, and the reason was personal rather than technical: "Our cold outreach on LinkedIn, led by James' strong reputation in the consulting industry, has been particularly effective." In the interview the founder says James can book tens of sales demos a week with potential enterprise clients.
Email marketing handled retention. The team sends weekly updates and exclusive deals, and began sending "first month free" emails to leads who had not been interested enough to buy at first. About 30% of those leads accepted the offer.
SEO became the second engine. By targeting relevant keywords and publishing blog posts or tool pages for them, the site drew almost 10,000 clicks from SEO in a single month. Ten tool pages were live at the time of the interview, including an AI Paragraph Generator, AI Email Generator and AI Script Generator. Two things matter about how they were built: every tool sits in the same niche, which the founder argues builds topical authority, and each tool page links to the others near the bottom, passing internal links between them.
Networking and partnerships added a third channel: local industry events, plus a referral program that offers clients a 40% cut when they refer work.
The numbers
All figures below are what the founder reported in the June 2024 interview, not current performance. The founder attributed the revenue to EchoBase AI, the product inside the company.
| Metric | Figure | Basis |
|---|---|---|
| Monthly recurring revenue | $10K | Founder-reported, June 2024 interview |
| Enterprise share | About 80% of revenue, from enterprise clients paying $299/month | Founder-reported, June 2024 interview |
| Subscribed enterprises | 30, some paying for extra seats | Founder-reported, June 2024 interview |
| Training and onboarding | $2,500 flat fee for 4 hours of calls plus materials | Founder-reported, June 2024 interview |
| Team | 3 founders, 2 employees | As stated at the time of the interview |
| Started | December 2023 | Founder-reported |
| Based in | Brisbane, Queensland, Australia | Founder-reported |
| Starting cost | Not stated | The founder described a very cheap Fiverr developer and a tiny budget, but gave no figure |
How it compares
Against 1M.chat's analysis of founder-reported cases, the business sits below its category median. That analysis covers 1,997 businesses with founder-reported revenue in Starter Story interviews from 2015 to 2026 — a self-selected sample of founders who chose to be interviewed, not a random sample of all businesses.
Within the AI tools & apps category, 84 businesses report revenue. Their median monthly revenue is $16.5K, and 26% of the category reports $1M a year or more. Median headcount is 2 employees, and 42% of the category was started by one founder. Among the 40 businesses in the category with channel information, the most common growth channels are organic social media (60%), SEO and organic search (60%), partnerships and affiliates (35%), and launch platforms such as Product Hunt, AppSumo and marketplaces (32%).
Artificial Workflow's $10K a month is below the $16.5K category median. Its two employees match the category median of two, and it was founded by three people rather than one. Two of its three stated channels — SEO and partnerships — appear in the category's top four, while its heaviest channel, LinkedIn cold outreach, is a form of organic social activity that the category data does not break out separately.
What a founder can take from it
Watch for the same build repeating. Artificial Workflow's pivot came from noticing it was writing the same solution for different clients. If your custom work keeps converging on one artefact, that artefact is the product.
Attach the channel to whoever has standing. The outreach worked because James was already known in consulting. Cold outreach still cost about 200 sales calls before the pattern became obvious.
Build tools on exact-match keywords, then link them. The strategy the founder describes is simple to copy in outline: pick high-volume terms you can build a small free tool for, publish one page per tool, keep them in one niche, and link them to each other. That produced almost 10,000 SEO clicks in a month for this team.
Make a specific offer to lukewarm leads. "First month free" converted about 30% of leads who had already declined to buy. A 40% referral cut does similar work on the partnership side.
Do not underfund the build. The cheap MVP shipped without concurrency handling and became a problem at four to five clients. The founder's own stated fix was a technical co-founder who could build the MVP and later lead development.
Sources
Revenue, team and start-cost figures are what the founder reported at the interview date, not current figures and not a prediction of what you will earn. Product and pricing facts come from the business's own website as checked on 30 September 2026. Method & sources · Disclosure