How long until the first agent is live?
The first agent goes into draft mode within four weeks.
Week one maps the work and connects your tools. From there the agent proposes changes for your team to approve, before it acts on anything itself.
Running Shopify Plus stores since 2012.
14+
Years on Shopify Plus
210+
Shopify stores built
$1.5B+
In client GMV
3
Offices: Melbourne, LA and London
What is Run
AI agents that run your store.
Not the ones that shop from it.
Run is Flux's 100-day partnership for Shopify Plus brands. We build AI agents that handle repeat work across your store, marketing and support, then run them against a target we agree with you. Getting recommended by shopping agents like ChatGPT is a separate job, covered by Lift.
Agents we build
Three agents.
Each measured.
Your strategist picks the first agents from a map of where your team's hours go. Our engineers build them, connect them to your stack and keep them running.
Tell us about your project01
Revenue agents.
They keep collections sorted, product pages current and email flows personal. Measured on conversion rate and email revenue.
02
Operations agents.
They handle catalogue updates, support triage and order exceptions. Measured on hours handed back and response times.
03
Store engineering.
Our Shopify Plus engineers ship fixes and features every week. Measured on site speed and release cadence.
How an agent goes live
From draft to autonomous.
Each agent earns trust first.
No agent touches your live store on day one. Each moves through the same stages, and your team decides when it's ready for the next.
Week 1
Map the work
List the repeat tasks, the hours they take and the risk if one goes wrong.
Weeks 2–4
Draft mode
The agent proposes changes. Your team approves or rejects each one.
Weeks 5–8
Supervised
The agent acts on low-risk tasks and logs every change it makes.
Weeks 9–12
Within limits
The agent runs inside rules you set. Exceptions go to a person in Slack.
Day 100
Review
Compare results with the baseline and choose the next agents.
Example scorecard
What you see every fortnight.
One target, one number and a short note on what changed. This example tracks hours of repeat work handed to agents.
Target #1 / week 8 of 14
Hand repeat work to agents
112 hrs
Agent hours this month
160 hrs
Target for day 100
70% of target · 48 hrs to go
Day 1
Week 8
Fortnightly note
Written by your strategist.
What changed
This example compares 112 hours of catalogue and support work done by agents in the last 30 days with 64 hours in the 30 days before.
What this means
Agents are moving towards Target #1. We'll check which tasks still need a person and why.
Next fortnight
Hand the next set of catalogue tasks to agents, then review progress against the remaining 48 hours.
The tools we connect
Agents work in your tools.
No new platform to learn.
Agents connect to the systems your team already uses and report back in Slack, so nothing moves to another dashboard.
Shopify
Products, orders and customers
Google Analytics
Traffic and conversion
Klaviyo
Email and SMS flows
Gorgias
Support tickets
Google Ads
Paid search spend
Meta
Paid social spend
Slack
Where agents report in
PostHog
What visitors do on site
Is Run right for you
When /run fits.
And when to start elsewhere.
Agents pay off when the same work comes round every week. If it doesn't, we'll say so.
Run fits when
- Your team repeats the same catalogue, merchandising or support tasks every week.
- You sell on Shopify Plus and run tools such as Klaviyo and Gorgias.
- You want results measured against a target, not reported as hours.
Start elsewhere when
- A Shopify Flow rule already does the job. We'll set it up and skip the agent.
- You want AI search to recommend your products. Start with Lift.
- You need a new store or a replatform first. Start with a build.
The next cycle
Agents keep learning.
The work compounds.
Every 100 days we set a new outcome and add agents where they'll move it. Want AI search to recommend your products too? Add Lift.
See how Lift works →Questions about Run
Questions.
Will agents replace my team?
No. They take the repeat work so your team can focus on decisions.
We map which tasks suit an agent and which need a person, then build around the team you have.
Do we need Shopify Plus?
Run is built for Shopify Plus brands.
Plus gives agents higher API rate limits and Plus-only features such as checkout customisation, so they can take on more of the work. On another Shopify plan? Talk to us and we'll tell you what's possible.
How is Run different from a retainer?
Run is tied to outcomes, not a bank of hours.
We agree a target up front, such as hours handed back or revenue from a flow, and report against it through to day 100.
Is it safe to let agents change our store?
Agents work inside rules you approve.
We start in draft mode, add approval steps for anything customer-facing and log every change an agent makes.
How is Run different from Shopify Sidekick or Flow?
Sidekick and Flow work inside Shopify. Run's agents work across your whole stack.
Sidekick is Shopify's admin assistant and Flow runs rules-based automations. We use Flow where a simple rule does the job, and build agents for work that spans Shopify, Klaviyo, Gorgias and your other tools.
What happens at day 100?
You set the next outcome, or add Lift.
We review results against your targets and agree the next outcome. Each cycle can extend agents into new parts of the store.
Hand the repeat work over.
Get more from the team you have.
Let's talk about your store, your goals and what a 100-day Run partnership could deliver.
