Marketing
Marketing that ships every week and proves what worked.
At a glance
A product launch has to go live in three markets by Tuesday.
- Deployed AI
About a day per market
Marketers rewrite and publish everything
- Operational AI
Hours, only for new or risky claims
AI runs localisation and scheduling; people own the message
Illustration
What matters
What marketing teams are judged on.
Pipeline from marketing
Campaigns judged by the deals they start, not the clicks they get.
One voice everywhere
Every asset on brand, in every market and channel.
Speed to market
From idea to live campaign in days, not weeks.
Spend that follows results
Budget moved to what works, with the evidence to show why.
Walkthrough
The same request, three ways.
The request
MessageHead of product marketing
Localise the 4.2 launch
UK, Germany and the UAE by Tuesday.
Same claims as the master. Tracking on everything.
Autonomy level 1
Deployed AI
A chat assistant. Someone asks, it answers, people do the rest.
- A person types
Write a LinkedIn post announcing our new analytics release.
- AI replies
Exciting news! Our latest release helps you make smarter decisions, faster. Discover what's new today!
- Person
Marketer. Adapts it for three markets, checks the claims by hand, chases sign-off by email and posts it.
3 steps · one to a person · about a day per market
- How long
- About a day per market
- Who does the work
- Marketers rewrite and publish everything
- What keeps it safe
- None: claims unchecked, approvals lost in email
- What goes to people
- Everything
Autonomy level 2
Specialised AI
It works from your data and tools and prepares the work. A person approves.
- Trigger
The request arrives in the marketing board.
- Tools
brand.glossary() · claims.library("4.2") · results.past_launches(3)Returned: The approved master copy, 14 claims with their proof, and style notes for Germany and the Gulf.
- AI replies
Three localised drafts in the voice of each market, every claim linked to its proof.
- Person
Product marketer. Checks the claims and approves the three drafts.
Approved - Action
Scheduled for Tuesday in each market's local time, with tracking links.
5 steps · one to a person · about two hours of review
- How long
- About two hours of review
- Who does the work
- AI drafts; people approve every asset
- What keeps it safe
- Every claim linked to its proof
- What goes to people
- Every asset
Autonomy level 3
Operational AI
It runs the process within set limits, logs every step and hands people the exceptions.
- Trigger
Every approved master, for every market in the plan.
- Rule
Localise approved masters only, and check every claim against each market's rules.
Comparative and environmental claims always go to legal.
- Action
The UK and UAE versions pass and are scheduled with tracking links.
- Person
Exception: goes to a person
Legal. The German draft says 'twice as fast as Brand X' and there is no proof on file. It is blocked and sent to legal with a claim that can be proven.
Escalated - Record
[Tue 07:00]Approvals, claim checks and versions kept for each market.
- Report
Day seven: results against the launch target, by market, with what to change.
6 steps · one to a person · hours, only for new or risky claims
- How long
- Hours, only for new or risky claims
- Who does the work
- AI runs localisation and scheduling; people own the message
- What keeps it safe
- Claims rules, approval log, legal on comparisons
- What goes to people
- New, comparative or regulated claims
Controls
- Owner
- Head of marketing
- Measure
- Launch on time; pipeline by market
- Limit
- Approved masters only
- Log
- Every claim check and approval
- Fallback
- Marketers publish by hand
- Stop
- Pause scheduling in one switch
Illustrative: names, figures and timings are invented.
Where to start
Processes that pay back first.
2Automate with approval
First drafts in your voice
Claims checked by a person.
2Automate with approval
Localisation for each market
Native review for key markets.
3Run with exceptions
Scheduling and publishing
Within the approved calendar.
3Run with exceptions
Reporting and attribution
From your CRM and ad platforms.
3Run with exceptions
Market and competitor monitoring
Signals flagged to the team weekly.
1Assist
Positioning and messaging
AI researches; people decide.
What we build and run
Built in your systems, run with your team.
- A content engine that writes in your voice, with sources
- Campaign kits for every launch and market
- Publishing and approvals inside your existing tools
- Reporting tied to pipeline, by market
What we measure
Numbers you already track.
- Pipeline influenced
- Cost per opportunity
- Time from brief to live
- Assets published each week
- Share approved first time
The Academy
Learn to run it yourselves: the Marketing track.
Put AI into content, campaigns and reporting without losing your voice.
Being built