Systems · Leadnix
PreviewFrom brief to campaign, with every claim checked before it ships.
A claim nobody checked is the one that gets read back to you — by a regulator, a competitor, or a customer with a screenshot. Give Leadnix a brief and it returns governed copy, media and a launch packet — the exact number of assets you asked for, with the source of every claim attached. What it will not do is publish something nobody approved, or spend your money.
Five multiple-choice questions. A person replies within one business hour.
The problem
The writing stopped being the bottleneck. The accountability did not.
A model can produce forty versions of a headline in a minute. That was never the hard part of marketing. The hard part is that somebody puts their name to every sentence that leaves the building: the performance claim, the price promise, the comparison, the testimonial, the small print under the offer.
So the tools got faster and the approval queue got longer. All of that volume arrives at the one person allowed to say yes, with no record of what the last yes actually covered and no way to see which sentences in the new draft are the ones that changed. Review becomes re-reading, and re-reading does not scale.
Then there is the quieter risk. A campaign tool that can publish can publish at three in the morning. A campaign tool that can spend can spend faster than anyone reads the dashboard. Neither of those is a copywriting problem, and no amount of model quality fixes them.
The mechanism
Brief in. Packet out. A name on anything that leaves.
Every stage leaves something you can inspect rather than a chat log you have to reconstruct, and the two stages that touch the outside world stop at a person. That is the difference between a pipeline and a prompt box.
- 01
The brief is the source of truth
Everything is written back to the brief — the positioning, the offer, the claims and where each one came from. The campaign snapshot is frozen at launch, so what shipped stays exactly what shipped, whatever anyone edits afterwards.
- 02
Brand context loads before anything is written
Your tone, the claims you have already approved, the words you never use, the comparisons you will not make. The agent starts inside those constraints instead of being corrected back into them three drafts later.
- 03
You get the count you asked for
Ask for one ad, get one ad. Not six with a note saying pick your favourite. Exact-count output sounds like a small thing until you have spent a year deleting the other five.
- 04
The media studio builds the assets
Video and image work is produced against the same brief as the copy, so the thing people see and the thing people read make the same promise.
- 05
The launch packet is assembled, not published
Copy, media, audience, budget and schedule land in one packet with the claim sources attached. Assembled is as far as an agent gets on its own.
The governed hook
Two things an agent does not get to do here.
Everything else in the pipeline is reversible — a draft is a draft, a packet can be thrown away. These two are not, so they are classified before they run and they stop at a person.
Publish
T3Anything that reaches the public — an ad, a post, an email to a list — is an external action. The agent prepares it and the claims review runs against it; a named person releases it. The approval is recorded with who approved, what they were shown and which version of the policy was in force at the time.
Spend
T3Releasing budget is a person’s action, every time, however small the number. The agent builds the media plan, prices it and makes the case for it. It does not fund it. When it tries, the refusal is written to the record with everything else.
The claims review is the part that earns its keep. Every factual claim in a draft is traced back to the source you approved it from, and anything new is marked as new. Your reviewer stops re-reading the whole document and starts reading the difference, which is the only version of this job that survives contact with a real campaign calendar.
What is in it
Four modules, each carrying its real status.
Campaign pipeline
PreviewBrief to governed copy, media and launch packets, with exact-count output and the claim trail attached to each asset.
Brand context
PreviewThe approved-claims library, your tone, your exclusions and your product facts. Every run loads it before it writes a word.
Media studio
PreviewVideo and image production against the same brief, so the creative and the copy do not drift apart between reviews.
Agency Studio
Design partnerClient workspaces, white-label apps on your own brand and usage rebilling with your markup. Specified and reviewed; built with the first agency partner.
For agencies
Design partnerSix months later the client asks who approved that ad.
You carry the liability for what you published on their behalf, and the answer is somewhere in a shared inbox. The agency version of Leadnix is a different shape from the in-house one, and it is the extension we most want a design partner for.
A workspace per client. Each client’s briefs, approved claims, assets and approval history live in their own workspace, fenced from every other client you run. An account manager who moves between three clients in an afternoon does not carry data between them.
Apps on your brand, not ours. The thing your client logs into carries your name and sits on your domain. Our governance runs underneath it, which is exactly where a client should never have to think about it.
Usage you can rebill. Consumption is metered per client workspace, so an agency can put its own markup on it and invoice it as a line the client understands, rather than absorbing it as a cost of doing business.
The part the client never sees. Governance they cannot switch off, and an approval record per client: who approved it, what they were shown, which version went out. That record is the product. The copy is the thing you were going to produce anyway.
Where this actually is
Deployed internally. No customers on it yet.
Leadnix is in preview. It is built, it is deployed and it runs. It has no external customers, and we have published no campaign results because there are none to publish — the only campaigns it has run are our own.
We could dress that up. Here is why we would rather not. Every growth platform you look at this quarter will show you a case study, and you already know how those are assembled: the friendliest client, the best available number, and no counterfactual anywhere. Showing you one would be asking you to trust our arithmetic on a sample of one. We would rather spend that paragraph on something you can check.
So the offer is a measurement instead of a testimonial. The X-Ray reads your own operation — what you actually publish, how long approval really takes, how much of the work repeats, and hands you the baseline. If the numbers are not there, you have still learned your own baseline, which was worth having anyway. That is the same deal we would want from a vendor our size.
Straight answers
The objections a marketing lead actually has.
Ask for one ad. Get one ad.
And nothing goes out carrying a claim that nobody approved. Start with the measurement, and you will have your own baseline in writing: what you publish, how long approval really takes, and how much of it repeats.
Five multiple-choice questions. A person replies within one business hour.
Preview: built and running for our own work, being packaged. Agency Studio is design partner.