There is no framework
Every campaign argues whatever its author believes that week. The cost: the team rebuilds the same strategic argument on every campaign, because the last one left no trace.
Message governance and attribution for B2B GTM · Design partner release
Naros holds your positioning as a structured system, drafts every campaign asset from one governed messaging decision, and keeps that decision identifiable into HubSpot and LinkedIn. Performance comes back by message, not only by channel.
For B2B SaaS marketing teams running demand gen on HubSpot · Six places
No customers yet. Everything on this page is either something you can verify yourself, or something we are telling you is not built.
01 / The problem
Every company at your stage already has the pieces — positioning, ICP definitions, a CRM, campaigns across five or six channels, dashboards on all of it. What nobody has is the thing that connects them. Your CRM reports the channel that produced the leads and has no field for the message that converted them. So the campaign that worked cannot be repeated on purpose, and the one that failed cannot be ruled out. It produces numbers. It cannot produce knowledge.
02 / Why your stack can't close it
This is not a gap you close by adding another tool in one of these categories. It sits between them.
Engineering has version control. Sales has a CRM. Finance has a ledger. Each of them took a thing the company depends on and made it a record instead of a habit. Messaging never got that layer, which is why it still runs on documents, briefs and whoever was in the room. Your company has a strategy. It governs nothing you actually say.
03 / How Naros works · one thing survives the journey
Naros adds one layer between the strategy you own and the tools you run. Every campaign is composed from one governed messaging decision, and that decision keeps an identity of its own, so what comes back is legible by message and not only by channel.
Everything in magenta is the same messaging decision. Follow it down the page.
The work you already did, or the company material that contains it. Naros does not replace it. Today none of it carries into execution, which is why the line beside it is broken.
Your framework stops being a document and becomes a system execution can read, edit and inherit from.
Every campaign is composed from one explicit decision, checked against a codified methodology before anything is generated.
That decision is given an identity of its own, so it can be recognised again when results come back.
The messaging decision, now a thing with a name. This is the object everything below is about.
38 rules passed · 1 advisoryNo CRM to migrate, no reporting to rebuild, no channel to re-platform.
Each channel still sounds like itself. Angles, hooks, formats and length stay free to differ. The argument underneath does not.
The reporting you already trust, with the one variable it never had. Naros measures what Naros governs, so activity it did not produce stays outside the picture by design.
04 / Reporting by message
Your GTM reporting already knows where performance came from. It has never known what you were saying when it happened.
| Block | Value proposition | ICP / Persona | Funnel | Leads | MQL % | SQLs | Pipeline |
|---|---|---|---|---|---|---|---|
| BLK-001 | Predict and prevent churn with prescribed interventions, not health scores | Growth-Stage · VP Customer Success | MOFU | 312 | 18.6% | 47 | €247K |
| BLK-002 | Cut the reporting cycle from three weeks to three days | Growth-Stage · VP Customer Success | MOFU | 198 | 17.5% | 31 | €189K |
| BLK-003 | Stop paying for renewals your team could have savedvolume, no depth | Enterprise · Head of CS Ops | TOFU | 274 | 9.1% | 12 | €61K |
Real product screen, restyled with the new design system · the figures are what the product shows
Look at the third row. It produced more leads than the second and a fifth of the pipeline. That is a messaging finding, not a channel one, and no report in your stack would surface it today.
What this is not
Naros attributes outcomes to messages under a defined, configurable model. It shows association, not causation. Where the evidence does not support an assignment, the outcome stays unattributed rather than credited to a guess. Naros measures what Naros governs, so activity it did not produce stays outside the picture by design.
05 / What runs today
Naros is in design-partner release. Everything on this page is either something you can verify yourself, or something we are telling you is not built. This table sits on the homepage rather than in a footer on purpose: found later, it would read like something we had hidden.
And what Naros
will not do
06 / Design partner program
If your team is running real campaigns and cannot answer which message is working, that is the problem we are building for. Design partners run their own messaging, their own campaigns and their own data through the product, and shape what gets built next.
Six places, open now · reviewed within five working days
No cost during the design partner period. What we ask for instead is candour: that you use Naros on campaigns that matter, and tell us where it gets in the way.
Behind every generation sits a documented GTM methodology, v1.0: market context, funnel stage logic, ICP to persona architecture, ACV economics, channel selection and a controlled vocabulary for message framing. Design partners get access to it during the pilot.
Naros · the mechanism, in full
If you are deciding whether to put real campaigns through an early product, this is the page to read. Every step below is either working today or marked as not built. Nothing here is aspiration written in the present tense.
Six places · open now
01 / The problem, in full
Every company at your stage already has the pieces. Positioning of some kind, ICP definitions, a CRM, campaigns across five or six channels, dashboards on all of it. What nobody has is the thing that connects them.
Every campaign argues whatever its author believes that week. The cost: the team rebuilds the same strategic argument on every campaign, because the last one left no trace.
It went into a drive and has no authority over the Tuesday the campaign gets built. The cost: the positioning you paid for never reaches the market it was written for.
Each person interprets the positioning their own way, and not one interpretation is recorded. The cost: the volume scaled. The governance did not.
All three end the same way. Your CRM reports the channel that produced the leads and has no field for the message that converted them. So the campaign that worked cannot be repeated on purpose, and the one that failed cannot be ruled out. It produces numbers. It cannot produce knowledge.
01 / The six steps
That asymmetry is the whole adoption cost. There is no migration, no reporting to rebuild and no channel to re-platform, because Naros sits beside the stack rather than inside it.
Naros reads the material you already have and structures it into a framework, at every level of your go-to-market.
Your positioning, sales decks, website, whatever contains the argument you make today. You do not start again. Corporate, ICP and persona today, with product level next. You review and edit every part of it, and where your material is too generic to execute from, it says so rather than accepting it.
This happens once, at onboarding. Everything below it happens per campaign, and the framework is edited from then on rather than rebuilt.
You choose the ICP, the persona and the funnel stage.
These are not free fields. They come from the framework you approved in step 01, and the internal methodology resolves that framework for the combination you pick — ranking the channels that fit the stage and narrowing the framings to the ones your material actually supports.
The four struck through assert a moat. Your framework has not established a differentiator, so they cannot be selected. Establish it and they become available. Channels are ranked, never blocked — pick the one it advises against and that choice is recorded against the message. The framing rule is the one place the product does say no.
The assistant composes the messaging decision with you, inside those limits.
It proposes the problem, the value proposition, the proof and the framing, drawn from the level of the framework that applies to that buyer. You edit any of it. What comes out is one decision with a name, and that name is what everything downstream is built from.
Checked as you edit, not after you launch. Twenty-two deterministic rules run while you edit, debounced, and block the save if they fail. Sixteen AI-assisted checks run after generation, asynchronously, and never block.
You approve the decision. Only then does anything get written.
Naros drafts each asset in the form its channel needs. Angles, hooks, formats and length stay free to differ. The argument underneath does not.
Real product screen, restyled with the new design system.
The assets land in your own HubSpot and LinkedIn accounts, as drafts.
You connect the accounts once. From then on Naros pushes each asset into your own HubSpot and LinkedIn accounts, and your team activates them from there exactly as it does today. Naros never sends or publishes anything on your behalf. That is a permanent decision, not a current limitation.
Nothing changes about who approves a campaign, who schedules it or where it goes out from. The only difference is what the asset arrives carrying.
Every asset carries the identity of the decision it came from.
So performance returns by message and not only by channel. Inside Naros, an attribution view shows what each message actually produced: engagement, conversions and funnel outcomes, read against the argument that was running at the time rather than against a campaign name.
Slice the same performance by
| Block | Value proposition | Funnel | Leads | MQL % | Pipeline |
|---|---|---|---|---|---|
| BLK-001 | Predict and prevent churn with prescribed interventions, not health scores | MOFU | 312 | 18.6% | €247K |
| BLK-002 | Cut the reporting cycle from three weeks to three days | MOFU | 198 | 17.5% | €189K |
| BLK-003 | Stop paying for renewals your team could have savedvolume, no depth | TOFU | 274 | 9.1% | €61K |
Real product screen, restyled with the new design system · the figures are what the product shows
Roadmap · not built yet
Once there is enough history behind you, the same evidence starts working in the other direction: which value propositions, framings and proof points have held up for a given ICP and stage, surfaced while you are composing the next decision instead of found months later. That is the point of accumulating the record, and it is the part of the loop still to come.
03 / Where the AI sits
Messaging volume rose sharply once anyone could generate a campaign in a minute. Governance did not. Naros generates too, and most of what makes it different is a description of what the model is structurally prevented from doing.
Only the first step touches your source documents. Everything after it works from the framework you confirmed, so no later step can introduce a claim your framework does not hold.
Where your material does not support a differentiator or a proof point, it says so and names the evidence that would close it, rather than writing something confident.
Tone, hook, format and length are the model's to choose. Who the message is for, what it claims and what backs it are not.
04 / Validation
Governance that cannot be overridden becomes a template your team routes around. Governance that is never recorded is a suggestion. Naros does neither.
Deterministic · blocking
Twenty-two rules, and they stop the save
They run while you edit, debounced, and a failure blocks the save. The hardest of them is the framing rule: the four angles that assert a moat stay unselectable while your framework has no established differentiator. That one runs against you rather than for you.
AI-assisted · advisory
Sixteen checks, and none of them stop you
AI-assisted, run after generation and asynchronously: channel fit for the stage, thin proof, a value proposition the framework supports weakly. You can proceed against every one of them, because the person running the campaign usually knows something the rule does not.
Recorded
Whatever you chose, the version remembers
Every override is written against the version of the message that carries it. A launched campaign keeps the version it ran with, so when the numbers come back you are reading them against what was actually approved.
05 / Proof
We have no customer results, so there is no logo wall here and no percentage without a source behind it. Here is what you can verify on your own systems in about twenty minutes.
Check one · 2 minutes
Then count the campaigns you have shipped since. If the second number is much larger than the first, the document is not governing anything, and nobody on your team did anything wrong.
What you are measuring whether your strategy has any route into execution at all.
Check two · 15 minutes
You will get the campaign name, the source and the lifecycle stage. Now try to retrieve which value proposition that contact actually encountered. There is no field, because nothing upstream ever wrote one.
What you are measuring whether the message exists as data in your stack. It does not.
And one you can run inside the product
The system will decline to write you a differentiator your own documents do not support. It marks the gap, and then it closes the door on every framing that would have depended on that gap being filled. This is the only hard restriction in the product, and it runs against you rather than for you. If you are evaluating a governance product, that behaviour tells you more than any feature list.
The four struck through assert a moat. Your framework has not established one, so they are greyed out and cannot be selected. Establish the differentiator and they become available.
08 / Your material and your data
You are being asked to upload your positioning and connect a HubSpot account. Two things are true today and can be checked. One is not written yet, and saying so is better than a paragraph that sounds reassuring and commits to nothing.
What happens to your material and your data
Naros does not store your contacts, and does not modify your data
People live in your CRM and stay there; what Naros keeps is counts and identifiers. It writes its own metadata against a campaign and does not touch your business properties.
Naros measures only what it governs, and names its AI provider
Activity it did not produce stays outside the picture by design. Use of the Anthropic API is declared explicitly in the onboarding consent, with its DPA — not folded into a general policy.
What happens at the end of the pilot: whether data is deleted, retained, or exported, and whether we can cite your results as a reference. That policy is being written before the first design partner conversation, and it will be agreed with you in writing rather than assumed.
What Naros does with each platform
Naros governs the message a campaign carries. To report performance by message rather than only by channel, it needs to know which campaign a lead came from. It does not read anything it did not produce.
HubSpot
Its own metadata against campaigns and contacts, so the message a campaign carries is recorded where the campaign lives.
Lifecycle and deal data back, so a lead can be followed from the campaign that produced it to the pipeline it became.
So that a lead can be attributed to the message it came from, not only to the channel.
Creates ad content, carrying the governed message into the campaign.
Identifies which campaign produced a lead.
So that lead can be attributed to the message it came from.
06 / What this means for your team
Not all of them arrive at once. This is a curve, not a switch, and it is worth knowing which part arrives when.
Week one
Which argument are we actually making, and does everyone have the same one?
Your positioning stops being a document nobody opens and becomes something execution reads. If two people on your team would describe your differentiator differently, you find that out immediately.
Every campaign after that
Who is this for, what are we claiming, and can we back it?
Composed in minutes from something that already exists, instead of rebuilt from scratch by whoever is running the campaign that week. The strategic argument stops being re-derived every time.
After five campaigns Modelled · no customer has reached this yet
Which of our value propositions is landing with this ICP, and which one only looked good in the document?
For the first time this is a comparison rather than an opinion, because the messages were structured before they ran.
After twenty Modelled · no customer has reached this yet
What should we scale, what has earned another test, and what should we stop repeating?
The second half of that question is the one that saves budget, and it is the one nobody can answer today.
And the one you get asked in the board meeting
Which message is producing pipeline with our best-fit buyers?
Today the honest answer is judgement. That is the answer this replaces.
This compounds, and nothing else in your stack does.
Every other tool in the messaging chain resets when the next campaign starts. Here, campaign twenty is composed against nineteen campaigns of evidence about what worked with whom. The messages become the asset, and they accumulate.
07 / Questions
What if we already have a messaging framework?
Good. Bring it. Naros uses your existing messaging work as source material, structures it into the framework execution runs from, and lets you review and edit every part of it. The objective is to make good strategic work operational, not to replace it.
What if our messaging isn't well defined?
That is more common than the alternative, and it is not a prerequisite you have to solve before you start. Naros builds the structured framework from the company material you already have, and challenges the answers that are too generic to execute from rather than accepting them.
Does Naros replace HubSpot?
No. Naros sits upstream of execution and alongside the stack you already run. HubSpot remains your CRM and your execution system. Naros adds the structured messaging decision those systems do not model, and hands it back so their reporting can finally carry it.
Does message-level attribution prove what caused a conversion?
No, and we are careful about the difference. Attribution assigns credit under a defined, configurable model. Causation requires a separate method. What Naros does is make the message observable, so it can become part of a better decision instead of remaining the one variable nobody could see.
More questions, including setup effort and how activation works →
06 / Design partner program
Six places, filling now. Design partners run their own messaging, their own campaigns and their own data through the product, and shape what gets built next.
Six places, open now · reviewed within five working days
Naros · legal
Naros is in design-partner release. This policy says what we collect, what we do with it, and what we have not decided yet. Where something is not settled, we say so rather than write a paragraph that sounds reassuring and commits to nothing.
Naros is a message governance and attribution tool for B2B go-to-market teams, currently in design-partner release. For the purposes of data protection law, Naros is the data controller for the information you give us directly, and a data processor for the data that flows through the platforms you connect.
When you apply as a design partner or use the product, we may hold:
To report performance by message rather than only by channel, Naros needs to know which campaign a lead came from. In HubSpot it writes its own metadata against campaigns and contacts, and reads lifecycle and deal data back. In LinkedIn it creates ad content and identifies which campaign produced a lead, so that lead can be attributed to the message it came from. It does not read anything it did not produce.
Naros uses the Anthropic API to generate and validate messaging. That use is declared explicitly in the onboarding consent, with its data processing agreement. It is not folded into a general policy you have to go looking for.
What happens at the end of a pilot — whether data is deleted, retained, or exported, and whether we can cite your results as a reference — is being written before the first design partner conversation, and will be agreed with you in writing rather than assumed.
You can ask us at any time what we hold about you, ask us to correct it, or ask us to delete it. Write to us and we will answer. As Naros is in early release, the formal response process is being set up with the first design partners; the commitment to answer is in force today.
Naros · legal
These terms govern the design partner programme. Naros is in early release, so they are written to be read, not to hide behind. Where something is not settled, we say so.
Naros is a message governance and attribution tool for B2B go-to-market teams. It governs the message a campaign carries and reports performance by message rather than only by channel. It is provided during the design partner period as described on this site.
Your positioning material and your platform data remain yours. Naros annotates your campaigns with its own metadata and does not modify your business properties. What happens to your data at the end of a pilot is agreed with you in writing before the first conversation, not assumed. See the privacy policy.
Naros is not a source of truth for causation. Attribution assigns credit under a defined, configurable model; causation requires a separate method. What Naros does is make the message observable, so it can become part of a better decision.
Naros is in early release. Some things are not built yet, and the site says which. We do not promise features that do not exist, and we do not claim results we cannot verify. If something on this site turns out to be wrong, we would rather correct it than defend it.
If these terms change, the change will be dated here and, where it affects an active design partner, told to you directly. Continued use after a change means you accept the new terms.