Brand migration

Some results, and the case for migrating your own brands.

Partner workshop · 15 September 2026

Say the one sentence: every brand that moved its site onto the portal this summer earned more, and we can show you why. Then: three parts, the case, which brands, how.

What to expect

Three parts, in this order.

PART ONE The case for migrating What happened to the brands that moved this summer, and how we measured it. PART TWO Where to start Every brand with a site elsewhere is worth moving. Which one you do first, and why. PART THREE How to move a site What the toolkit does, what it asks you, and what it never does without your yes.

The evidence, then the decision, then the mechanics.

This is the map, so give it fifteen seconds and move on. Part one is the numbers and it is the longest. Part two is how to pick a brand. Part three is the how, and that is where the prompts are. Say that the prompts and the figures are all in the takeaway afterwards, so nobody needs to write anything down.

Part one: the case for migrating

Migrated brands: how we measured.

For this we took the traffic and the registrations from a group of brands that have already moved a site onto the portal, and measured what changed.

HowAdded together, not averagedEach brand against its own before, then added up. An average would let the smallest count as much as the biggest.
WhatNo advertising countedSignups that came from an advert are left out of both sides, before and after.
The periodSix weeks each sideThe same number of days before and after go-live.

Three boxes, left to right: how we measured, what we counted, over what period. They are the honesty of the whole thing, so take them in that order. Say the first one slowly. It answers both who these brands were compared against, which is nobody but themselves, and why the headline is not an average. If pushed on why not average: an average gives every brand an equal vote whatever its size, so it would read about seventy per cent where adding them up reads a quarter, and only one of those is a number anyone earned. The middle box stops the obvious objection before it is asked: paid signups are out of both sides, so nothing here was bought. The last box is the like for like, the same number of days each side. If asked how many brands: answer it straight and move on.

Part one: the case for migrating

Every brand that moved took more signups.

+24%more signups you did not pay for

Not one of them went backwards.

One number, and say it slowly: a quarter more signups, and not one brand went backwards. If anyone asks what counts as a signup here: someone who arrived at the site and joined, with no advertising behind them. That covers people who found the site, typed it in, or were sent by an AI assistant. Do not call it organic: the AI ones are inside this number, and they are the next slide. Held back, use only if asked. Payout added together rose 23% over the same weeks. That 23% is every source, paid included, which is why it is the one they can check against their own report; on the signups nobody paid for it is +56%. Lead with 23% and only reach for 56% if someone asks whether the money figure matches the signup figure. The platform-wide comparison is off the slide on purpose: brands built on the portal sit inside that group, so "why am I not seeing this" is a fair question and not one to take cold. If it comes up, say we are measuring brands that moved a site from elsewhere, and offer to check their brand.

Part one: the case for migrating

AI assistants are sending people, and that channel grew fastest.

The extra signups the move brought, with the part an AI assistant sent marked inside them and its own growth beside it The extra signups the move brought Every extra signup the move brought The part an AI assistant sent: about one in ten of them about 1 in 10 the part an AI assistant sent 6.7× nearly seven times as many as before An AI assistant means ChatGPT, Perplexity, Gemini, Copilot or Claude.

A trickle before the move, a stream after it.

The bar is the extra signups the move brought, in other words the 24% from the last slide. About one in ten of them came from an AI assistant, and that is drawn at its real size. The line that matters: this is a slice out of the growth, not a separate channel added to it. If anyone asks about the share of all signups rather than the extra ones, that is the next slide: one in 45. Say the multiple out loud rather than the percentage, it is easier to hear: signups overall up a quarter, the AI ones nearly seven times. It starts from a small base and the next slide says how small. If asked whether AI is just up everywhere. Not on our numbers: no other brand we measured saw an AI source reach even ten signups on either side of the same weeks. Offer that if it is put to you, do not lead with it, and do not explain why, because we have not measured why.

Part one: the case for migrating

Still small, but climbing fast.

Share of the non-paid signups referred by AI assistants, before and after moving 1 in 45 of the signups you did not pay for came from an AI assistant 0.4% 2.2% Before After

Up from about one in 250 before the move.

Grey dot is before, teal is after. This is the honest size of it, and it is the slide that keeps the last one credible. Do not oversell, and do not say it did not exist before: it did, the three brands took 33 AI-referred signups between them in the six weeks before moving. A trickle becoming a stream is the true shape of it. Two things to have ready. One in 45 of what: of the signups with no advertising behind them, not of every signup a brand takes, so a brand running adverts would see a smaller share of its total. And why not the same number as the last slide: seven times is the count of AI signups, the share only rose five-fold because the total grew a quarter underneath it. If anyone asks why the last slide said seven times and this one does not: seven times is the count of AI signups, and the share only rose five-fold because the total grew a quarter underneath it. Do not put that on a slide, it is an answer and not a message. If asked whether they pay better: we have not measured that, the numbers are small and they point both ways. Say so and move on.

Part one: the case for migrating

Every brand that moved took more initials.

+44%more initials in the same six weeks*

* Measured across all signup channels for the brands in the study.

Initials are the first purchase a new member makes. Added together they rose nearly a half, and all three brands are up, which is why the heading says every brand and not a total. This is the one figure they can check line for line: it is the Initial Sales column on their own campaign report over the same six weeks, and it matches to the penny. Invite the check. Do not say each signup is worth more. Initials per signup rise added together and fall on two of the three brands, so it is not a claim that survives being looked at. If someone says their revenue per user fell. It does for a while, and their own report will show it falling further than ours, because that report counts every member right up to today, so the ones who joined earlier have had far longer to earn. The signups arrive at once and their revenue arrives over months. That is arithmetic, not decay, and the total is what got bigger. If asked whether the 44% is only the traffic they did not pay for. It is not, and answer it straight: this is the Initial Sales column on their own report with every source in it, which is exactly why they can check it. On the signups nobody paid for the same figure is +46%, so the honest answer costs us nothing. Held back, use only if asked. Payout added together rose 23%, members who paid within a fortnight rose 58%, the revenue from those signups rose 62%, all of them every source on the same basis. Take out the brand that grew most and initials are still up 24%.

Part one: the case for migrating

Who moved, what changed, and what we think did it.

The brands that moved
  • Established brands, years of standing behind the domain.
  • Already earning traffic they never paid for.
  • Small sites and large ones alike.
What we saw
  • AI assistants started sending people.
  • Organic signups rose as well.
  • Rankings moved too, though rankings are not what these numbers count.
What we think did it
  • Structured data on every page.
  • Pages laid out so an agent can crawl and read them.
  • Content that stays fresh.
Which one did the work is hard to pick apart. Covering all of them is the point.

Left to right: who these brands were, what changed for them, and what we think did it. The middle box is measured, the right one is not, and saying which is which is the persuasion. If asked what organic means here. Signups that arrived with no advertising tag on them. Our only label is the campaign tag, so somebody sent by a search engine and somebody who typed the address in look the same to us and are counted together. Do not split them out, we cannot. On rankings. We did not measure ranking positions, so do not quote one; if someone asks, say we have seen rankings move on brands we have looked at since, and that what these numbers count is people arriving and joining. On the domain. Years of standing is the profile, not a score we measured, so do not put a number on it. Do not credit any single feature, and do not say it was the schema. If someone's brand is new or small, say it plainly: this is simply where a move shows up fastest and largest, not the only place it is worth doing. A newer brand still belongs on the portal, it just has less old ground to make up.

Part two: where to start

Where to start.

1Start with the brand that already earns you unpaid traffic.
2Find the brand in the portal.It is already there, even though its site is served from somewhere else.
3Point the toolkit at the old site, and it does the move into that brand.

The headroom is in whatever a crawler could not read: a thin site, a bloated one, or markup that never described the pages.

Say the framing before the steps: every brand with a site elsewhere belongs on the portal eventually, so this is an order to work in, not a shortlist to qualify for. Step two is the one that surprises people, so land it: the brand already exists in the portal whatever is serving its website today, so there is nothing to create and nowhere new to sign up. Step three is the whole of part three, which is next. On the caption, all three shapes are headroom for the same reason: a crawler could not read the pages properly, so nothing sent anyone to them.

Part three: how to move a site

Nothing goes live that you have not seen.

One line starts it
"Migrate my site, it is at your address." It reads the site, then comes back with a plan: which addresses come across, which you leave. Nothing is built before you agree it.
You work through it together
It groups the pages by the patterns it found on the site and takes a set at a time. You see one designed and three filled before it builds the rest.
Preview it on a temporary address
The whole thing goes up on the portal on an address only you are using. Your live site carries on serving visitors the entire time.

This is the slide that answers "what if it builds something I did not want". One line starts it, you agree the plan before anything is built, you see a set before it fills the set, and the whole thing sits on a temporary address until you say otherwise. Land the last box hardest. Nothing at the old site changes, not one visitor is redirected, until the DNS is switched, and that is a separate decision taken later. Held back, use if asked. When it asks to build a set it names the number, so "shall I build all 162" is the question and a yes to one page builds one page. Anything it would replace is backed up first.

Part three: how to move a site

What comes with you, and what you can add.

  • Your site structure comes across. No page drops off the internet. Most are a direct replacement, and where an address has to change a 301 sends the old one to the new.
  • Titles, descriptions and structured data travel with the page. And the new pages add what was never there: the schema a page was missing, the description it never had. You can change any of it afterwards.
  • Once it is here, the pages can use your own members. Their interests and demographics, your polls and your reviews. Real words, which no competitor has.

The plan brings the whole estate across, and changes what needs changing on the way.

Three ticks, and the story is carry and add, not carry alone. First tick is the one people worry about: the address structure comes with you and no page drops off the internet, because the address is what rankings and links attach to. Most pages are a straight replacement; where one genuinely has to change, a redirect sends the old address to the new, and that mapping is part of the plan you agree. Second tick is where the back-to-basics argument from part one lands: the page data travels, and the gaps get filled, so a page that never had structured data or a description gets one. Say plainly that it is theirs to change afterwards. Third tick is what only the portal can do. Held back, use only if asked. Member numbers are never written on a page, exact or rounded, and the rule is built in. Menus become platform menus, both halves, with no hand-editing when one changes. The reading is polite: it obeys robots.txt and takes one page at a time. If asked about rankings: we keep the addresses and the page data, which is what rankings attach to; we do not promise a ranking.

Part three: how to move a site

Audit any site today, before you move anything.

audit the site at https://www.example.com

What it reads

  • Whether search engines and AI assistants are allowed to read the site at all.
  • Robots and sitemaps, addresses that redirect, and pages nothing links to.
  • Titles, descriptions, structured data and share cards, page by page.
  • Broken links, heavy images, and copy that never got finished: placeholders, filler text, template leaks.

What comes back

  • An actionable report the agent can then help you work through.
  • Grouped by what it costs you: now, worth fixing, or fine to leave.
  • A spreadsheet listing every page each finding applies to.
  • It knows whether the brand is on the portal or not, so it runs on sites you have not moved yet.

Point it at any address, on the portal or not, and it reads the pages the way a crawler does.

This is the one thing everybody in the room can do this afternoon, whether or not they ever move anything, and it is the easiest yes in the deck. Land the "on the portal or not". It will read a site that is still hosted somewhere else, from the sitemap, with nothing to install, so a partner can find out what is wrong with a site before deciding whether to move it. On a brand already on the portal it does more, because it can also cross-check what is served against the page list. Left card is what it looks at, right card is what you get back. The first line of the left card is the one to say slowly. If an AI assistant is blocked at the front door, none of what we saw on the AI slide can happen for that brand, and that is a thing they can find out today. Do not promise any fix is automatic: it offers each one and waits. Do not quote a running time either; it depends entirely on how many pages there are.

Part three: how to move a site

What to say.

Migrate www.example.com onto my brand

That is the whole of it. It leads from there, telling you what it found and asking what it needs.

* Two more worth knowing: "what version is this?" and "upgrade my toolkit".

One line, and it is the only thing anybody has to remember. Name the brand in it: without a brand named the toolkit asks which one rather than guessing, which is safe but slower. Then stop talking about prompts. The follow-ups came off this slide on purpose, because it leads the conversation: it tells them what it found and asks what it needs, so listing questions they might ask implies homework that does not exist. If someone wants the detail, the guide is on the slide and it carries every stage with what a good answer back looks like.

Next actions

What to do next.

1Pick one brand.The one already earning traffic you never paid for, on the site nobody has had a reason to touch in years.
2Audit it, if you want to.Worth doing, not required. One prompt shows you what a crawler cannot read today, and it works whether or not the brand is on the portal. Skip it and the move turns up the same things as it goes.
3Open the guide, then start the move.Say "migrate my site" with its address. You need nothing else ready: it decides nothing before it has read the site, and you read the plan before you agree to any of it.

We are here if you get stuck. The feedback button is on every one of these pages.

Three actions, then the links. They run in order and each one is doable today: pick a brand, audit it, open the guide and start it. The guide is the first link and it went live this morning, so point at it and say it carries the prompts and the checklists. Step two is optional and say so. Nobody should read it as a gate: a partner who wants to start migrating today should start, and the move surfaces the same problems as it works through the site. The audit is the gentler way in for anyone not ready to commit. Do not offer to do the picking for them. We are here to help if they get stuck, not to take a list of brands away and come back with an order. Held back, use if asked what they need ready. The address of the site, the brand on their account, and the toolkit connected. Nothing else, because it decides nothing before it has read the site. Do not put a time on it: a big site takes a while and works through steadily, so say it depends on the size and that they are not sitting over it. Close on the feedback line.

Brand migration  ·  partner workshop, 15 September 2026
01 / 14