The short answer
Score each acquired brand on how many people search for it by name, how much its queries overlap a sister site, and who buys from it. Freeze a 90-day Search Console baseline before any redirect, then read the old domain and the receiving pages together on day 7, day 30 and day 90. Move one brand at a time. The same setup doubles as marketing analytics across brands: every domain's Search Console and GA4 numbers, ranked against each other.
- A brand people search for by name usually earns its own domain; a brand nobody searches for is a merge candidate
- Measure the old domain and the receiving pages together, in clicks per week, so ranges of different length compare fairly
- Check sister brands' top 100 queries against each other before deciding which one should own a topic
Keep, merge or redirect: deciding each acquired site's future
Every brand you buy arrives with a domain, years of rankings and a team that likes its logo. Fewer sites is an easy finance case and a harder search one. Google calls merging domains a site move, and its guidance on moving a site with URL changes says visibility "may fluctuate temporarily during the move" (checked November 2026).
So give each acquired domain one of three outcomes, picked from evidence rather than the brand team's mood or the CFO's spreadsheet.
- Keep: the brand stays on its own domain and keeps its own site.
- Merge: its content moves page by page into a parent-site section, each old URL redirecting to its twin.
- Redirect: the brand retires. Only old URLs with a real equivalent on the parent redirect to it. Pages with no equivalent return 404 or 410; they are not funneled into a category or the home page.
| Signal | Keep the domain | Merge into parent | Redirect and retire |
|---|---|---|---|
| Searches for the brand name | A large share of its clicks | A modest share | Almost none |
| Query overlap with a sister site | Little; it answers its own searches | Heavy, but its pages are better | Heavy, and its pages add nothing |
| Who buys from it | A different buyer or channel | The parent's buyers | The parent's buyers |
| Contracts, dealers or listings tied to the name | Keep, renaming costs more than hosting | Partly, phase the name out | None worth protecting |
| Content worth saving | Yes, where it already ranks | Yes, moved one URL at a time | Little; redirect true equivalents, let the rest return 404/410 |
| What to watch afterwards | Its own click trend | Receiving pages vs baseline | Old domain clicks fading out |
A working framework, not a Google rule. The first two rows come from Search Console; the rest come from your commercial teams and deal files.
Baseline before anyone touches a redirect
Once redirects go live, the old domain's numbers start to drain, and any baseline taken afterwards is polluted. Take it before.
DuckDive turns the baseline into a question instead of a project. One Google sign-in, with the login that already sees every brand, and each acquired domain becomes its own website in DuckDive with its Search Console and Analytics history imported and stored. From then on, whichever assistant your team uses, Claude, ChatGPT or Cursor among them, pulls the before and after numbers for one brand or the whole group without a spreadsheet per domain.
Ask your agentBefore we redirect Corbel Flooring's domain, give me a baseline: its top 100 Search Console pages and queries by clicks over the last 90 days.
Tools it calls list_websites get_search_console_pages get_search_console_keywords get_search_console_summary
The pages list sets your redirect map's priority: the URLs with the most clicks get a hand-checked destination first. Google's advice is to map old URLs to their new equivalents and let removed pages return 404 or 410. A high-click page with no twin on the parent is a build task before go-live. The queries list tells you which searches the receiving pages must still answer.
Freeze these before go-live
- Save the agent's answer with the date. Agent ranges roll forward daily, so "last 90 days" will never mean these 90 days again.
- Export the same data from Search Console yourself. Attach it to the migration ticket, so the before picture lives outside any tool.
- Write down the go-live date. Every checkpoint counts from it.
- Note what counts as a conversion on both GA4 properties. A quote form on the brand site and a showroom booking on the parent are not the same event.
- Confirm Search Console ownership of both domains. Google's Change of Address tool requires you to be an owner of both the old and the new property.
- Agree a pass mark with the brand team. Decide now how close to baseline the day 90 number must be before the next brand moves.
Checkpoints on day 7, day 30 and day 90
Three reads tell a normal migration from a bad one. Each adds two numbers: clicks still landing on the old domain, and clicks on the parent's receiving pages. Alone, either misleads; the old domain always falls and the receiving pages always rise. Only the sum shows whether searchers followed.
- Receiving pages
- The parent-site URLs that old brand URLs now redirect to, for example everything under /flooring/.
- Clicks per week
- A range's clicks divided by its length in weeks (30 days is 4.3 weeks). It lets a 7-day read sit beside a 90-day baseline.
- Click
- Search Console counts a click when someone clicks through from a Google Search result to your site.
- Average position
- Google's average of the topmost position your site held in results, for the query or page in that row, per the Performance report help.
Day 7: is anything broken?
Expect a dip. Google says that for medium-sized websites it can take "a few weeks or more" before it gradually shows the new URLs, "and for larger sites, even longer". Ask for the last 7 days to catch mechanical failures, such as a receiving page with zero clicks, which can mean a redirect lands on the wrong page or a 404. Save it dated; by day 8, "last 7 days" no longer covers launch week.
Day 30: is it recovering?
On day 30, "last 30 days" starts on go-live day, so it is the whole first month. The combined number should be climbing toward baseline. High old-domain clicks mean Google has not processed many redirects yet.
Day 90: did it hold?
This read gates the next brand, so it must show where traffic settled. Ask for the last 30 days again, days 61 to 90, in clicks per week. A last-90-days figure would average in the launch-week dip and understate recovery; keep it only as the cumulative cost of the move. Google's Change of Address help suggests moving sites one at a time and waiting "till traffic stabilizes before moving the next site", and says combining several sites at once "can cause some confusion and traffic loss" (checked November 2026).
When a brand team asks how it is going between checkpoints, a mid-course read does the job:
Ask your agentSince the Corbel redirect went live three weeks ago, how do Search Console clicks to the parent site's flooring pages compare with Corbel's old weekly clicks?
Tools it calls list_websites get_search_console_pages get_search_console_summary
Worked example: folding Corbel Flooring into the parent site
Ironleaf Building Products bought Corbel Flooring and merged it: Corbel's pages were rebuilt under a new /flooring/ section on the Ironleaf site, each old URL redirecting to its match. The team saved the day 7 and day 30 answers on the migration ticket. Here is the day 90 read, with those two pasted in.
It's day 90 since the Corbel redirects went live. Our saved day 7 and day 30 reads are pasted below. Add today's read from the last 30 days, put all three against the baseline in clicks per week, and tell me if we passed our rule of staying within 10% of baseline.
list_websitesfinds the Ironleaf and Corbel Flooring websitesget_search_console_pagesIronleaf, last 30 and last 90 days, keeping only URLs under /flooring/get_search_console_summaryCorbel's old domain, clicks for the last 30 and last 90 daysget_search_console_pagesCorbel's old domain, last 30 days, old URLs still clickedget_ga4_landing_pagesIronleaf, last 30 days, conversion rate on the /flooring/ landing pages
Passed. Over days 61 to 90 the two sites drew 1,178 clicks a week together, 3% under the baseline of 1,213 (15,600 clicks in the 90 days before go-live) and inside your 10% rule. Clicks per week; the day 7 and day 30 rows are from your saved answers.
| Checkpoint | /flooring/ | Combined |
|---|---|---|
| Baseline | 0 | 1,213 |
| Day 7, last 7 days | 298 | 960 (-21%) |
| Day 30, last 30 days | 854 | 1,078 (-11%) |
| Day 90, last 30 days | 1,145 | 1,178 (-3%) |
Across all 90 days the two sites averaged 1,135 a week (-6%), about 1,000 clicks fewer than the baseline period. That is what the move cost, not where it ended.
The old domain still drew 140 clicks in the last 30 days, and 118 of them went to one URL: /installation-cost-calculator. It was mapped to the /flooring/ category page, which has no calculator, so searchers land somewhere that does not answer them. Rebuild the calculator on Ironleaf and point the old URL at it, or let the old URL return 404 or 410.
Over the last 30 days the /flooring/ landing pages converted at 2.0%, against 2.2% on Corbel before the move. Confirm Corbel's quote form event was recreated on Ironleaf before reading into the gap.
The calculator is the textbook case Google's site-move guide warns about: sending old URLs to one destination that does not match them "might be treated as a soft 404 error". Its Change of Address help adds that old URLs without an equivalent page on the new site can keep appearing in results. One more caution: filtering by path only works while the flooring URLs sit inside the parent's top 100 pages; on a very large parent site, check the long tail in Search Console.
Ask your agentWhich old Corbel Flooring URLs are still getting Search Console clicks in the last 30 days, and what does each one redirect to on the Ironleaf site?
Tools it calls list_websites get_search_console_pages
The agent cannot see your redirect rules, so paste the redirect map into the chat and let it match the two.
Freeze the baseline for your next brand move
Connect the brand you plan to fold in first, and have its 90-day pages and queries saved before the redirect ticket is even written.
Free for 2 sites · No credit card
Sister brands competing for the same searches
Two brands you kept can still chase the same queries from separate domains. Neither team notices, because each watches only its own Search Console. The group pays twice for content that splits the clicks, and often neither page reaches the top three.
There is a policy angle too. Google's spam policies describe doorway abuse as sites or pages "created to rank for specific, similar search queries", and give "multiple websites with slight variations to the URL and home page" as an example (checked November 2026). Real brands with distinct products are not that; near-identical sister pages built to take two slots for one query drift toward it.
The test: intersect each brand's top 100 queries by clicks for the last 90 days. A few shared queries are normal. A cluster on one topic means two teams write for the same searcher.
Ask your agentList the queries that appear in the top 100 Search Console queries of both Pallister Doors and Westmere Cabinetry over the last 90 days, with each brand's clicks and average position.
Tools it calls list_websites get_search_console_keywords
With more brands, ask for every pair. Four brands make six pairs; ten make forty-five. Tedious by hand, quick for an agent, since each brand's query list is one call.
Worked example: two brands splitting the clicks for one query set
Ironleaf kept Pallister Doors and Westmere Cabinetry as separate brands, and both publish buying guides.
Where do Pallister and Westmere compete in search? Shared queries from both top 100 lists, last 90 days, with clicks and average position for each.
list_websitesfinds both brand websitesget_search_console_keywordsPallister Doors, top 100 queries, last 90 daysget_search_console_keywordsWestmere Cabinetry, top 100 queries, last 90 days
Seven queries appear on both lists. They drew 1,595 clicks for Pallister and 1,445 for Westmere. Three are true splits: both brands sit between positions 4 and 9 and neither leads. Cells show clicks, with average position in brackets.
| Query | Pallister | Westmere |
|---|---|---|
| shaker cabinet doors | 140 (7.8) | 610 (2.4) |
| replacement cabinet doors | 220 (5.1) | 260 (4.6) |
| glass cabinet doors | 95 (8.9) | 180 (6.7) |
| cabinet door hinges | 160 (6.4) | 150 (6.9) |
| pantry door ideas | 390 (3.2) | 85 (9.4) |
| barn door for pantry | 310 (2.9) | 40 (11.2) |
| custom closet doors | 280 (4.0) | 120 (7.5) |
Every split query is about cabinet doors, Westmere's product. Give Westmere that topic: Pallister's cabinet door guide redirects to Westmere's equivalent or narrows to what Pallister sells. Pantry and closet door queries stay with Pallister.
Before acting on a split, check two things the query list cannot show. Which page ranks: the query list does not say, so check it in Search Console (the FAQ below shows how). And which brand converts: compare GA4 landing page conversion rates for the pages involved. A position lead matters less than a filled-in form.
Comparing brands that live on separate domains
Adding every brand's sessions into one group total hides what the board asks: which brands are getting stronger. Separate GA4 properties and Search Console sites mean Google offers no shared view. The usual workarounds are re-tagging every site into a shared property or paying for a roll-up; for what each route costs, see side-by-side GA4 numbers for brands on separate properties.
One comparison does most of the work in the keep, merge or redirect call: branded search share, meaning clicks from queries containing the brand's own name as a share of clicks in its top 100 queries. A brand customers ask for by name has equity a redirect would throw away.
Ask your agentQuarterly brand read for Ironleaf, Pallister, Westmere and Corbel: Search Console clicks for the last 90 days, the share of clicks in each site's top 100 queries that contain the brand's own name, and GA4 conversion rate. Rank them by branded share.
Tools it calls list_websites get_search_console_summary get_search_console_keywords get_ga4_overview
| Brand site | Clicks, 90 days | Branded share |
|---|---|---|
| Pallister Doors | 21,700 | 58% |
| Ironleaf (parent) | 48,200 | 41% |
| Westmere Cabinetry | 30,400 | 22% |
| Corbel Flooring | 15,600 | 9% |
Pallister's buyers ask for it by name, so it kept its domain. Corbel's traffic came almost entirely from generic flooring searches the parent could answer, which put it on the merge list. Westmere sits in between and gets the next look. Compare conversion rates within a brand over time, since each brand defines its own key events.
Watch out
Branded share is counted inside each site's top 100 queries, the most a query list returns. On big sites the missing long tail tends to be generic, so the share reads high. Compare it brand to brand and quarter to quarter, not as an exact figure.
If a brand you did not touch falls mid-migration, do not blame the move yet. Check whether the drop is hitting every brand in the group or only one before anyone reverses a redirect.
Connecting every brand before the first move
Connect every brand, not only the one moving.
-
Get one Google login onto every brand
Each brand's team adds the group login to its GA4 property and Search Console site. For a Change of Address move, that login must own both Search Console properties.
-
Sign in to DuckDive, or hand setup to your agent
Use Continue with Google with that login, or connect DuckDive in your AI app and paste the setup prompt; you click Continue with Google once. Adding the connector by hand instead? The setting hides in a different place in every AI app; search its settings for "MCP" or "Connectors".
Paste this into your agentSet up DuckDive for my website. Follow https://analytics.duckdive.ai/
agents.md -
Add each brand domain as its own website
The parent and each brand get one website each. The agent suggests sites from your Google account and asks when there are several. Whoever connects needs an admin role and must leave the actions box on during sign-in, or the agent can read numbers but cannot add a single brand.
-
Link each brand's Search Console site and GA4 property
A brand with its own paid search or Local Services Ads spend gets those two accounts attached to its website too; each link starts that source's history import. A Search Console site can sit under only one DuckDive website, and so can a GA4 property. After a merge, the agent will move one across to the parent only once you say yes; bulk onboarding with an agent covers how that hand-off works.
-
Check the import, then freeze the first baseline
Imports normally finish within minutes, longer for brands with years of history. If numbers look empty, ask the agent for sync status, then run the baseline for the first brand on your list.
More detail on what the Search Console tools return, row limits included, is on the Search Console connector page. Would rather not set up an agent? Ask Drake, the assistant inside DuckDive.
Brand teams' worries about consolidation
The people who built a brand will push back. Most objections settle with a number agreed before go-live.
"Our customers search for us by name"
Then branded share will show it, and the matrix says keep. If you merge anyway, check after the move that searches for the old name now reach the parent's own query list.
Ask your agentAre people who search for Corbel by name landing on Ironleaf yet? Show me any queries containing "corbel" in Ironleaf's top 100 Search Console queries for the last 30 days, with clicks and position.
Tools it calls list_websites get_search_console_keywords
"We will lose rankings we spent years building"
Some dip is normal while Google catches up, and Google says 301 and other permanent redirects "don't cause a loss in PageRank". Show the team the pass mark, the day 30 trend, and that their section is reported as its own line, not buried in the parent's totals.
"Our leads come from ads, not search"
Ad accounts don't move with a domain; their landing pages do. Link the brand's Google Ads account and compare its account summary before and after the final URLs change.
Jobs DuckDive leaves to other tools
Redirects and Change of Address are yours to build and file. Plan on these from elsewhere too:
- Crawling. Redirect chains, 404s and canonical tags need a site crawler. DuckDive sees clicks, not HTTP status codes.
- Indexing and sitemaps. Google points migrators at Search Console's Sitemaps and indexing reports, which DuckDive's tools do not read.
- Backlinks. Who links to the old domain is a job for a link research tool.
- Query-to-page pairs and long lists. Queries and pages come back as separate lists of up to 100 rows each.
- Long before-and-after windows. Agent questions stop at a 12-month window, while Search Console's own Performance report has held 16 months of data since Google rebuilt it in 2018. For a brand moved two years ago, your saved baselines are the only record.
- Revenue. Whether a merged brand still sells lives in your ERP or CRM.
Still buying brands? Checking a seller's traffic before close is covered under vetting a website you plan to buy. Fund-level monitoring lives on the private equity page, and paid-search reliance across companies on the portfolios overview.
Before you sign off on a brand's domain
Does DuckDive set up redirects or file the Change of Address request?
No. Your developers build the redirects and you file Change of Address in Search Console. DuckDive reads the data that shows whether the move worked: Search Console clicks, queries and pages for both domains, plus GA4 sessions and conversions.
How far back can the agent compare before and after a move?
Twelve months back is as far as an agent's question reaches, and Search Console itself keeps 16 months. Save the dated baseline answer and your own Search Console export before go-live, so the before picture survives as the ranges roll on.
Should we remove a retired brand's website from DuckDive after the move?
Not before the day 90 checkpoint, because its fading clicks are half of every read. After that, keep your own exports of its baseline rather than relying on any tool to hold a removed site's history.
Can the agent tell me which page ranks for each overlapping query?
Not directly. The Search Console tools return top queries and top pages as separate lists. For a query both brands share, open the Performance report in Search Console, filter by that query and check the pages tab on each site.
Two of our brands already share one GA4 property. Can DuckDive split them?
Only partly. DuckDive reads a GA4 property as a whole. If each brand lives under its own path, the agent can group landing pages by path to compare them. If they share URLs, it cannot separate them.
Is it safe to point an AI agent at every brand's analytics?
An agent sees only the websites its user's DuckDive account can already see, under that user's role, so a brand manager sees sister brands' numbers only if their account includes those websites. The link reads data. Adding brands or linking sources works only for admins who left actions on at sign-in; deleting data is not something the connector can do at all. Conversations with the agent never reach DuckDive; the AI vendor's own terms govern the tool results it receives.
Sources
- How to move a site, Google Search Central (site moves with URL changes), checked November 16, 2026.
- Change of Address tool, Search Console Help (ownership rule and moving sites one at a time), checked November 16, 2026.
- Performance report (Search results): Overview and basic setup, Search Console Help (metric definitions), checked November 16, 2026.
- Spam policies for Google web search, Google Search Central (doorway abuse section), checked November 16, 2026.
- Introducing the new Search Console, Google Search Central Blog (announcement of 16 months of Performance data), January 8, 2018, checked November 16, 2026.
Last reviewed . Numbers in EXAMPLE DATA blocks are fictional.
