Maggie Beer Holdings has been all over the news this week for a $9.5 million loss caused by a "Google blunder".
Both halves of that are wrong. Google didn't do anything, and the migration didn't cost $9.5 million.
The real story is more useful than either, so here it is.
What the $9.5m actually is
Three separate numbers have been fused into one headline.
The $9.5 million is Maggie Beer Holdings' statutory loss after tax for FY26. It covers the entire group, not just search, and it actually halved from $24.3 million the year before.
Inside it sits $5.1 million of non-cash impairments. An impairment is a company accepting that something on its books is worth less than the number written down. MBH bought Hampers & Gifts Australia for $40 million in 2021 and is now progressing a sale at around $10 million. They cut the recorded value to match, and that cut shows up as a loss. No money left the bank. Most of that gap is a 2021 acquisition price, not last Christmas.
Strip the impairments out and the operating picture barely moved. Trading EBITDA loss narrowed to about $500,000 from $700,000. Gross margin improved 1.6 points to 49%. Group net sales were flat at $75.7 million.
The part you can put on the migration is HGA net sales falling 6.3% to $41.4 million, so roughly $2.8 million of revenue, plus $800,000 of emergency advertising. Bad. Not $9.5 million.
What did happen
MBH's own reporting is the strongest evidence here, and it's damning enough without help.
Hampers & Gifts Australia moved its two sites, The Hamper Emporium and Gifts Australia, to Shopify in June 2025. The company says it lost up to 38% of its Google visits by January. It spent around $800,000 on advertising trying to buy that traffic back, in the most expensive weeks of the year, and HGA still finished down 6.3%.
June, for a business that makes its money at Christmas.
URLs stopped existing
The mechanism isn't mysterious. The old URL structure wasn't fully redirected.
Right now 878 referring domains point at 404 pages on giftsaustralia.com.au, and another 644 on thehamperemporium.com.au. Roughly 1,500 domains worth of links somebody earned over years, landing on error pages. The dead URLs follow the old site's patterns exactly: product pages like /t2-teatime-gems-teacup-gift-set/T145AK897, categories like /occasions/christmas/hamper and /valentines-day.
That's what was riding on the Christmas category tree the December before the replatform. Four pages, each in the same range, on the exact terms a gifting business needs in November and December.
A redirect map is not glamorous work. It's a spreadsheet. It is also the single highest value thing anyone produces during a replatform, and it has to be signed off before the new site goes live, not built in a panic when someone notices traffic is down.
Migrations don't fail on launch day
This is the part that catches people, and it's why "the site went live fine" is a useless test.
Google needs time to recrawl a site of that size. When URLs vanish without redirects, the losses arrive as a slow drip over eight to twelve weeks rather than overnight. Both HGA sites looked healthy through July. The company's own account has the damage clear by January.
Which means the window to find and fix this closed while everything still looked fine. By the time it was obvious, they were inside their peak trading period with no runway.
They had already done this once
Here's the part nobody has reported.
MBH ran the same play eight months earlier. Maggie Beer Products, the pâté and verjuice business at maggiebeer.com.au, migrated to Shopify in October 2024. In the four months after, the number of pages earning it a Google ranking fell from around 2,728 to 1,451. That's a 47% drop, or about 36% once you strip out what comparable Australian retail sites did over the same months.
So why did nobody raise it before running the same migration on the gifting sites?
Because the number everyone watches didn't move. Organic traffic to maggiebeer.com.au was roughly 44,000 estimated visits in December 2024 and 42,000 in December 2025. That business sells through supermarkets and wholesale, so its search volume sits on a handful of brand and recipe terms that held up fine. MBP net sales were up 8.2% to $34.3 million in FY26.
A third of the ranking pages gone. Sessions flat. Eight months of nobody looking.
Then the same migration went onto the two sites where organic search is close to the whole business.
A note on the data
Worth being upfront about something, because anyone pulling these numbers themselves will hit it.
Ahrefs changed how it indexes keywords around October 2025. Run a control site with no migration at all and you'll see its long-tail keywords fall 90% in the same window. That artefact sits on top of every site in the database, and it overlaps exactly with the period the HGA sites were declining.
So I'm not quoting an Ahrefs traffic figure for the HGA sites, because I can't separate the migration from the tool. The 38% is MBH's own analytics and it's better evidence anyway. The Maggie Beer Products numbers above are usable because that migration and its aftermath both sit a year before the artefact, and I've adjusted them against a control basket regardless.
If you're benchmarking anything across late 2025, check a control site first.
Four things that would have prevented this
Map every URL before launch
Every indexed page on the old site gets a relevant destination on the new one. Not the homepage. If a page has rankings or links it needs somewhere to go, and legacy URLs count, because they still hold links.
Set the real checkpoint at week eight
Launch week tells you the site works. Week eight tells you whether Google agrees. Put it in the calendar before you go live, and keep looking at week twelve.
Watch ranking pages, not sessions
Maggie Beer Products is the proof. A third of its ranking pages disappeared and traffic told them nothing for eight months. Sessions are noisy and seasonal enough to hide a serious problem for the better part of a year.
Don't replatform into your peak
A gifting business should be doing this in February. You want three quarters of runway to find and fix things before the money months.
We run a lot of migrations. They're hard work and mostly unglamorous, and the most valuable deliverable is never the new design.
I don't know who built these sites and this isn't about naming anyone. Replatforming is genuinely hard, and the technical work of a Shopify build is a different discipline to protecting search visibility. From the outside it reads as an ownership gap. Nobody owned the search outcome, so nobody was watching the numbers that would have caught it.
Maggie Beer herself had nothing to do with any of it. She sits on the board of the holding company. Her name is in the headline because it's the name on the ticker.
If you've got a replatform coming up, get an SEO in the room before the build starts. It's a rounding error against $800,000 of emergency ad spend and a Christmas you can't get back.
Financial figures are from Maggie Beer Holdings' FY26 results and annual report. Ranking and backlink data is from Ahrefs, with the caveat noted above.