Don’t Change What Isn’t Broken: What a $6 Million Website Upgrade Teaches Every Marketing Team
A Melbourne skincare brand turned over $6 million a year at its peak. Then it upgraded its website.
7NEWS reports that founder Briony Kennedy moved Adorn Cosmetics from its existing website to a new ecommerce platform, on quoted revenue increases of between 34 and 94%.
The day the new site launched, she says, sales dropped by about 30%. The business was spending around $1 million a year on Instagram and Facebook ads, and those ads were no longer linking back to the website.
She laid off staff. She sold the family home to keep the company alive. She moved platforms a second time. According to the report, it was too late, and she now faces bankruptcy.
Her advice to other business owners is one line.
“No matter what is dangled in front of you, really question if things are working. Don’t chase the new shiny thing.”

We do not know what happened inside that project, and we are not going to guess. There is a dispute, and only one side of it has been told publicly.
But the pattern she describes is one we see every month. A website that was working gets replaced, and the replacement costs more than the build.
This article is about that pattern. Why it happens, what actually breaks, when a rebuild is the right call, and what to check before you sign anything.
Why a working website gets replaced
It usually starts with a fair concern. Leads are down. The site looks tired. Somebody is worried about security. A vendor or an agency shows a forecast with a big number on it.
A rebuild is also the easiest thing in our industry to sell. It has a start, an end, a before and after, and a figure on a proposal. Everyone gets to look at new designs in a workshop.
Fixing what you have is the opposite. It is unglamorous and hard to photograph. It means taking two seconds off a page template, repairing tracking that has been broken since March, and rewriting a form that already exists.
It also tends to produce a better commercial result for less money and far less risk.
Web design is not the enemy. Design matters a great deal. But a rebuild is a solution, and most businesses buy it before they have diagnosed the problem.
What actually breaks in a migration
A rebuild or a platform move changes almost everything at once. Each item below is small. Any one of them can cost you sales on launch day.
Your URLs. Every page address that changes needs a redirect from the old one. Miss them and your Google rankings, your backlinks and every bookmarked link point at a dead page.
Your ad links. Every paid ad points at a landing page. If that page moves or disappears, you keep paying for the click and the visitor lands nowhere. This is the failure described in the 7NEWS report, and it is one of the most expensive because the spend carries on while the sales stop.
Your tracking. Analytics, conversion events and ad pixels are wired into the old site. They rarely survive a migration untouched. When they break, your ad platforms lose the signal they use to find buyers, and your reports stop telling the truth at the exact moment you need them.
Your speed. New sites are often slower than the ones they replace. More scripts, heavier pages, bigger images. Slower pages convert less and rank lower.
Your checkout and forms. The path from “interested” to “paid” is the most valuable part of the site. A new one is untested with real customers until launch day.
Your search history. Google has years of trust built up in your existing pages. A migration asks it to learn the site again. Handled well, the dip is small and short. Handled badly, it is neither.
Slow website wasting your marketing spend?
- Uncover performance issues
- Identify SEO opportunities
- Security gaps, and quick wins
None of this is rare. It is the normal risk of changing a website, and it is why the question “should we rebuild?” deserves more than a design workshop.
Work with someone who gets marketing, and who cares
Most website migrations are run as a technical project. The pages get moved, someone checks that they load, and the site goes live.
But the things that break are marketing things. Ad links, tracking, landing pages, search rankings, the path to an enquiry. A team can deliver exactly what was scoped and still switch off your revenue, because nobody on the project knew which pages earn the money or which campaigns point where.
So if you do have to migrate, do it with someone who understands your marketing and cares what happens after launch. That means every live campaign mapped to its new landing page. Every conversion event tested before and after go-live. Search rankings watched daily for the first month.
A simple test: ask who will be checking your ads and your leads the morning after launch. If nobody has an answer, the project is not ready.
Five questions to ask before anyone opens a design file
Ask these in order, and be honest about the answers.
1. What is actually broken?
Write the list. Not “it looks dated”. Specific faults: this page is slow, this form loses people, this cannot be edited without a developer. If the list is short, fix the list.
2. Is the site slow for real visitors?
Not one test on a fast laptop. Google Search Console shows what real people on real phones experienced over the last 28 days. That is the number that counts.
3. Is your tracking telling the truth?
If your analytics is not recording the actions that matter, you are not making a decision. You are guessing with a dashboard open. We find broken or partial tracking on most of the sites we audit.
4. Where do people give up?
Session recordings and heatmaps show the exact moments visitors leave. It is usually a handful of fixable points, not the whole site.
5. Which pages already earn?
Almost every site has a small number of pages carrying most of the value. If those pages convert badly, that is your project. Rebuilding the other ninety changes nothing.
If you cannot answer all five, you are not ready to commission a rebuild. You are ready for a diagnosis.
When a rebuild is the right call
We are not against rebuilds. We do them, and sometimes they are exactly what a business needs.
Rebuild when:
- The platform is obsolete or unsupported. If it no longer gets security updates, staying is the bigger risk.
- The codebase is unsafe or cannot be maintained. When every small change breaks something else, the site is costing you more than it looks.
- The site is so clunky your team cannot use it. If marketing cannot publish a page without a developer, the site is holding the business back.
- The business has changed. A new brand, a new model or a new product line the old structure cannot carry.
- The list of faults is longer than the list of what works.
The difference is how the decision gets made. A good rebuild starts from evidence: a diagnosis that shows the faults cannot be fixed in place. A bad one starts from fatigue with how the homepage looks, or a forecast nobody tested.
And when a rebuild is right, it should be run as a risk project, not only a design project. Redirects mapped before launch. Tracking tested before and after. Campaigns checked link by link. A plan to roll back if the numbers fall.
Rebuild or fix: an honest comparison
| Full rebuild | Fixing what you have | |
|---|---|---|
| Timeline | 4 to 9 months | Improvements shipping within weeks |
| Risk to rankings | High if the migration is handled poorly | Low. You improve what already ranks |
| Risk to tracking and ads | High. Links and events commonly break | Low. Tracking is repaired first, not last |
| Where the money goes | Design, content, build, migration, testing | Straight at the pages that already carry value |
| Shape of the result | One big change, then slow decay | Gains that build quarter on quarter |
| When it is right | Obsolete platform, unsafe code, changed business | Nearly every other situation |
How we approach it
At WP Creative we work through the WPO Framework™. Four stages, always in this order.
Stabilise. Fix tracking, hosting, security and anything urgent. The site becomes reliable and measurable before any bigger decision gets made.
Diagnose. A full audit of speed, user experience, search and conversion, using real visitor data. The output is a list of what to fix, ranked by commercial impact. This is also where a rebuild gets recommended, if the evidence supports one.
Optimise. Work down the list, most valuable first. Speed on the busiest pages. Conversion paths on pages that already rank.
Scale. Quarterly cycles. Test, measure, improve, repeat.
The principle behind it: we do not promise to do everything. We promise to do what matters most next.
Most of the time, that is a fix and not a rebuild. Which is why the first thing we do with any new site is find out what is actually broken.
Find out what is broken before you spend a dollar
We now run our WPO audit with AI. It is faster than a manual review and it flags more: security holes, slow pages, broken tracking, dead ad links, and the leaks between a click and a lead.
If you are weighing up a rebuild, or you have a quote for one on your desk, get the audit first. It is free, there is no obligation, and the report is yours to keep.
Or come and see us in person. We are at The Business Show Australia on 25 and 26 November at ICC Sydney, stand B1330. Jean-Luc Giuca from our team is speaking on Wednesday 25 November at 2:45pm in Seminar Theatre 5, on where marketing budgets are lost after the click. Bring your website and your rebuild quote, and we will give you an honest second opinion. Entry is free, so register here.
Frequently asked questions
Should I rebuild my website or fix the one I have?
Fix it, unless the platform is obsolete, the code is unsafe, or the business has changed so much the old site cannot carry it. Diagnose first. The answer usually becomes obvious once you can see what is actually broken.
Why do sales drop after a new website launches?
The common causes are broken ad links, missing redirects, lost tracking, slower pages and an untested checkout. Each one can be prevented with a proper migration plan.
How risky is a website migration?
It is one of the riskiest things you can do to a site that earns money, because it changes URLs, tracking, speed and conversion paths at the same time. The risk is manageable when it is planned for and someone owns the marketing side.
What should I check before signing a rebuild contract?
Ask for the redirect plan, the tracking plan, who checks every live ad after launch, what happens if sales fall, and how you exit if it goes wrong. Be careful with any revenue forecast that is not based on your own data.
Will improving an old website really beat a new one?
Often, yes, and sooner. Improving pages that already have traffic and trust carries less risk than starting again, and the gains arrive in weeks instead of months.
About WP Creative
WP Creative is a web performance agency based in Sydney. We are a 30+ person team across Sydney and Kathmandu, with 1,200+ Australian businesses served, 15+ industry awards and 100+ five-star Google reviews.
We are the web agency behind better marketing. We look after the website layer, speed, stability, tracking and what happens after the click, so marketing teams get more from what they already spend.
Get Your Free Website Audit
($3,000 Value)
- Uncover performance issues
- Identify SEO opportunities
- Security gaps, and quick wins