Somewhere in your business, someone has a slide that says the platform is holding you back. The roadmap is slow, the developers complain about the codebase, and a competitor just relaunched on something shinier. The recommendation writes itself: it is time to replatform.

Here is the problem with that slide. Replatforming is being framed as a technical upgrade when it is one of the largest discretionary capital decisions your digital operation will make this cycle. Judge it as a tech project and you will green-light it on the wrong evidence. Judge it as an investment and most of the proposals that reach your desk fall apart.

The question is not which platform. It is what the switch is worth.

Platform debates default to features. Shopify versus a composable stack, Magento versus headless, this CMS versus that one. Features are the easiest thing to compare and the least likely to determine the outcome, because almost any modern platform can run almost any established brand. The teams that struggle after a migration rarely struggle because they picked the wrong technology. They struggle because the move cost far more than the business case admitted and returned far less.

So the first move is to stop asking which platform is best and start asking what changing platforms is actually worth to the P&L. That reframes the whole conversation from a technical preference into a return on capital, which is the only frame a board should accept.

Count the full cost, not the licence fee

The sticker price of a replatform is the licence and the build. The real price is a longer list that rarely makes it into the proposal. Re-integrating every system that touched the old platform: ERP, PIM, payment, tax, shipping, your warehouse, your CRM, your analytics. Rebuilding the operational knowledge your team has accumulated over years. The marketing capacity you divert into a migration instead of into growth. And the ramp, the months after launch where the new stack is slower than the one your team had mastered.

For an established brand, a serious replatform is usually a six-figure programme once those lines are honest, and often well into seven figures across a year when you count internal time and lost momentum. None of that is a reason not to do it. It is a reason to size the prize before you commit, because a six-figure cost needs a seven-figure reason.

Model the downside, because the downside is where replatforms die

Every business case models the upside: faster pages, better conversion, a happier team. Almost none model the downside with the same rigour, and the downside is what sinks these projects.

The most common one is organic traffic. A migration done without obsessive attention to URL structure, redirects and technical parity can cost a meaningful share of organic visibility, and recovery is measured in quarters, not weeks. For a brand where organic search drives a large slice of revenue, a temporary fall of even ten to thirty percent is not a glitch, it is a number your finance director will see in the quarterly results. Then there is the stabilisation dip in conversion while the team relearns its tools and the inevitable bugs surface in the first weeks of real traffic.

You do not avoid this risk by hoping. You price it. A credible case carries an explicit downside scenario and a recovery timeline, and it survives that scenario still being worth it. If it only works in the perfect case, it does not work.

There is a softer cost too, and it is easy to wave away. For the duration of the programme your best people are pointed at the migration instead of at the market. The senior developer untangling a payment integration is not building the feature that would have lifted revenue this quarter. That opportunity cost is real money even though it never appears on an invoice, and on a long programme it can rival the build itself.

A test that cuts through the noise

When a replatform proposal reaches you, put it through three questions before anyone debates technology.

First, is the constraint structural or operational? A structural constraint is something the current platform genuinely cannot do at any reasonable cost: a market it cannot support, a commerce model it cannot run, a ceiling you have actually hit. An operational constraint is a team that is under-resourced, a backlog that is poorly managed, or a build that has been neglected. Replatforming fixes structural constraints. It does not fix operational ones, it just resets them on a more expensive base, and six months later you have the same backlog on a new logo.

Second, what is the smallest change that removes the constraint? Often the honest answer is not a replatform at all. It is fixing the integration layer, addressing the performance debt, or putting a capable team on the existing stack. If a far cheaper intervention gets you most of the value, the replatform is being used to solve a problem it is not the right tool for.

Third, can you state the return as a number with a payback period? Not "it will be faster" but "we expect this to add X to conversion or Y to revenue, against a fully loaded cost of Z, paying back in N months." If nobody can put that sentence together, the project is not ready, however compelling the demo looked.

What this looks like in practice

A brand on an ageing Magento build is convinced it needs to move to a composable stack. Run the test. The constraint turns out to be a slow, under-invested front end and a brittle set of integrations, both operational. The smallest fix is a front-end rebuild and a proper integration layer, at a fraction of the cost, with none of the migration risk. The composable move gets deferred until there is a structural reason for it, and the brand spends the saved budget on growth instead.

A different brand has genuinely outgrown its platform: it cannot support the markets it is expanding into and the commerce model it now needs. That is structural. Here the replatform is the right call, and the work is to size it honestly, model the downside, and protect organic visibility through the move. It is the kind of considered platform work we have run for established brands like Herman Miller. Same decision framework, opposite answer, because the underlying constraint is different.

The option nobody puts on the slide

There is always a fourth option that rarely gets a line in the proposal: do nothing structural this year, and invest the same money into the current platform and the team running it. It is not exciting and it does not come with a launch. For a large share of the brands that think they need to replatform, it is the better return, and naming it forces every other option to clear a higher bar.

None of this is an argument against replatforming. Done for the right reason, it is one of the highest-return decisions a digital business can make. It is an argument for treating it as what it is: an investment that should clear the same bar as any other use of that capital. If you are weighing one up and want a second read on the business case before it goes to the board, that is exactly the kind of decision our ecommerce consultation exists to pressure-test with you.