A company pays tens of millions for an ERP. Eighteen months later, its teams are still managing stock on an Excel file circulating over WhatsApp. The software is there, paid for, deployed. The management accounting modules are switched off, and nobody looks at the dashboards more than once a quarter.
That case is far from isolated, and it does not only concern ERPs. It is exactly what happens to a website rebuilt from scratch, to a new acquisition channel, to the latest marketing tool. Two things that have nothing to do with each other have been confused: buying an asset, and obtaining a result.
Purchase does not make the transformation
The digital transformation consultant Mohamed Cheikh Sidiya describes this paradox in an opinion piece published by Le Calame on 30 April 2026. Companies acquire serious management software, of the Dynamics 365, Sage or Odoo kind, and never really use it. His formula sums it up: digital transformation is not a purchase, it is a discipline.
It is an opinion piece, not a third-party quantified study. The amounts mentioned are qualitative and serve as illustration, not statistical proof. But the mechanism it describes is one we see everywhere, in every market and every language. A manager quoted in that same piece puts it better than any pitch: “I have more information than ever, and I make worse decisions than before.”
That is the heart of the problem. The tool kept its technical promise. It generates data, reports, charts. And yet the decision-making did not improve. Because the result was never in the tool. It was in the way it is used, and that way is not bought with a licence.
A website follows exactly the same slope
Transpose it to a website. An SME decides to rebuild its own. The brief fits in one line: “Our site looks dated, we need to modernise it.” Three months later, the new site is live. Faster, cleaner, prettier. Everybody is pleased on launch day.
Six months later, the question that counts arrives: how many visitors does this site turn into customers? And the answer, most of the time, is “we do not know”. The site was bought as a finished object. Nobody defined what it should produce, nobody measures what it produces, nobody adjusts it. The fine asset sleeps, exactly like the ghost ERP.
The slippage is even more treacherous today than it was yesterday. HubSpot’s State of Marketing 2026 report puts brand awareness and reach as the number one marketing priority for the first time, at 35%, ahead of sales and revenue growth, at 29%. At the same time, nearly 30% of marketers report a fall in their search traffic, as internet users put their questions to generative AI tools rather than to Google. In other words, the ground is moving underfoot. Buying a site with no quantified objective was already risky when traffic rose on its own. When it starts falling, it becomes an expense you never recover.
The result is played out before and after, not during the purchase
If the value is not in the asset itself, where is it? In two places nobody really invoices, and which nonetheless decide everything.
Upstream, in the scoping. Before producing a single mockup, you have to answer a simple and unpleasant question: this site, this tool, this channel, what should it gain the company, in numbers, by when? “Modernise the image” is not an objective. “Double qualified quote requests in six months” is. The nuance looks like nothing. It is what separates a steerable project from a blind purchase. Without a quantified objective written down in black and white, there is no way of knowing later whether the investment served any purpose.
Downstream, in the discipline of use. A tool only produces a result if it is genuinely used, measured and corrected. That is precisely the step the companies in the opinion piece skipped: modules switched off, dashboards consulted once a quarter, a quiet return to the Excel file. The technology did not fail. The use never started. For a website, that discipline has a concrete name: you look at the conversion figures at three months, identify the screen where visitors drop off, correct it, and measure again at six months.
The brief before the brief: if a supplier quotes you for rebuilding your site without first asking how much it converts today or what business deadline lies behind the redesign, they are selling you a service. They are not scoping your project. That is convenient for them, because it is a quick sale. It is expensive for you, because you walk away with one more asset that will produce nothing.
Why this is so hard to hear
Let us be honest about the commercial mechanics. An agency’s reflex is to sell the deliverable and consider the work finished on delivery day. The client wants a beautiful site, they get a beautiful site, everyone signs the acceptance report. It is comfortable on both sides.
Except that this logic manufactures precisely the ghost assets described above. An agency that never questions the brief, that executes a specification without scoping the objective, delivers a product it knows has a one in two chance of producing nothing. It has sold, but it has not served.
That is why upstream scoping and downstream measurement are not optional boxes to tick. A two-hour scoping workshop with your team turns “rebuild the site” into a quantified business objective, and maps the journey that leads a visitor to the action that matters to you. An impact report at three months and then at six months checks, with real data rather than intuitions, whether the investment is working. Those two steps cost time before the sale and impose a demand afterwards. They are also the only verifiable difference between an asset that sleeps and an asset that pays.
Before signing your next digital purchase
Whether you are about to order a website, a management tool, a campaign or a new channel, three questions are worth more than ten quotes.
The first: what precise business result, expressed in numbers, should this purchase produce, and by when? If you cannot answer, this is not the moment to buy, it is the moment to scope.
The second: who, internally, will actually use the tool day to day, and with what follow-up discipline? An asset with no owner of use always ends up as a switched-off module.
The third: on what date, and with which indicator, will you check that the investment served its purpose? Without a measurement appointment set in advance, nobody will ever make it.
A deliverable is not a result. The finest asset in the world, paid for at full price, is worth nothing until it has been scoped against an objective and genuinely put to work. It is less flattering than inaugurating a new site. It is the only thing that decides, in the end, whether the money spent comes back.
