The cost of not deciding: what an outdated website costs every month
How to put a monthly number on a website that is slow, dated or unmaintained, and why the number usually ends the debate.
The short answer
Not deciding about an outdated website feels like the cautious option, and it is a decision to keep paying a monthly cost without receiving anything for it. The cost arrives in five ways, none of them as an invoice. Visitors lost to a slow site on phones. Enquiries and orders lost to poor forms, a poor mobile experience and a dated impression. Staff hours spent on workarounds the site should handle. Risk carried from an unmaintained platform, which arrives eventually as an incident. And opportunities declined because the site cannot support them. Each can be estimated from numbers you already have, roughly but usefully. The decision turns on one comparison: the monthly total of the current situation against the cost of a rebuild divided by the years you expect to use it. Count into the monthly total the hours spent working around the site, the leads that do not arrive, and the fixes that recur. The debate tends to end once both numbers are on the page.
The five costs and how to estimate them
| Cost | How to estimate it | Where the numbers are |
|---|---|---|
| Visitors lost to slowness | Mobile visits times the share that abandons at your speed versus a fast benchmark | Analytics; real-user speed data |
| Enquiries and orders lost | Visits times the gap between your conversion and a well-built site’s, times value per enquiry or order | Analytics; sales history |
| Staff hours on workarounds | Hours per week spent on manual steps the site should handle, valued honestly | Ask the team |
| Risk carried | Estimated cost of the incident an unmaintained platform will have, spread over the expected months until it | Incident guides; your own exposure |
| Opportunities declined | Campaigns not run, markets not entered, integrations not possible because of the site | The list of things people stopped asking for |
Putting it on one page
- Pull three months of analytics: mobile share, top pages, conversion, enquiries.
- Get the real-user speed data for the top templates on mobile.
- Ask the team for hours spent on site-related workarounds each week.
- Estimate each of the five lines roughly, with the assumption written next to it.
- Total them per month, and put the rebuild or improvement cost per month over its life next to it.
- Read both numbers aloud in the meeting.
Why the number ends the debate
Because it turns a choice between spending and not spending into a choice between two kinds of spending, one of which buys nothing. A monthly loss that exceeds the monthly cost of fixing it is not a difficult decision; it is an unmade one. Where we have done this exercise, the loss line came out several times the fix line, and the surprise was how long it had been that way.
What this means for you
Estimate what your current site costs each month in lost visitors, lost enquiries, staff hours, carried risk and declined opportunities, from numbers you already have. Put the total next to the cost of fixing it spread over its life. Decide on purpose, whichever way the numbers point. The one outcome to refuse is the default, because the default is the monthly number, paid indefinitely, for nothing.
Frequently asked questions
How can a website that works be costing us money?
By losing what it should be winning. A site that takes seconds on a phone loses a share of visitors before they see anything; a form that is hard to use on mobile loses enquiries; a dated appearance loses trust; a checkout with surprises loses orders; and a platform nobody maintains carries a risk that will eventually arrive as an incident. None of it appears as an invoice, which is why it feels free.
How do we estimate lost visitors and enquiries?
From your own numbers: mobile traffic times the share that typically abandons at your page speed, against a fast benchmark; enquiries per visit compared with a well-built form's typical rate; the value of an enquiry from your sales history. The estimates are rough; the order of magnitude is what matters, and it is usually enough to decide.
What about the risk line?
An unmaintained platform will have an incident: a compromise, an outage, a plugin failure. Estimate the cost of one, downtime, clean-up, lost sales, reputation, and spread it over the months you expect to go without one. That monthly figure is the insurance premium you are paying by carrying the risk, and it is usually not small.