Establish whether the problem is the site, the traffic, or the offer, because only the first is solved by replacing it. Then ignore what it cost and decide on what the next twelve months need.

The fact that gets in the way

What the site cost is the first thing anybody says about it and the least relevant.

That money is spent. It is not recoverable by continuing, and it does not make the site more likely to start working.

The reasoning that keeps a failed site running is that abandoning it would waste what was spent, when in fact the waste already happened and the only question left is what to do next.

Which is easy to say and genuinely hard to act on, particularly when the person who chose the supplier is in the room.

Establish what actually failed

Before replacing anything, because three different problems look identical from the outside.

Nobody arrives, which is a visibility problem and is not solved by a new site.

People arrive and leave, which is a site problem and may be.

People enquire and nothing converts, which is a sales or offer problem and is definitely not.

Your analytics answers the first two in about ten minutes: how many visitors over twelve months, and what proportion did anything.

Rebuilding a site with fifty visitors a month produces a better site with fifty visitors a month, which is the most common way this money gets spent twice.

The three failures and their remedies

The fourth is where a surprising number of businesses are, and it means the diagnosis has not actually been made yet.

Salvage before deciding

Even a failed site contains things worth keeping.

The domain, which is usually the most valuable asset and has nothing to do with whether the site worked.

Any page addresses that have accumulated links or visitors, however few.

The content, which may be poorly presented rather than poorly written.

Photographs, particularly originals at full resolution.

And the record of what was tried, which stops the next attempt repeating it.

Extract all of that before anybody touches anything, because it becomes considerably harder once a replacement project starts.

A worked example

A firm had spent a significant sum on a site three years earlier that had produced, as far as anybody could establish, four enquiries.

The instinct was to rebuild it properly this time with a better supplier.

The analytics showed roughly sixty visitors a month, almost all of them people searching the business name.

Nobody was finding them for anything they sold, because the site had no pages about individual services and the business listing was largely empty.

They kept the site, spent a fraction of the rebuild budget on six service pages and a photography day, and completed the listing.

Traffic rose to several hundred a month within two quarters and enquiries followed.

The site had never been the problem. It had simply been the only visible thing to blame.

When replacing genuinely is the answer

There are clear cases and they are worth naming so this does not read as an argument for never rebuilding.

A site nobody can update, so no amount of content work is possible.

A site that cannot be made usable on a phone.

A platform that is abandoned, insecure, or that you do not control.

A structure that cannot express what the business now sells.

Or a site that has traffic, that people leave immediately, and where the specific reasons have been identified and cannot be fixed in place.

In those situations the money is spent on removing a constraint, which is different from spending it in the hope that a nicer site performs better.

Decide on the next twelve months

The framing that removes the sunk cost from the conversation.

Ignore what has been spent. Ask what the business needs the site to do over the next year, what the cheapest way to achieve that is, and whether the current site can be part of it.

That question can be answered without anybody defending a previous decision, which is why it produces better outcomes.

It also frequently produces a smaller answer than a rebuild: six pages, some photographs, a completed listing, and a review routine.

If the honest answer is still a rebuild, it will survive the question, and it will be a decision rather than a reaction.

Handling it internally

The part that is not technical and determines whether the right decision gets made.

Somebody chose the supplier, approved the design, and defended the spend. Making the review about the site rather than about that person is the difference between an honest assessment and a defensive one.

Frame it as what we know now that we did not know then, which is accurate and does not require anybody to have been wrong.

Write down what was learned: what the brief missed, what was not asked, what would be specified differently.

That record is the actual return on a failed project, and it is worth more than the recriminations that usually take its place.

The counter-case

Sometimes a site has not failed and is simply being judged too early or against the wrong measure.

A site launched four months ago has not had time to accumulate visibility, and concluding it does not work is premature.

A site for a business that is mostly referral-based may correctly produce very few enquiries while doing valuable work reassuring people who already have your name.

And a business with no measurement in place may be receiving enquiries it is attributing elsewhere, particularly phone calls.

Before writing anything off, confirm it has been given a year, that you have measured properly, and that you have asked customers how they found you.

The decision

  1. Check twelve months of traffic and what visitors did.
  2. Diagnose which of the three failures you have.
  3. Confirm you measured before concluding anything.
  4. Salvage domain, content, images and addresses.
  5. Ignore what it cost.
  6. Ask what the next year needs, and the cheapest route.
  7. Write down what the brief missed.

Step two prevents the most expensive version of this, which is buying a second site to solve a problem the first one never had.

Deciding between repair and replacement is covered in deciding to rebuild instead of repair.

Frequently asked questions

Should I rebuild a site that produced nothing?

Not until you know why. Nobody arriving is a visibility problem, arriving and leaving is a site problem, and enquiring without buying is an offer problem. Only the second is solved by rebuilding.

How do I diagnose it?

Twelve months of analytics answers most of it: how many visitors, and what proportion did anything. If you have no measurement, that is the first thing to fix.

Does the money already spent matter?

No. It is not recoverable by continuing and does not make the site more likely to start working. The only question is what the next twelve months need.

What should I salvage?

The domain above all, any page addresses with links or visitors, the content, full-resolution photographs, and a record of what was tried.

When is replacing genuinely right?

When you cannot update it, cannot make it work on a phone, cannot control the platform, or the structure cannot express what you now sell.

How do I keep the review honest?

Frame it as what we know now that we did not know then. Somebody chose the supplier and approved the design, and making it about the site rather than the person produces a better decision.

West Coast Media Solutions Inc. provides web design, web development, hosting, digital marketing, and business consulting to organisations across Canada, drawing on more than twenty-five years in the field.

Site that never produced anything?

Check twelve months of traffic first. Rebuilding a site nobody visits produces a better site nobody visits.

Start a Conversation