My client's new website lost them $223,370.27 in profit last month.

The month before the shiny new website went live we spent $172,238.74 on Google Ads.

We've spent roughly the same every month for the last 3 years. And we've managed to keep a R.O.I of 230%. For every $1 spent on Google Ads the client earned $2.30 in initial profit.

That makes our $172,238.74 ad spend worth $396,149.10 in profit.

Last month we only spent $75,121.23, worth $172,778.83 in profit. If I've done my math we're down $223,370.27.

Here's what happened, how we're planning to fix it and two things I learned that made me change my mind.

This story starts with the usual office politics.

Client hires a new manager who wants to sweep clean. New manager brings in people he's worked with before and sidelines existing service providers. New team insist on a new website.

This is nothing new. It happens every day.

In the past I've been brought in to replace other service providers. This time I was the one being shifted out.

What's interesting is how quickly our Google Ads performance crashed with the new site. We saw a dip after a week. After a month the call centre operators were watching TikTok all day because there were so few leads coming in.

Here's what went wrong...

TLDR: Client's web developers broke the conversion tracking.

We get leads via enquiry forms, and phone calls when people click a button on the website.

• The web developers broke the enquiry form conversion tracking by accident.
• The new broom manager broke the phone call conversion tracking on purpose. We were spending about $1 600 a month tracking call conversions with CallRail. The new manager decided that was too expensive and closed the account.

We use Target CPA bidding strategy for most of the advertising. That means we tell Google how much we want to pay for a lead. We let Google know how many leads we got using the enquiry form and call conversion tracking (now broken). Google then does their best to get us leads at our target cost per lead.

This approach has produced good leads at our target price per lead for years. But when those cretins broke the conversion tracking we couldn't tell Google that we were still getting leads.

That meant that the bidding AI thought its ad targeting wasn't working. It started changing when and where our ads showed in a desperate attempt to get leads.

Eventually it gave up and stopped showing our ads as often. The month before the new site our ads showed 151 168 times. Last month that was down to only 88 374. Fewer ad impressions meant fewer clicks, fewer visitors and fewer leads. And a call centre with nothing to do.

I suppose this could end as one of those stories where I boast about how they couldn't do it without me. But the truth is anyone managing PPC ads knows that without conversion tracking you're flying blind.

Luckily it only took them two months of pain before they agreed to fix the broken conversion tracking and pay for CallRail again.

In a couple of days the bidding AI should know we're getting leads. I don't know if it'll be able to recover automatically or if I'll have to intervene. In a month or so we'll know for sure.

I learned a few surprising things from this...

Speed doesn't matter as much as I thought it did.

The new site is sloooow. It takes 6.7 seconds to load on a mobile phone. The old site loaded in 1.3 seconds.

I'd expected that the slow load would kill the conversion rate.

It didn't. The conversion rate (leads/clicks) remained constant at about 17%. (I was able to calculate the conversion rate from data in the CRM even though it was broken on the site.)

At first I thought that the new site might have been slower but better designed, but that wasn't the case. The copy and design of the landing pages was almost exactly the same as the old site.

I have a couple of ideas why the slow page didn't hurt the conversion rate:-

• Most of our leads use older or cheaper (slower) mobile phones which might make them used to waiting for sites to load.
• The client offers a free consultation where some competitors charge for consultations. This might be enough of an incentive to keep someone waiting for the page to load.

Let me know if you have any other ideas on the subject.

A header navigation bar with lots of links doesn't harm conversion rate for mobile users.

It's common practice to have a very stripped down navigation bar across the top of landing pages. The thinking is that you want the visitor to call or fill in the form rather than go wandering around your website.

The old site's landing page navigation bar had the logo and a clickable phone number. Nothing else.

The new site's developers couldn't figure out how to put a different navigation bar on the landing pages, so they used the full navigation bar from the site. It has the logo and the usual links to things like contact, about, privacy, services, blog and so on.

I'd expected that this would lead to a lower conversion rate because people would leave the landing page for other pages. But, as I mentioned above, the conversion rate didn't change.

In hindsight it's obvious why a busy navigation bar didn't lead to a poor conversion rate. Most of our traffic is on mobile phones. The navigation bar shrinks to fit the narrow screen.

On a mobile it only has the logo on the left and a hamburger menu on the right. (In case you're not familiar with the term, a hamburger menu is three stacked lines that signify click on this and you'll see the rest of the menu).

I guess if we had more desktop traffic the busier navigation might have reduced the conversion rate.

Hope you've found this interesting.

Related articles

Why good Google Ads campaigns still lose clients

The AI audit ambush

Most Google Ads Reports Focus on the Wrong Things

Helping people make better Google Ads decisions

When there are no good options in Google Ads

Why I've Stopped Giving Quick-Fix Google Ads Advice

Topics you'll find on this site