You can't run profitable Google Ads if your pipes are leaking

You can't run profitable Google Ads without working plumbing - the systems that carry data from the website, to the business and back to Google.

This data is crucial:

• The business owner uses it to decide where to invest. Which services should we promote? Where does the budget go?
• The ad manager uses it to improve performance. Should we change campaign type? Should we use search partners?
• Google uses it to optimise in real time. Should this ad show? What should the bid be?

The problem is that the data is siloed. It's in Google Ads. It's in call tracking systems like CallRail or WhatConverts. It's in CRMs, spreadsheets and post-it notes.

In most cases one of two things happen:

• The data stays in siloes. Everyone sees different numbers. The owner sees ad spend. The ad manager sees leads. Sales sees bookings. No one is basing their decisions on the complete picture.

• Data flows between silos via a mishmash of native integrations, Zapier and duct tape. This is brittle and prone to silent failures. When something goes wrong nobody notices until campaign performance suffers. Then the ad manager gets it in the neck. They waste hours or days trying to unearth the root cause, fix it and repair the damage.

The plumbing is important enough to be designed, monitored and maintained. So I'm designing, monitoring and maintaining it.

I've done this a handful of times for larger clients where a bespoke solution made sense. I've also built several public-facing tools that solve parts of the plumbing problem, so the work isn't new to me.

What’s new is making the same interface useful for businesses generating 30 leads a month and 3,000.

This week I’ve been working on the layer that monitors the data flow and flags problems.

Profitable Google Ads needs four kinds of data:

• Lead capture. Did we receive the lead details and tracking data when someone filled in a form or called?
• Lead feedback. Did someone record what happened next: qualified, not qualified, or won?
• Conversion upload. Did Google receive the signal that this click became a lead, a qualified lead, or a sale?
• Lead delivery. Did the lead reach someone in sales so they could follow up?

If data stops flowing through any one pipe, performance suffers. We need to know that each pipe is functional.

Here’s the early version of the data health dashboard.

Conversion flow is predictable. Google fetches once per day. If it misses its window by more than six hours, the dashboard turns red. So you don't find out a week later that tracking broke.

Feedback flow is predictable too. If leads sit unmarked for too long, the dashboard turns red.

And the delivery flow is also predictable. Leads go out by email or webhook. If delivery fails, the dashboard turns red.

Lead capture is unpredictable.

A business that gets 50 calls a day needs to know if it only got five. But a business that gets 20 calls a month expects quiet days.

Figuring out the difference between a broken pipe and normal dip is tricky:

• If the system triggers too many false alarms, people ignore them.
• If the system doesn't send an alarm when it should have, it's useless.

I modelled ideas using rolling averages, weighted rolling averages and user-defined thresholds. None of them felt reliable.

So for now, I’ve settled on something simpler. The dashboard shows when the last lead arrived. It’s not perfect. But combined with what I know about each of the accounts I manage it's enough to warn me if something's not working.

If you’re responsible for performance but you’re not completely sure the data can be trusted, that’s uncomfortable. I'd love to show you what I'm working on and get your thoughts. Message me and we'll make it happen.

Related articles

Why a Google Ads account audit isn't enough

Why I'm rethinking how I audit Google Ads accounts

Testing Google Ads for an RV Campground

Why negative keywords are becoming less effective in Google Ads

What’s the right bidding strategy for a brand new Google Ads account?

The campaign looks good but the ads aren't profitable

Topics you'll find on this site