Bad data used to mean bad reports, now it means poor ad delivery

We’ve all seen dashboards that don’t make sense when you look into the numbers, but now that same data could be training your campaigns to spend your budget chasing the wrong people.

As automation takes over more of the ad-buying process, from creative generation to bidding, data has become one of the last inputs advertisers can control, and perhaps the most important. That’s because automation can only optimize for the signals you give it.

Think about it: Which is worse, a brilliant ad shown to the wrong audience or an average ad shown to the right one? The first spends your budget reaching people you don’t want. The second might get ignored, but if someone does engage, at least they’re the right person.

But can you honestly say the last time you set up a campaign, you spent more time verifying the data than thinking about the ad copy?

The cost of bad data has changed

Several years ago, bad tracking was a reporting problem.

If a tag fired twice, a conversion was mishandled, a value came through incorrectly, or your offline conversions stopped working for a few weeks, the result was a dashboard that didn’t add up. It was annoying, but had little impact. Eventually, someone would question the numbers during a monthly review, you’d trace the issue, fix it, and the data would be good for the next review.

However, that same data now feeds the algorithm buying your paid media. Smart Bidding doesn’t wait for you to interpret a report or reach your monthly review – it reads your conversion data and acts on it before you’ve even noticed an issue.

The same number, now wrong, has a different outcome. A bad number in a report requires an explanation in a meeting.

A bad number in a conversion used for bidding costs you because the algorithm doesn’t know it’s wrong. It optimizes toward that signal the moment it sees it, and it does so efficiently.

Google doesn’t understand your funnel or your business

While conversion actions are labeled in Google’s interface as “lead,” “opportunity,” and so on, those labels are just for organization. The platform doesn’t actually understand where the conversion event sits in your funnel.

All it sees is a conversion event with a numeric value attached to it (usually representing a currency value), so it has no idea that a newsletter sign-up is worth $2 in eventual value, a lead is worth $60, and an opportunity is worth $400. Google sees three conversions. It has no idea one is worth 200x another.

The algorithm isn’t optimizing for your business outcome. It’s optimizing for the data you’ve given it. If the data is wrong, the optimization will be, too.

For example, if every form submission fires the same conversion with the same default value, there’s no clean way to distinguish tire kickers from high-value inquiries, so the algorithm treats them identically. And because tire kickers are usually cheaper to acquire, it floods you with them. 

The cost per lead drops from $40 to $25, and the dashboard makes your cost per lead look more than 35% lower, but the pipeline quickly dries up as genuinely qualified inquiries quietly halve in number.

Dig deeper: Why better signals drive paid search performance

3 ways bad data quietly wrecks delivery

Bad data can take different forms, but these are the three issues most likely to derail campaign delivery.

1. Wrong event

Optimizing for a top-of-funnel action like a page view, when the real conversion events occur further down the funnel, causes the algorithm to buy more and more of those cheap events without the lower-funnel activity actually following through.

2. Wrong value

Counting all conversions equally (or assigning them a flat placeholder value) when their actual value varies by 10x. The algorithm chases volume of the lower-value conversions because they’re easier to acquire.

3. No data

This one isn’t discussed enough. Nothing kills a campaign faster than a complete break in the data.

On Day 1, the algorithm wonders where the conversions are. By Day 2, it starts to assume they aren’t coming. By Day 3, it’s making serious bidding changes. Within a week, most campaigns will have throttled themselves to almost nothing.

How to pick the right signal for Google

So how do you fix it?

Take a typical lead generation business as an example. Some leads will never convert, while others are worth 10x as much as the rest.

If your form asks the right qualifying questions, you already know which is which. But if you’re optimizing for every lead submitted using a target CPA, you’re telling Google they’re all equally valuable.

Imagine an account spending $20,000 a month at a $40 target CPA and generating around 500 leads. Only 150 qualify, and maybe just 50 are genuinely high-value. A lead’s expected worth is $60, a qualified lead is $200, and a high-value lead is $600. That’s a 10x spread in value.

You have several ways to improve the optimization signal:

• Optimize for a qualified lead: Create a new conversion action, such as “qualified lead,” and fire it only for leads with value. You can then move your target CPA to this conversion action, knowing it’ll ignore leads with no value. The advantage is that you’re training the campaign on a more meaningful signal. The downside is that every lead with value is still treated equally.
• Assign conversion values and use the target ROAS: Add a currency value to the qualified lead based on the potential revenue it could generate if it converts to a sale. You can then switch the campaign to target ROAS, allowing Google to optimize for return instead of simply counting leads with no value. However, it may still buy larger numbers of lower-value leads if it can acquire them at the right price, rather than prioritizing higher-value ones.
• Optimize for a high value lead: Create a “high value lead” conversion event that fires only for your top-tier leads, with or without a conversion value. You can then optimize with either target CPA or target ROAS, depending on whether you want to focus on acquisition cost or return. The advantage is higher-quality leads. The downside is that, depending on your spend, the data may be too limited to support this approach until you scale.

These are just a few possible optimization signals without even going deeper into the funnel. You can apply the same approach to lower-funnel events as well by creating separate conversion actions for milestones such as contacted lead, qualified contact, or high-value contact.

Dig deeper:

Targeting and measurement can be different

It sounds simple, but the conversion event you optimize for and the one you report on aren’t, and arguably shouldn’t be, the same. One trains the algorithm. The other tells you how that training is performing.

In our previous example, a client or internal stakeholder might want to see cost per lead, which is a perfectly valid metric. Meanwhile, the campaign is optimizing for the Qualified Lead conversion, not the original lead.

You keep the original lead conversion running purely as a reporting metric, so stakeholders still get their cost per lead while the campaign bids on the qualified lead signal that actually drives business value. Same campaign, two conversions, two very different jobs.

Which brings us back to where we started: Did you spend more time verifying the data than writing the ad? In an automated account, data is now your strategy.