Are you accidentally polluting your conversion data?

I bumped into an interesting problem on Reddit.

Google Ads was delivering a lot of cheap leads, but only a handful turned into sales. The sales team was burned out because they were putting in the work but not seeing the rewards.

The poster asked:

How do I teach Google that I want fewer leads, but significantly better ones?

They were happy to pay more for leads that were likely to turn into sales.

The obvious fix is offline conversions

Several replies recommended optimising the campaign for something that happens later in the sales process: a qualified lead or a sale. Offline conversions, in Google speak.

Optimising for something that happens in the business after the lead arrives is usually the right move. Here’s why:

• If you set the goal to be as many form fills as possible, Google looks for more people who are likely to fill in your form. In practice, this risks getting you lower-priced but lower-quality leads.
• If you set the goal to a qualified lead, Google looks for more people who are likely to become qualified leads. You tend to get leads who are more likely to turn into sales, although at a higher cost per lead.
• It’s the same if you set the goal to sales. Google learns to look for more people who are likely to become customers.

A good goal has three characteristics:

• It happens frequently. The more often the goal happens, the more information Google has to learn from. If you only make a handful of sales from Google Ads each month, sales may not happen often enough to be a useful goal.

• It happens quickly. The shorter the delay between seeing the ad and reaching the goal, the better. It might take two minutes from seeing an ad to filling in the form. It might take three months between seeing the ad and closing the deal.

• It reflects business value accurately. For example, form spam is worth nothing. But if your conversion tracking system tells Google that all form submissions, including spam, are valuable, Google tends to find more spam. This is the junk lead death spiral.

Choosing the right goal is a trade-off. I explain this in more detail in Which Google Ads conversions should I optimise for?.

But once you've chosen the right goal, you still need to give Google clean data to work with. And the Reddit post sparked a thought: your sales process could accidentally be polluting your conversion data.

I’ll explain in more detail in a minute. But first, do you think your sales process affects how many leads become sales?

Obviously, yes.

• An emergency plumber isn’t going to do a lot of jobs if he calls leads the next day.
• An overworked physio isn’t going to sign many new patients if they have to wait three weeks for their first treatment.
• A contractor isn’t going to win as many quotes if she doesn’t follow up after sending them.

Your sales process also affects the feedback you send to Google. This, in turn, affects who Google shows your ads to.

A few examples should make it clearer.

A difficult form means some people won’t fill it in.

The Reddit poster had deliberately added friction to the form by asking a lot of questions. I guess this was an attempt to improve lead quality. (I don't think this works.)

From what I can tell, the poster didn’t use the answers to tell Google if one lead was better than another. Everyone who completed the form was counted as a lead.

Here’s how this goes wrong:

• Google shows the ad to someone who is likely to be a good lead. That person doesn’t fill in the form because it looks long and difficult, so Google thinks that kind of person isn't interested.
• Google shows the ad to someone with different characteristics which make them a poor fit. But they’re patient and fill in the long form, so Google thinks it's a win.

The business process - the long form - affects what Google learns from. In this case, Google is learning to find people who don’t mind filling in long forms.

How interested is a three-week-old lead?

The poster said the business was too busy to speak to every lead. They sent automated emails, but it could be two or three weeks before a salesperson called.

Someone who was interested three weeks ago has probably solved the problem by now. They’re unlikely to still be a qualified lead.

If the business used qualified leads as its goal, Google would learn that many leads weren’t qualified. Not because Google had shown the ads to the wrong people, but because of how long it took the business to speak to them.

Again, the business process - slow contact - affects what Google learns from.

A demotivated sales team isn’t going to put in the effort.

The poster said that the volume of leads and low sales rate were becoming demotivating for the sales team.

I don’t know exactly how this affected their lead handling. But it’s been my experience that when salespeople expect the next lead to be a waste of time, they don’t make much effort. They call more slowly, make fewer contact attempts and don’t follow up consistently. This leads to fewer sales.

Again, this affects what Google is learning from.

So what’s the answer?

The easy thing to say would be to improve lead quality, whip the sales team back into action and give Google a goal with real business value.

I’ve worked with a lot of businesses over the years. Improvements like this look good on paper, but in practice humans resist change. It takes a gentler approach.

If this were my problem, I’d create a qualification checkpoint immediately after the enquiry.

I’d begin with the next 100 leads and make sure each was handled in the same way:

• Call within an agreed period.
• Make an agreed number of contact attempts.
• Ask the same short set of qualification questions.
• Apply qualification criteria that were defined in advance.

The business would need to decide what “qualified” means. It'd probably be some mix of location, requirements, budget and when they're planning to buy.

The salesperson should record two things for each lead:

• Is this lead qualified?
• Was the agreed process followed?

The first answer should be “yes”, “no” or “unknown”.

Someone who was contacted promptly but didn’t have the budget might be unqualified. Someone who wasn’t reached is not necessarily unqualified. They get the "unknown" label.

The second question tells us if we can trust the first answer. If the salesperson called three weeks late or made only one contact attempt, the result tells us more about the process than the lead.

This gives us a qualification checkpoint early enough to minimise the effect of the rest of sales process.

Once the business is applying this process consistently, I’d use qualified leads as the goal for the campaign.

It meets our requirements for a good goal:

• It’s likely to happen often.
• It happens soon. If the team prioritises calling, it is likely to get an answer within a few minutes or within a day or so if several attempts are needed.
• It reflects business value reasonably accurately. Qualification is still a human judgement, but written criteria and a consistent process make the data much cleaner.

This should start a virtuous circle. Google learns what a qualified lead looks like from cleaner data. Google looks for more people like them. The business gets more qualified leads. Qualified leads are more likely to buy.

But there's a catch.

You can’t remove the pollution by leaving badly handled leads out of the conversion upload. Google will still see clicks that failed to produce qualified leads. The process has to become consistent.

If the way I think resonates with you I'd like to give you a free copy of my book Profitable Google Ads. You can download it here.

Related articles

Why I delay Google Ads conversions on purpose

I send conversion data more slowly in exchange for data that's permanent, auditable and recoverable.

Google Ads conversion tracking for Pipedrive

Connect Pipedrive to Google Ads and fix the offline conversion gap automatically, in about 30 minutes.

What is Google actually learning from?

Tag-based tracking only tells Google what the browser saw. Conversion tracking must be right, reliable and auditable.

Offline conversion tracking with Contact Form 7

Contact Form 7 is one of the most popular form plugins for WordPress. But, if you're running Google Ads, there is a problem. Contact Form 7 can tell you that somebody submitted a form. It can't tell Google Ads if that person became a customer.

The Google Ads WhatsApp conversion tracking guide

Here's the complete guide to tracking WhatsApp conversions from click to sale.

Offline Conversion Tracking with Fluent Forms

Fluent Forms is a great way to capture leads from a WordPress website. But if you run Google Ads there is a gap in the system. Google can see the click and it can see the form submission. But it can't see what happens after the lead arrives. Here's how I helped a client close that gap.

Topics you'll find on this site