This month, a really good question came in:
“My campaigns are generating plenty of leads and hitting our CPA target, but the sales team says the quality is getting worse. How can I optimize for better leads rather than just more conversions?”
The short answer is: If you want better leads, you first need to teach the platform what a better lead looks like.
It’s very easy to lose track of real-world performance when platform-reported performance looks great. Your CPA may be within target and lead volume may look healthy, while your sales team is telling you those leads aren’t turning into meaningful opportunities.
Those two things aren’t necessarily contradictory.
The ad platform may actually be doing exactly what you asked it to do: finding more of the conversions you told it were valuable. If every form fill is treated as the same conversion equally valuable, the bidding system doesn’t inherently know that one became a strong sales opportunity and another went nowhere.
That means improving lead quality starts with the definition of success you’re giving the platform.
In this post, we’ll look at how to define a better outcome, pass that information back to the platform, and give automated bidding enough meaningful signal and time to optimize toward it.
A quick disclosure: I am a Microsoft Advertising employee, and while I’ll reference Microsoft Advertising capabilities where they’re useful, these principles can apply to paid media programs across platforms.
1. Define What A Better Conversion Actually Looks Like
Before changing your bidding strategy or implementing a new measurement tool, ask a more fundamental question:
What actually makes one lead better than another for your business?
For some advertisers, the initial form fill may be meaningful enough. For others, a more granular segmentation might be needed:
- Marketing-qualified lead (MQL): Not quite ready for Sales, but shows enough interest to get added to mailing lists.
- Sales-qualified lead (SQL): Demonstrates needed characteristics that are predictive of a successful customer and shows enough interest in engaging in a demo/trial.
- Turned into an opportunity: The nurture stage between initial contact to closed deal which indicates connection with the Sales team.
- Generated revenue: Closed deal/sale that can now enter nurture flows.
The goal should be to get as close as you reasonably can to the deepest meaningful funnel outcome you can reliably measure.
If 100 leads came through but only 10 were SQLs, you don’t want to teach the bidding system that all 100 were equally valuable simply because they completed the same form.
The closer your conversion tracking gets to the outcome the business actually cares about, the more meaningful the information you’re giving the bidding system.
That’s also why marketing and sales alignment matters here. If the platform says CPA is great while Sales says lead quality is deteriorating, both things can be true.
Before optimizing for “better leads,” make sure everyone agrees on what better actually means. Check that your Sales team is using current lead quality to inform their decisions rather than applying historical bias, as lead quality can change and improve over time with campaign optimizations.
See also: The Conversion Setup Errors That Break Smart Bidding
2. Tell The Platform Which Outcomes Are More Valuable
Once you’ve defined the outcome, the next step is giving the platform information that helps distinguish higher-value actions from lower-value ones.
This is where real-world qualification and revenue data become particularly useful. The best signal any ad platform can work off of is conversion values, and passing accurate values will ensure budgets are directed effectively. This is why conversion value rules are so effective: by adapting the conversion value based on probability of value, you can make it even easier for ad platforms to deliver the right leads.
If you aren’t going to use real-world numbers or outcomes, your CPA or ROAS goals may need to be treated as directional guidance rather than representations of actual business results.
There will also be cases where your ultimate outcome doesn’t happen frequently enough to give conversion-based bidding sufficient volume and signal to effectively optimize performance.
That’s where an earlier but still meaningful conversion can help. A lead becoming marketing-qualified, sales-qualified, or progressing further into your funnel tells you more about lead quality than an unrelated engagement metric.
A useful benchmark for automated bidding is at least 30 conversions within 30 days. Following significant changes to budgets or targets, conversion-based bidding can take as many as roughly 50 conversions to calibrate according to multiple platforms.
See also: Ask A PPC: What Are Learning Periods In Digital Marketing?
3. Build The Feedback Loop
Once you know which outcomes matter, you need a way to pass that information back to the platform.
If you can’t yet measure true value, identify the most meaningful signal you can track reliably and use conversion value rules to assign it directional value. If you’re able to measure and optimize against real value, use offline conversion imports to bring that data into your ads platform. This can fill in the gaps so platforms optimize against real value.
That’s where mechanisms such as offline conversions and Conversions APIs come in.
Offline conversions let you connect outcomes that happen after the immediate website experience back to the original ad interaction. Think about someone who clicks an ad, fills out a form, talks to a sales contact, and eventually becomes a customer.
Microsoft Advertising, for example, documents the Click ID, conversion time, and conversion goal name among the core information used for offline conversion imports.
And Conversions API, commonly called CAPI, lets conversion information move programmatically from your systems into the ad platform rather than relying exclusively on periodic manual uploads.
Microsoft Advertising’s Conversions API lets advertisers send conversion and engagement events directly from their servers using Universal Event Tracking (UET), and can pass business signals including revenue, product details, click IDs, hashed identifiers, and consent status.
Here’s the important part:
The mechanism isn’t the strategy. The quality of the signal you’re sending through it is.
Before automating the flow of conversion data, make sure you trust the data you’re automating.
If your CRM is treating poor-quality leads as successful outcomes, passing that information back faster won’t solve the problem. You’ll simply create a more efficient feedback loop around the wrong definition of success.
This also applies to gaps in volume or value of conversions. Advertisers should confirm where the gap is appearing:
- Conversion definition: Are the right conversion actions being tracked, and if they’re intended to be micro-conversions, are they being set up to accurately convey directional value instead of being predictive of real revenue outcomes?
- Lead qualification: Where in the MQL/SQL spectrum does the lead fall, and if it’s not a fit for either, what are the main reasons? Even if the “lead” is an existing customer, there’s an action item there in only targeting new customers till the lead flow is more balanced.
- CRM progression: Is your CRM set up to neatly pass leads from one stage to another and easily document that?
- Form or landing page experience: when people come to your site or fill out a lead-form ad, is it easy for them to pass their information to you? If they’re on your page, do they seem lost or frustrated in figuring out how to engage you?
- Duplicate/invalid lead volume: Are you double-counting conversions due to multiple tracking pixels or manual entry?
- Sales follow-up: Are the leads you’re currently paying for getting a follow-up within 90 minutes of submission? If not, that could be a factor as prospects find their solution elsewhere.
Auditing for these factors, as much as the technical passing through of conversion data, is a critical step in closing the loop.
See also: Is Your Conversion Data Misleading You? Seven Common Google Ads Tracking Issues
4. Give Bidding Enough Signal And Time To Learn
There is one more piece that can easily get overlooked: conversion delay.
Your campaign may look one way when you initially review results and very different after enough time has passed for leads to move through the sales process.
Microsoft Advertising conversion tracking supports configurable conversion windows, depending on the goal, from one minute up to 90 days.
CAPI introduces another consideration. On Microsoft Advertising, CAPI currently has a 90-day look-back window. Consider which outcomes realistically happen soon enough to pass back through that mechanism.
Your systems still need time for sales to work the lead or for a customer to actually convert.
Don’t Confuse Faster Data Transfer With Faster Business Outcomes
Your reporting cadence should reflect your actual sales cycle, and automated bidding needs enough meaningful data and time to learn.
That may mean resisting the instinct to make aggressive changes because yesterday’s platform numbers don’t yet contain outcomes that take weeks to materialize.
What If Conversions Aren’t The Right Measure Of Success?
Everything above assumes the campaign’s purpose is ultimately a conversion-driven business outcome.
That isn’t always the case.
If you’re running an awareness or consideration campaign, it may generate sales and booked demos, but these should be reported as secondary key performance indicators rather than the appropriate immediate measures of success. Instead of forcing every campaign into the same CPA or ROAS framework, start by being explicit about the business outcome the campaign is intended to drive and select measurement signals appropriate to that objective.
The important thing is consistency between what you want the campaign to accomplish and what you’re asking the platform to optimize toward.
Use Behavioral Signals To Understand What Conversion Data Can’t Tell You
Conversion totals also don’t explain everything happening between an ad interaction and the eventual outcome.
That’s where diagnostic signals can be useful.
Search behavior can help you understand whether you’re creating interest. Search term reporting, for example, can surface changes in branded searches or the kinds of queries coming from your audience.
Microsoft Clarity can help diagnose what happens after the click. Behavioral analytics such as recordings, heatmaps, and engagement insights from your website can help surface whether visitors are meaningfully exploring the site, encountering friction, or leaving without taking the intended action. Clarity AI Visibility can add another discovery signal by showing when content is cited in supported AI-generated experiences and which AI grounding queries triggered those citations.
These signals aren’t substitutes for your business outcome.
Instead, use them to help answer questions the conversion number alone can’t: Are we creating interest? Where are we losing it? Is campaign performance the problem, or is something happening later in the customer journey?
Final Takeaways
If you want better leads, don’t just ask your ad platform for more conversions. Teach it what a better conversion looks like.
Start by defining the deepest meaningful outcome you can reliably measure. Then pass qualification, revenue, or other value signals back where appropriate, so the platform has a better understanding of what success means to your business.
Use offline conversions, CAPI, or other feedback mechanisms to connect those real-world outcomes with advertising activity. And make sure you’re giving automated bidding enough meaningful signal and enough time to learn from it.
If you choose not to pass those real-world outcomes back, that’s valid too. Just be honest about what your platform metrics represent and keep your internal source of truth at the center of your business decisions.
The platform can only optimize toward the definition of success you give it.
So when lead volume looks great, but quality doesn’t, don’t immediately ask the bidding system to find more.
First, make sure you’ve taught it what “better” means.
More Resources:
- Google Ads Is Automating More Decisions, Your Measurement Needs To Get Better
- How To Maximize Paid Ads Profitability With A Strategic Landing Page Audit
- How To Report The PPC Metrics Your CFO Actually Cares About
Featured Image: Paulo Bobita/Search Engine Journal
