Are you optimising PPC campaigns for trial sign-ups or for submitted enquiry forms, but only a fraction of these “conversions” turn into paying customers? With products that have a longer sales cycle, such as SaaS platforms with a 30-day trial or complex B2B services, the data between first contact and an actual deal is easily lost. But what if you could tell the advertising systems exactly which of these leads are genuinely good and have the potential to pay?
A lecture at the PPC Camp conference presented a real-life example: deploying lead scoring in a Performance Max campaign reduced the cost per acquisition (CPA) by 28% and increased the conversion rate by 23.8%. Sounds like a clear success, doesn’t it? But there is a catch that is not usually talked about: the overall return on investment (ROI) paradoxically fell by 28%. The reason? The model was not calibrated correctly and optimised for user activity in the free version, not for actual orders. And this very lesson is key to a successful implementation.
Summary for Those Who Don’t Have Time to Read the Whole Article
- What lead scoring is and why you need it: It is a points system that rates the quality of leads. With a longer sales cycle it is crucial, because it distinguishes mere sign-ups from genuinely valuable contacts.
- Technical connection to Google Ads: Data about user behaviour (events) is sent to BigQuery, where it is assigned a score. It is then sent back to Google Ads as a conversion value via Google Ads Data Manager.
- Setting up the scoring model: Track the key actions that signal interest (uploading a product, entering a phone number, a company ID) and assign them a monetary value. Give the main conversion the highest value.
- Real results and lessons learned: A correct setup can reduce CPA by tens of percent. But a badly calibrated model can lead to a drop in ROI if it optimises for the wrong signals.
- Key pitfalls: Implementation is technically demanding — expect it to take twice as long as estimated. GCLID is significantly more effective (up to 60% more data) than matching via email.
Why Optimising for Sign-Ups Is Not Enough
With products that have a longer decision cycle — whether a SaaS with a free trial, B2B enquiries or freemium models — a sign-up is only the first step. It is a signal of interest, not of business success. The problem is that algorithms such as Smart Bidding in Google Ads do exactly what you tell them. If your goal is to maximise the number of sign-ups, the system will bring you the people who are most likely to sign up. But those may not be the ones who eventually pay.
Without lead scoring you are essentially feeding the Google Ads algorithms the wrong data — a large volume of conversions that have no real business value. The result? Campaigns learn to bring in more and more similar but ultimately low-quality leads. You waste budget on acquiring users who will never go through the whole purchase process. Getting a sign-up is just the beginning; what follows is the key conversion rate optimisation at every stage of the customer journey.
What Lead Scoring Is and How It Changes PPC Campaign Performance
Lead scoring is a method in which you assign potential customers (leads) a points or monetary rating based on their behaviour and characteristics. The goal is to identify which leads are closest to buying and have the highest potential value. The key difference is in what you do with this data. Many companies use lead scoring only as an internal reporting tool for the sales department.
The real power, however, is unlocked only when you connect this system back to the advertising platforms. The principle is simple: based on their actions (e.g. activity in the app, downloading a price list, filling in a company ID), you assign each lead a specific score, which you convert into a monetary value. You then send this value back to Google Ads as a conversion value. Smart Bidding then no longer optimises for mere sign-ups, but for genuinely valuable leads that show buying signals. The whole process requires close cooperation between the PPC specialist and the web analyst.
How to Technically Connect Lead Scoring to Google Ads
A correct technical setup is the foundation of success. The architecture of the whole solution usually looks like this: Data layer (collecting events on the website/in the app) → Data warehouse (e.g. Google BigQuery) → Google Ads Data Manager → Google Ads.
Connecting data from your internal database with advertising systems is an ideal job for Google BigQuery, which can process large volumes of data and connect information about user behaviour with the user’s unique identifier. From practice we know that relying only on email matching is not enough. Implementing the collection of the GCLID (Google Click Identifier) brought 60% more matched data. While Google Sheets often failed with daily data imports, Google Ads Data Manager proved to be a significantly more stable solution.
What will you need? Prepare for cooperation between the data team, the web analyst and the PPC specialist. A robust data layer, correct GCLID collection and ideally the deployment of server-side measurement are a must. The whole data collection process is today best handled through server-side Google Tag Manager (SGTM), which ensures higher reliability and accuracy. You can find more on this topic in our article on server-side tracking.
How to Set Up the Scoring Model — and Why the First Version Is Always Just the Beginning
The most important step is to define which user actions signal genuine interest and what value to assign to them. Don’t try to be perfect from the start. Begin with a simple model based on the most important events. Here is a real example from practice for a SaaS product:
- Uploading a product into the system: CZK 15/day (maximum CZK 450 over 30 days) — a signal of active use of the trial.
- Entering a phone number: CZK 30 — a signal of serious interest and willingness to be contacted.
- Entering a company ID: CZK 15 — a signal that this is a business customer.
- Ordering a paid plan: CZK 1,800 — the main conversion action.
- Maximum value of a single lead: CZK 2,295.
And here comes the key lesson mentioned in the introduction. If you assign the same high value to all orders, including activation of the free version, the system will logically start optimising for the easiest path to a high score — activating users in the free version who will never pay. That is why ROI fell by 28%. The solution is to discriminate against low-value conversions, for example by assigning them a significantly lower score, or by not including them in the model at all. Lead scoring is an iterative process. Deploy the first version, measure the impact, evaluate it and adjust the model. And be prepared for the real time and financial cost to probably be double the original plan.
Want results like −28% CPA? Book a free introductory consultation with us.
Final summary: Lead scoring is not just a function in a CRM system for salespeople. It is a strategic tool that can fundamentally change the performance of your PPC campaigns, especially if you work with a longer sales cycle. It does require an initial investment in data infrastructure and close cooperation across teams, but it brings measurable results in the form of higher-quality leads and a lower cost per real acquisition. Don’t start grandly — deploy a first, simple version, measure its impact carefully and improve it gradually.
Frequently Asked Questions
What is lead scoring?
A points system that rates the quality of potential customers (leads) based on their behaviour and characteristics. The goal is to distinguish promising contacts from less valuable ones.
Is lead scoring worthwhile for smaller companies too?
It depends on the length of your sales cycle and the volume of leads. If you have more than 100 leads a month and the conversion from first contact to purchase takes weeks or months, lead scoring will very likely pay off.
What tools do I need to connect lead scoring to Google Ads?
The basis is a data store such as Google BigQuery, plus Google Ads Data Manager for transferring the data, server-side Google Tag Manager (SGTM) and a quality data layer for collecting the necessary events.
How long does implementing lead scoring take?
Count on 2–3 months. The technical solution is often more demanding than it seems at the start. Real experience shows it is wise to multiply the planned time by two.
Sources
- Kubíková, B. (2025). Lead scoring for Shoptet version 2.0. Lecture at PPC Camp. YouTube — publicly available