Lead scoring that shows which inquiry counts first.
And steers advertising toward closed deals.
Dometrics builds lead scoring models from a company's own sales data: every inquiry receives a transparent score in the CRM, sales calls the right contacts first, and Google Ads and Meta learn to optimize for qualified leads instead of form submissions.
- Managing director · Mechanical engineering, 180 staff Quote request · Upper Austria · 6 min ago 92 Now Industry fitsViewed pricingBudget stated
- Procurement · Food industry, 420 staff Callback requested · Vienna · 22 min ago 78 Today Company sizeSecond visitRead case study
- Marketing · Small business, 6 staff General question · Salzburg · 1 hr ago 34 Nurture Industry differsNo timeframe
- Private individual Contact form · 2 hrs ago 12 Decline No purchase intent
Plenty of leads, little clarity.
In B2B sales, the number of inquiries is rarely the problem. The problem is that every inquiry gets the same treatment, even though their value differs by orders of magnitude.
Sales works in order of arrival
First in, first called. The promising inquiry from Friday evening waits until Monday lunchtime, while time flows into contacts who will never buy.
Marketing counts forms
Success is measured in leads, not in orders. Campaigns that deliver many cheap but weak inquiries look best in the report.
Ad platforms learn the wrong thing
Google Ads and Meta optimize for the signal they receive. If that signal is a submitted form, they find people who submit forms, not customers.
Knowledge lives in gut feeling
Experienced salespeople spot good inquiries intuitively. That knowledge is undocumented, not transferable and leaves the company with the employee.
From the form all the way to the order.
The model learns from the company's own history: which characteristics did the inquiries that turned into orders share? The result flows in two directions, into sales and back into the ad accounts.
- 01 · Data
Connect sources
Form inputs, website behaviour, firmographic data and the CRM history with the outcome of every past inquiry.
- 02 · Model
Learn patterns
Fit with the customer profile and buying signals are weighted and validated against inquiries with a known outcome.
- 03 · CRM
Score with reasons
Every new inquiry receives a score and the three most important reasons behind it, directly in the CRM.
- 04 · Advertising
Feed quality back
Qualified leads and closed deals are sent to Google Ads and Meta as offline conversions with value.
- 05 · Control
Refine
Quarterly review: did high scoring leads actually buy more often? The model is adjusted accordingly.
Point system versus data-driven scoring.
Many CRM systems offer a simple point system. It is a good start, but it reflects what someone assumes, not what the data shows.
| Criterion | Manual point system | Data-driven lead scoring |
|---|---|---|
| Basis | Assumptions: job title 10 points, download 5 points | Weights derived from real past deals |
| Updates | Rarely revised, slowly goes stale | Retrained regularly and checked against outcomes |
| Explainability | A total without context | Score plus the key reasons for each lead |
| Ad accounts | Remain untouched | Receive qualified leads as a conversion signal |
| Verifiability | Impact unknown | Hit rate per score class is measured continuously |
How lead scoring works and what it requires.
Lead scoring answers a simple question: which inquiry deserves attention first? In sales, that order determines close rates. In marketing, it determines whether budget flows into reach or into revenue. Good scoring makes both visible and connects the two departments through the same number.
Fit and buying signals
Every score has two dimensions. Fit describes how closely a contact matches the ideal customer: industry, company size, region and the role of the person making the inquiry. Buying signals describe how concrete the interest is: the type of inquiry, pages visited such as pricing or case studies, requested quantities or dates. A contact with high fit and a weak signal needs nurturing. One with a strong signal and weak fit deserves an honest no. High on both dimensions means: call now.
Why data beats gut feeling
Traditional point systems assign points based on assumptions. A data-driven model derives the weights from the company’s own history: which characteristics did the inquiries that actually became orders share? This often reveals that characteristics which seem intuitively important say little, while inconspicuous signals carry real weight. The model is not a black box. For every score, the CRM shows the most important reasons, so sales can see why an inquiry sits at the top.
Good to know: A reliable model needs a history with known outcomes. Without one, scoring starts with clear rules based on the sales team’s knowledge. From day one, the outcome of every inquiry is then documented properly, so the model can take over later.
The second lever: reporting quality to the ad accounts
The biggest effect often appears not in sales but in the ad account. Google Ads and Meta optimize for the conversions they receive. If the signal is a submitted form, the algorithms reliably find people who submit forms. If qualified leads and closed deals are reported back with their value instead, budget shifts toward the search terms and audiences that actually produce orders [2][3]. This requires clean conversion tracking, so every inquiry can be attributed to its campaign.
| Reported signal | What the platform optimizes for | Typical result |
|---|---|---|
| Form submitted | Many inquiries, as cheap as possible | Low CPL, fluctuating quality |
| Qualified lead | Inquiries that fit the customer profile | Fewer but better leads |
| Closed deal with value | Inquiries that generate revenue | Budget follows order value |
Legal framework
Lead scoring processes personal data and therefore requires a documented legal basis and transparent information for the data subjects. As long as a human decides whether to make contact, it is generally not a decision based solely on automated processing under Art. 22 GDPR [1]. The assessment of each individual case belongs with the data protection officer. Dometrics provides the technical documentation.
Key takeaway: A score is only as good as what happens next. Only when sales and the ad accounts act on it does a number become a revenue lever.
How high scoring inquiries turn into quotes faster is covered in automated quote creation. A related application built on the same data foundation is AI revenue forecasting.
What lead scoring includes.
- Data analysis: which characteristics separate won from lost inquiries
- Cleansing and standardization of the CRM history
- Scoring model covering fit and buying signals, explainable per lead
- Integration of the score into the existing CRM
- Sales alerts for particularly high scoring inquiries
- Offline conversion import with values into Google Ads and Meta
- Dashboard with hit rate per score class and cost per qualified lead
How the project runs
Data check
Review of the available CRM data. Is the history sufficient for a model, or does scoring start with clear rules first?
Model
Building and validating the scoring on past inquiries with a known outcome.
Integration
Score and reasons in the CRM, feedback to the ad accounts, alignment of sales processes.
Control
Quarterly comparison of score and outcome, retraining and adjustment of thresholds.
Frequently asked questions
What is lead scoring?
Lead scoring rates every inquiry by how likely it is to turn into an order. The score usually combines two parts: fit with the ideal customer profile, such as industry, company size and region, and buying signals such as the type of inquiry or behaviour on the website.
How much data does AI-based lead scoring need?
As a guideline, a few hundred inquiries with a known outcome, including a sufficient number of won deals. With less data, Dometrics starts with rule-based scoring built on the sales team's knowledge and switches to a model as soon as enough history is available.
Which CRM systems can be connected?
All common systems with an API, such as HubSpot, Pipedrive, Salesforce or Microsoft Dynamics. Industry-specific systems and ERP solutions can also be connected, provided they make data available via an API or a regular export.
What does lead scoring do for Google Ads and Meta?
The platforms optimize for the signal they receive. If only qualified leads or real deals with value are reported back instead of every form, bidding strategies shift budget toward the audiences and search terms that actually generate revenue.
Is lead scoring compliant with the GDPR?
Yes, if it is set up properly. Lead scoring used to prioritize sales work is generally not a decision based solely on automated processing with legal effect within the meaning of Art. 22 GDPR, because a human makes the final call. Information obligations still apply, and the legal basis should be documented. The legal assessment of each individual case belongs with the data protection officer.
How can companies verify that the scoring works?
Through the hit rate per score class. If the model works, high scoring leads buy significantly more often than low scoring ones. This analysis runs continuously in the dashboard and forms the basis of every retraining.
