4SYSTEM ACADEMY
4SYSTEM
Contact us
4SYSTEM

Lead quality: linking form, CRM and commercial results

Written by 4SYSTEMPublished on Updated on

8 min read

The quality of leads in a CRM is measured based on shared definitions and tracking of decisions, not just the number of forms. Distinguish between received request, valid request, qualified contact and opportunity, then note the reason and the person responsible for each visit. Link these states to a documented source of arrival when possible, without confusing technical attribution and declared origin. Also examine the files still pending: an unprocessed request is not a poor quality lead. This method allows you to know if the problem is with targeting, offer, data or support.

In this guide

  1. To remember: quality depends on an explicit rule
  2. Define CRM lead quality in stages
  3. Separate reasons for non-qualification
  4. Link the form to the file without losing the provenance
  5. Separate sales contact and marketing subscription
  6. Table of definitions to be validated between teams
  7. Compare cohorts that have had time to be treated
  8. Fictional example: a campaign accused of bad requests
  9. Put the findings at the service of a correction
  10. Frequently asked questions
  11. Next step: harmonize the patterns

To remember: quality depends on an explicit rule

The same request may seem interesting to marketing and be considered too imprecise by salespeople. Without common criteria, the report compares opinions rather than states. HubSpot distinguishes between marketing-qualified leads and sales-qualified leads.[45] This supplier example helps separate decisions, without imposing a universal nomenclature.

Define CRM lead quality in stages

Start with receipt: has the system actually created a folder? Validity then describes the respect of minimum criteria, such as an intelligible need and an exploitable means of response. Qualification requires a business decision on suitability for the offer. The opportunity corresponds to a business milestone defined by your team.

For each state, write an entry condition, a responsible person and a proof. Avoid “hot” or “interesting” labels without criteria. They can be useful in a conversation, but become difficult to compare when populating a table.

Also provide waiting and closing states. “To be examined”, “necessary information” and “not accepted” must not disappear in favor of a binary choice. An incomplete application may still be relevant; a complete file may remain out of offer. The method must maintain these distinctions.

Separate reasons for non-qualification

Create understandable reasons: outside scope, unusable data, duplicate, unsuitable project, lack of response after processing or request not yet examined. Don't mix a technical error, a business decision and an operational delay.

Define the use of each pattern. “Outside scope” can guide clarification of the page or targeting. “Unusable data” may call for a form check. “Not reviewed” refers to capacity and organization. Not everyone justifies reducing an acquisition budget.

Keep the note short when the category is not enough, but limit the information to what is useful and appropriate. A marketing report does not need a full copy of commercial transactions. Have access and purposes validated by the people responsible for the data.

Link the form to the file without losing the provenance

Use a request reference to track its transmission and creation. A single email address can associate several distinct needs with the same person. Preserve case history and document merge rules, rather than overwriting the previous request.

Check the original fields. A source declared by the person, a campaign parameter and analytical information answer different questions. Keep their labels separate and their boundaries visible. The “unknown” value is preferable to an origin invented during an import.

The measurement plan guide GA4 of 4SYSTEM helps distinguish site progress and results. The GA4 reference defines a key event rate per session as the share of sessions having this event.[54] This rate does not automatically qualify the contacts created subsequently: the CRM must retain its business decision.

Separate sales contact and marketing subscription

A quote request and a subscription are not the same purpose. The CNIL recalls the conditions of electronic prospecting according to the public, in particular the prior consent of individuals in the general case and the rights of opposition.[4] Have the basis of each use verified without transforming a qualification into a campaign authorization.

A contact can be qualified for the processing of their need and opposed to the newsletter. Your fields and selections must represent these two states. Automation that activates marketing with every sales pass loses this distinction.

Do not submit personal email addresses or comments in the settings for analytics reporting. Prepare authorized correspondence and useful aggregated information with your technical and data manager. The goal is to understand the journey, not to move the entire business case into each measurement tool.

Table of definitions to be validated between teams

Condition or reasonProposed definition to be adaptedResponsiblePossible decision
Request receivedFolder created in intended systemIntegration ManagerCheck transmission and errors
Valid requestMinimum usable informationFirst Reading TeamProcess or request clarification
Marketing QualifiedAgreed relevance criteria metMarketing according to the processTransmit with useful context
Sales QualifiedNeed recognized as relevant by salesSales managerInitiate next defined step
OpportunityFile entered into the proposal processSales teamFollow proposal and decision
Out of offerIncompatible scope documentedPerson who examinesClarify page or targeting if necessary
Not reviewedNo business decision yet renderedFolder ownerFix support
ClosedFinal reason known and recordedResponsible for the fileAnalyze without clearing the provenance

HubSpot categories illustrate some distinctions, but your process may be simpler.[45] Do not create a step without use. Each statement should help a person decide, and each passage should be explainable in a review.

Compare cohorts that have had time to be treated

Group requests by their entry period or another stable rule, then examine their maturity. A recent cohort naturally contains files without a decision. Comparing it directly to old closed files may suggest a deterioration in quality.

Keep volume, pending states, and criteria visible. If the qualification definition changes, note the break and avoid attributing any deviation to the channel. If the team lacks capacity, examine the treatment before concluding that the acquisition provides poor contacts.

Only calculate a rate when the numerator and denominator are compatible. Qualified Contacts, Requests, and Sessions are different units. GA4 also distinguishes between rates per session and per user in its metric reference.[54] A column of percentages does not become comparable because the labels all contain the word "conversion".

Fictional example: a campaign accused of bad requests

An imaginary company observes more forms from a campaign. The CRM indicates few qualified contacts. Before scaling back the campaign, the team reviews the statuses and discovers that several requests have not been assigned, while another group is actually out of scope.

In this illustrative scenario, files without a manager are classified “not examined” and processed separately. Out-of-offer requests are reread with the landing page to look for an ambiguous promise. A request that arrives twice remains a documented duplicate, not two poor quality contacts.

The team then sets up a review of the reasons and a next action for each valid file. It does not claim that the campaign is profitable or that quality has improved immeasurably. The example shows how to separate organization, targeting and collection before deciding.

Put the findings at the service of a correction

Choose an action that matches the documented dominant pattern. An ambiguity of scope requires clarification of the offer. A transmission error requires a technical correction. A lack of treatment requires an owner and a recovery process. Don’t prescribe the same “form optimization” for all cases.

Prepare a validation test and a reading period adapted to decisions. A corrected form can be checked immediately on a fictitious file; commercial qualification requires the processing of actual requests according to the criteria. This evidence does not arrive at the same time.

Keep contradictory findings. A channel can produce files outside of supply and relevant needs. A result that is too weak to conclude must remain described as such. Lead quality is a reading of the process and decisions, not a definitive judgment on people.

Frequently asked questions

Should we use an automatic lead score?

Only if the criteria, data and usage are understood and verified. A score can help prioritize, but does not replace professional qualification or the search for unprocessed files. Start with explicit states and reliable patterns. An opaque number makes disagreements more difficult to correct.

Is an unconverted request necessarily of poor quality?

No. The need may be relevant without leading to a sale, or the processing may be incomplete. Distinguish between suitability for the offer, maturity, commercial decision and response capacity. Keep known reasons and files still pending. The absence of a sale does not automatically reveal a defect in the targeting or the form.

Can we directly compare two channels?

Compare their populations, periods, maturity and criteria before reading the results. Keep origins unknown and processing differences visible. The same qualification rule must be applied, and files not examined must not be assimilated to refusals. If the volumes or provenance are insufficient, limit the conclusion.

Next step: harmonize the patterns

Bring marketing and sales together around fictitious requests or properly protected cases, then verify that they assign the same states. The 4SYSTEM services can help frame connections and definitions. Correct provenance and support first before classifying channels.

Sources

EN