> ## Documentation Index
> Fetch the complete documentation index at: https://docs.leadpipe.com/llms.txt
> Use this file to discover all available pages before exploring further.

# Understanding data quality

> Judge Leadpipe data by usefulness, not by whether every field is filled

## Start with usefulness

Judge Leadpipe data by whether it supports the workflow you care about.

For some teams, a company, landing page, and recency signal are already useful. For others, the record needs a work email, phone number, or stronger firmographic context before it is actionable.

## What usually matters most

| Field or signal | Why teams care                                    |
| --------------- | ------------------------------------------------- |
| `Name`          | Helps your team understand who was likely on site |
| `Company`       | Ties the visit to an account or target market     |
| `Job title`     | Helps qualify role, department, and seniority     |
| `Emails`        | Makes direct follow-up possible when available    |
| `Phone`         | Supports sales workflows that call fast           |
| `Source`        | Helps marketing judge acquisition quality         |
| `Landing`       | Adds context about what triggered the visit       |
| `Recency`       | Helps teams act while the signal is fresh         |
| `Firmographics` | Helps narrow traffic to the right accounts        |

## How to think about strong versus partial records

<AccordionGroup>
  <Accordion title="Strong record">
    A strong record has a clear identity, strong context, and enough detail to support action right away.
  </Accordion>

  <Accordion title="Partial record">
    A partial record can still be useful for prioritization, reporting, suppression, or audience building.
  </Accordion>
</AccordionGroup>

## Why quality can vary

Quality depends on your traffic mix, your market, and how tightly you define what useful looks like.

For example:

* clean B2B traffic is usually easier to operationalize than broad consumer traffic
* filtered traffic is easier to trust than unfiltered traffic
* a sales workflow may need richer records than an ads or reporting workflow

## How to improve the quality of what your team sees

* Suppress existing customers, competitors, internal traffic, and bad-fit segments
* Exclude low-value pages if they create too much noise
* Save views around your actual ideal customer profile
* Route only the highest-signal slice into your first workflow

## What not to do

* Do not judge the product by the noisiest traffic on your site
* Do not push everything into downstream tools on day one
* Do not assume every workflow needs the same level of data completeness

## Next steps

* Read [Suppress unwanted visitors](/guides/suppress-unwanted-visitors) to control noise and spend.
* Read [What to do with identified visitors](/guides/what-to-do-with-identified-visitors) to choose the right use case for the data you have.