KPIs for Website Traffic: Visits, Engagement, and Leads

KPIs for website traffic should connect visits with engagement and business outcomes, rather than stop at a rising session count. Track source-level session growth, engagement rate, lead-session conversion, and qualified lead yield using consistent periods. Give each measure an owner who can act when performance changes.

Choose measures that answer different questions

Traffic volume answers whether acquisition is bringing visits. Engagement answers whether those visits meet the source system's activity criteria. Lead outcomes answer whether the visits produce inquiries that fit the business's needs.

These are distinct questions. A support article can satisfy a reader without generating a sales inquiry. A product landing page may be expected to generate one. Segment by page purpose before judging either page against the same target.

A KPI differs from a metric because it has an accountable owner, a target, and a decision. Keep measures that diagnose a problem available without making every traffic count an executive KPI.

Keep the export definitions intact

Begin with an approved traffic export containing period, session source, landing page, sessions, and engaged sessions. Add counts of sessions with valid inquiries. For a commercial quality measure, bring a lead register summarized to the same acquisition cohorts using documented qualification rules.

Google Analytics describes a session as a period of website or app interaction. The same person can contribute multiple visits, so sessions are not a distinct-person count. Keep the source's session settings consistent when comparing periods.

Google's engagement documentation defines engaged sessions using duration, key-event, or page-view criteria. Preserve the exported definition and its settings instead of substituting an informal idea of an interested visitor. Covirage is not affiliated with Google.

Use session source throughout this comparison. A first-acquisition source and the source assigned to a current visit answer different questions. Do not join them under one unlabeled source column. Likewise, compare landing pages as entry pages, rather than mixing them with every page viewed during a session.

Worked figure: growth by traffic source

The following fictional US industrial supplier, Cedar Ridge Components, compares October 1-7 with October 8-14, 2026. These equal reporting windows use the same traffic definitions. For current-period acquisitions, lead outcomes are measured within seven days of each session, so qualification observation is complete by October 21, 2026.

Each valid new lead in this example comes from a different session, and each qualified lead is a subset of those valid leads. The rows exhaust the company's included traffic scope; there is no omitted source category in this illustration.

Session growth = (Current sessions − Prior sessions) / Prior sessions × 100

Engagement rate = Engaged sessions / Sessions × 100

Lead-session conversion = Sessions with a valid lead / Sessions × 100

Qualified lead yield = Qualified new leads / Sessions × 100

Growth needs the prior-period denominator. The other measures below use current-period sessions. All displayed percentages are rounded to two decimal places; totals use the original counts.

For organic search, growth is (3,600 − 3,000) / 3,000 × 100 = 20.00%. Engagement is 2,160 / 3,600 × 100 = 60.00%; lead conversion is 108 / 3,600 × 100 = 3.00%; qualified yield is 54 / 3,600 × 100 = 1.50%.

For the company, sessions increase by 7,500 − 6,000 = 1,500, giving 1,500 / 6,000 × 100 = 25.00% growth. Current engagement is 3,990 / 7,500 × 100 = 53.20%. Lead conversion is 252 / 7,500 × 100 = 3.36%, and qualified yield is 99 / 7,500 × 100 = 1.32%.

What the traffic owner should investigate

Paid campaigns contribute 3,000 − 2,000 = 1,000 additional sessions, organic search adds 3,600 − 3,000 = 600, and email loses 1,000 − 900 = 100. The bridge is 1,000 + 600 − 100 = 1,500, matching the total increase.

Paid visits are growing fastest, but organic search and paid campaigns have the same lead-session conversion rate. Paid traffic's lower qualified yield directs the next review toward audience fit and lead quality. It does not prove the paid channel caused weaker qualification; the campaigns may reach different markets or product lines.

Email has fewer sessions and stronger current conversion. Investigate whether comparable campaigns were sent and whether the audience changed before expanding activity. The table contains no previous-period lead outcomes, so it cannot establish that total leads or qualified leads improved over time.

Continue at page level with landing page KPIs, where conversion quality and cost are worked together. For the broader management review, choose a small set from the KPI examples by team and retain these source measures as supporting detail.

Checks that prove the review

First, the current sources sum to 3,600 + 3,000 + 900 = 7,500 sessions. Match that to the source export for exactly the same filters. If the source includes an unassigned category, keep that category in the real reconciliation.

Second, engagement has a useful identity:

Each remainder is total sessions minus engaged sessions. Across the company, 3,990 + 3,510 = 7,500. Under the complementary definition cited above, bounce rate is 3,510 / 7,500 × 100 = 46.80%, and 53.20% + 46.80% = 100.00%.

Finally, reconcile valid and qualified lead counts to the approved register after exclusions. Confirm that a visit is counted once within its source cohort and that lead sessions do not exceed sessions. These checks matter more than the visual layout when choosing dashboard software.

Where website traffic KPIs go wrong

• Averaging source percentages. Total rates need total numerators divided by total denominators. A small email audience should not receive the same weight as a larger source.
• Replacing missing outcomes with zeros. A missing qualification export is incomplete evidence, not proof that a channel produced no qualified leads.
• Mixing incompatible periods. Unequal reporting windows, holidays, and incomplete current days can move visits without changing acquisition performance.
• Ignoring page purpose. Support and recruiting traffic should not automatically be judged by the inquiry goal assigned to product pages.
• Losing unmatched rows. Incomplete source-to-lead matching changes coverage. Show that gap before interpreting channel differences as business performance.

Review your traffic exports in Covirage

Bring your traffic-by-source and landing-page export with the matching aggregate lead outcomes to Covirage. Covirage maps the columns, its tools compute and check the comparisons, and the external AI model explains the results with every figure traced to its rows. Create a free Covirage account, which takes about a minute, then set up your business and bring a sample of your own data.

See self-service analytics for business teams for the supporting approach.