This checklist gives B2B teams the website analytics best practices they need, in the order they should tackle them. Each recommendation stands alone, with practical instructions and the consequences of skipping it. Work through the list from the beginning, or use these web analytics tips to investigate specific problems. Agencies can also use it to audit a new client's setup.

1. Write the questions before you install anything

Before you configure any tracking, lay out the questions you want your analytics to answer.

When you agree on the questions upfront, you have a clear purpose, and everyone can focus on the same goals. It also makes it more likely you’ll measure the right information to evaluate your performance.

Start with specific decisions your team makes regularly, such as identifying which campaigns bring relevant visitors or which pages encourage inquiries, and write down the questions you need answered. Agree what you'll need to measure for each question before configuring anything, using our web analytics strategy guide for the wider planning process.

Skip this step, and you could end up tracking hundreds of events without being able to answer the questions that prompted you to collect the data.

2. Instrument only what the measurement plan needs

Configure event tracking only for actions that help answer a question in your measurement plan.

This keeps your events focused and avoids superfluous detail. It also makes testing easier, as everyone knows which actions matter and what each event is supposed to record.

Review the questions you set earlier and identify the actions that need tracking, like a completed inquiry form or a click on an important website button, and leave out interactions with no clear purpose. Set up the relevant events in Google Tag Manager, then use its preview mode to perform each action yourself and confirm that the correct tag fires.

If you don’t do this, you'll spend time maintaining events nobody uses, while cluttered reports make important activity harder to spot.

3. Agree one naming convention and write it down

Agree on strict naming conventions for everything you track, then document the rules somewhere everyone can find them.

Naming rules mean everyone records the same action consistently, even when multiple people manage the tracking. It also keeps your reports organized, so you can easily compare results over time.

Create a shared naming register that specifies capitalization and separators, with examples showing exactly how event names should appear. When configuring events in Google Tag Manager, check each name against the register and update the document whenever your team agrees on a change.

Otherwise, the same action could appear under several different labels, splitting the figures across multiple report rows and leaving your team struggling to work out which numbers to use.

4. Exclude internal and bot traffic on day one

Make sure your website reports don’t include visits from your own team or automated traffic.

Your team will regularly visit your site to test forms or update content. All those visits can inflate your traffic figures, while bots can make it even harder to see how real visitors behave.

Start in Google Analytics 4, where you can identify internal traffic using your company's known IP addresses, then test the rule before activating the data filter to make sure you're excluding the right visits. Check your reports for remaining bot activity, although GA4's automatic filtering won't catch every bot, so some unwanted traffic may continue.

If you skip this, you might count your own employees and bots as potential customers, leading to misleading conversion rates and unreliable comparisons between reporting periods.

5. Decide how you record consent state, and report the gap rather than ignoring it

Record how visitors respond to your cookie consent options and account for how it affects your analytics.

Some visitors will decline optional tracking, so you won't capture all their activity in your reports. Recognizing those gaps helps you interpret your B2B website traffic and campaign results more accurately. This matters more on a B2B site, where traffic is lower than on a consumer site, so a small consent gap can hide a real change in a segment such as visitors from one target industry.

Use your existing consent setup to record whether visitors accept or decline tracking, and document how each choice affects the events your analytics collect. When reporting performance, explain where the figures may be incomplete, using our guide to cookieless website analytics for more detail.

Ignore these gaps, and you could end up comparing figures from different periods without realizing that changes in tracking coverage are skewing results.

6. Track form starts as well as form completions

Track when someone starts filling in a form on your site, as well as when they submit it.

A completed form tells you someone made it to the end, but misses anyone who gave up halfway through. Comparing starts to completions shows how much interest drops off before they click submit.

In Google Analytics 4, set up separate events for form starts and successful submissions, using Google Tag Manager if the forms need additional tracking. Test both actions yourself to make sure each triggers the right event, then compare starts and completions over the same reporting period to see where people may be abandoning forms.

If you only count submissions, a poorly constructed form can keep losing interested visitors for months without anyone spotting the problem.

7. Split branded and non-branded organic from the start

Separate out searches for your company name from those where people are looking for the products or services you provide.

People searching for your brand often already know who you are. Looking at non-branded searches separately helps you see how well your website attracts people who may be discovering your company for the first time.

In Google Search Console, review the queries that bring people to your site and identify searches that include your company name, including common variations and misspellings. Document which terms count as branded, then compare their performance with non-branded queries over time, using our seo web analytics guide for more detail.

If you lump everything together, a surge in branded searches could hide falling visibility among potential new customers.

8. Tag every campaign with the same UTM rules, without exception

Use identical UTM naming rules for every campaign link, and we mean every. A free UTM builder applies the same format to every link, so nobody has to remember the rules.

Consistent tags keep campaign results together, making reports easier to read and compare. They help you judge which campaigns attract companies that match your ideal customer profile.

Write down your rules for source and medium, plus campaign names (including capitalization and separators), then apply the same naming rules to the utm parameters on every campaign link. In Leadfeeder Web Insights, the Campaigns report shows one row per campaign and counts how many companies it brought to your site and how many fit your ICP.

If you skip these rules, one campaign can appear across several rows under different names, so your figures don't add up, and nobody can confidently use your report.

After a week of consistent tagging, open the Campaigns report and check that each campaign appears on one row.

9. Define conversions by type, never as a single total

Report each type of website conversion separately instead of combining everything into a single number.

For example, a completed inquiry form means someone contacted your business, while a visit to your demo page shows a different level of interest. Keeping these actions separate helps you see which campaigns drive inquiries and which encourage visitors to learn more.

Choose the website actions you want to count, such as form fills or demo-page visits, and give each conversion type a clear name so everyone knows what it covers. In Google Analytics 4, configure the relevant key events and compare each conversion type across campaigns and reporting periods, rather than relying on an overall total.

Otherwise, a jump in demo-page visits could make your performance look better, even if completed inquiries are falling.

10. Track your high-intent pages separately

Keep a closer eye on the pages that suggest visitors are seriously considering buying from you.

Someone reading your pricing, documentation, integrations, or security pages may be further along in their buying research than someone browsing your blog. For B2B marketers, monitoring this activity separately helps you spot changes in buying intent faster and respond accordingly.

Create a group of URLs for these high-intent pages in your analytics setup, then monitor their traffic and engagement separately from the rest of your website. Track which pages attract more attention over time and look for changes in visitor behavior, using our guide to website visitor intent for more detail.

If you treat all pages equally, you could miss valuable signs that potential customers are actively evaluating your products.

11. Compare like segments, not site totals

When assessing performance, make sure you’re comparing the same groups of website visitors.

Site-wide averages can hide big differences between channels, especially when one starts bringing in a larger share of traffic. For example, a rise in overall engagement may mean more visitors are arriving from a channel that already performs well.

In Google Analytics 4, choose a channel or device category and compare its results with the same group over an equivalent reporting period, keeping the conditions consistent. If the headline figure has changed, check whether that movement also appears within the individual groups before deciding that anything significant has happened.

Otherwise, you could double down on an apparent improvement or spend time fixing a supposed problem when all that's changed is the mix of people visiting your site.

12. Use a comparison window that matches your sales cycle

Select a reporting period that reflects how long your customers typically take to make a buying decision.

B2B buyers often spend months researching their options, especially when several people are involved in the decision. Monthly reports alone can give you a misleading impression of performance because they capture only a small part of that process.

Use your CRM to identify your typical sales cycle, then adjust your reporting window accordingly, using a quarter as a starting point. Keep that window consistent when comparing results over time, although you can still use shorter periods to check how individual campaigns are performing.

Sticking to monthly figures may lead you to react to short-term changes while missing the bigger picture of how interest develops over the buying cycle.

13. Work out the size of normal variance before you call a change

Check how much your website metrics typically fluctuate before deciding that a change in performance needs attention.

Traffic figures naturally rise and fall, even when your marketing activity stays the same. A sudden jump or drop might look alarming, but it could be perfectly normal for your website.

Look back at several comparable reporting periods and record how much your key metrics typically vary, so you have a reasonable range to measure new results against. When a figure falls outside that range, investigate what happened and check whether anything unusual, such as a major campaign launch, affected the earlier figures you're using for comparison.

Without this baseline, you could waste hours investigating perfectly ordinary fluctuations while missing changes that deserve a closer look.

14. Never report a number you cannot explain the movement in

Investigate any substantial change in your website figures before you share it with your leadership team.

A sudden rise in traffic might look like good news, but your colleagues will want to know what's behind it. If you can't explain the movement, they may draw conclusions the numbers don't support.

Start by comparing the metric with the same period you normally report on, then look at recent campaign activity for anything that might explain the difference. Check whether anyone changed the tracking setup, and make it clear which explanations you can confirm and which still need investigating before you write up your findings.

Don’t risk increasing spending or changing a campaign based on a movement caused by a tracking error.

15. Segment by company, not only by session

Segment your website traffic by company to see whether you're attracting the businesses you want.

Sessions tell you how often people visit, but they don't reveal whether those visits come from your ideal customers. Company-level reporting also helps agencies show clients whether their campaigns are reaching the right businesses.

This is part of the broader practice of website intelligence: reverse ip lookup resolves each visit to an organization, not a person. In Web Insights, the Firmographics report groups companies by ICP segment, including Non-ICP, industry, size, and country. Each segment shows company counts, the ICP ratio (the share of identified companies that fit your ideal customer profile), and average Intent Score, so you can compare ICP fit across segments and see where relevant interest is concentrated.

Without this view, a busy website could look successful without attracting enough potential customers.

Install the tracking script this week, because company data starts on installation and cannot be added retrospectively. Try to ensure that next quarter’s review has a full quarter of company data.

16. Give every metric one written definition and one owner

Write down exactly what each metric means and make one person responsible for keeping that definition up to date.

A label such as “conversions” can mean different things to marketing and sales, even when both teams are looking at the same report. Giving someone ownership means there's someone in charge when how you measure something changes.

Create a shared metric register with the definition, data source, and named owner for each figure your team reports. Ask the owner to check the entry whenever tracking changes, using our web analytics glossary to keep the terminology consistent and avoid confusion between teams.

Leave definitions to individual interpretation, and you'll waste meetings arguing over figures that should have been agreed long ago.

17. Put the commentary above the chart, not below it

Put a short explanation above every important chart so readers know what they're looking at before they study the figures.

Most people scan reports, especially when they're busy, and may miss something important if they're left to interpret a chart themselves. Giving them the essential information upfront helps them understand what changed and why it matters to the business.

Write a sentence explaining the result and place it directly above the chart, making sure it tells readers something useful rather than simply describing the numbers they can already see. Use the chart to support your explanation, adding any relevant context or limitations, and see our web analytics report guide for more advice on presenting findings.

If you leave readers to work things out themselves, they may reach different conclusions from the same figures.

18. Audit the setup once a year against the measurement plan

Set aside time each year to check your analytics setup against the questions your team originally wanted answered, even if everything appears to be working properly.

Websites change as teams launch campaigns and revise priorities, so last year's tracking may miss something important. Regular audits catch those gaps before they affect the reports your colleagues use to make decisions.

Compare your current setup with the original measurement questions, then test whether important website actions still register correctly and whether new requirements need tracking. Assign someone to fix outdated configurations, document each change, and test the corrections before the next reporting cycle.

Skip this check, and broken tracking can go unnoticed, leaving your team making decisions based on unreliable figures.

19. Delete what nobody uses

Remove reports and tracking elements that your team no longer needs to make decisions.

Unused reports still take time to maintain, especially when someone has to check the figures each month. They also crowd out the information people actually look for, making your regular reporting harder to use than it needs to be.

Review your existing reports, see which ones people open, and ask their owners when they last used the findings to make a decision. Retire anything that's no longer useful, but keep the historical records you need for comparisons, and check our advanced website analytics guide when you're ready to do more with the data you keep.

Otherwise, your team could spend hours updating reports nobody reads while the work that matters gets less attention.