This conversion rate optimization checklist is a working sheet for the moment a B2B site has traffic but too few demo requests. It covers 48 checks in seven groups, from tracking to sales follow-up, and every check has a pass condition you can verify in under ten minutes. Copy the tables into a spreadsheet, or download the template as CSV.
The checklist is the "what to look at" layer. For the process around it (how to schedule a one-week audit, segment the data and score findings), read our guide to running a CRO audit. The two are meant to be used together. For where an audit fits in a wider optimization program, see the conversion rate optimisation guide.
How to Use This Checklist
- Pick three to five pages. The homepage, the main product page, pricing, the demo request page and your highest-traffic landing page cover most B2B pipeline. Do not audit the whole site.
- Run tracking checks first. If group 1 fails, every number you look at afterwards is wrong. Fix tracking before you judge any page.
- Mark each check Pass, Fail or Unknown. "Unknown" is a finding in its own right: it means nobody owns that part of the funnel.
- Write one line of evidence per Fail. A screenshot, a number from analytics or a quote from a session recording. Findings without evidence get argued away in the review meeting.
- Score and prioritize. Use the scoring sheet at the end to decide what to fix now, what to A/B test and what needs more research.
Our view: most teams skip straight to group 3 (page copy and design) because it is the fun part. We would start with groups 1, 4 and 7 instead: broken tracking distorts every later decision, form and follow-up fixes act on visitors who have already shown intent, and none of them need traffic for an A/B test.
1. Tracking and Data (8 checks)
| # | Check | Pass condition | Status |
|---|---|---|---|
| 1.1 | Primary conversion is tracked | Demo request or contact submit fires one event per real submission, verified with a test submit | |
| 1.2 | No double counting | Thank-you page reloads and back-button visits do not create extra conversions | |
| 1.3 | Micro-conversions tracked | Clicks to pricing, demo page views and CTA clicks are events you can report on | |
| 1.4 | UTM parameters survive | Campaign source reaches the CRM lead record, not only the analytics tool | |
| 1.5 | Internal traffic excluded | Office IPs, staff and agency traffic are filtered out of reports | |
| 1.6 | Consent setup tested | You know what share of visitors decline cookies and how that affects reported conversion rates | |
| 1.7 | Company-level view exists | You can see which companies visit key pages, not only anonymous sessions (see visitor identification) | |
| 1.8 | Pipeline outcome linked | Leads can be traced to opportunities and closed-won deals in the CRM |
2. Funnel and Segments (6 checks)
| # | Check | Pass condition | Status |
|---|---|---|---|
| 2.1 | Funnel is written down | Each step from first visit to booked meeting has a named metric and an owner | |
| 2.2 | Biggest drop identified | You can name the step with the largest relative drop-off for the last 90 days | |
| 2.3 | Results split by device | Mobile and desktop conversion rates are reported separately | |
| 2.4 | Results split by source | Paid, organic, direct and referral traffic are compared on the same conversion | |
| 2.5 | ICP vs non-ICP split | Conversion rate for target-fit companies is visible separately from everyone else | |
| 2.6 | New vs returning split | Returning visitors are tracked as a separate segment with its own conversion rate |
3. Key Pages: Message and Layout (10 checks)
| # | Check | Pass condition | Status |
|---|---|---|---|
| 3.1 | Five-second test | A stranger can say what you sell and who it is for after five seconds on the page | |
| 3.2 | Headline names an outcome | The H1 states a result for a specific buyer, not a category label | |
| 3.3 | One primary CTA | Each page has one main action; secondary actions are visually quieter | |
| 3.4 | CTA says what happens next | Button copy describes the outcome ("See pricing for 50 seats") rather than "Submit" | |
| 3.5 | Message match with ads | Landing page headline repeats the promise of the ad or email that sent the visitor | |
| 3.6 | Proof near the CTA | A relevant logo, quote or result sits within one scroll of the main CTA | |
| 3.7 | Proof matches the visitor | Case studies shown are from an industry or company size the visitor will recognize | |
| 3.8 | Pricing is findable | Pricing or a clear "how pricing works" page is reachable from the main navigation | |
| 3.9 | Objections answered | Security, integration and implementation questions are answered on the page or one click away | |
| 3.10 | No dead ends | Every blog post and resource page has a relevant next step, not only a footer |
4. Forms (8 checks)
| # | Check | Pass condition | Status |
|---|---|---|---|
| 4.1 | Field count justified | Every field is used by sales or routing; anything else is removed or enriched later | |
| 4.2 | Work email only where needed | Free email domains are blocked only on forms where sales would reject them anyway | |
| 4.3 | Errors are clear | Validation messages appear inline and say how to fix the problem | |
| 4.4 | Form tested on mobile | A full submission on a phone works, with the right keyboard for each field | |
| 4.5 | Expectation set | The page says what happens after submit (call length, who calls, response time) | |
| 4.6 | Privacy note present | A short line explains how the data is used, with a link to the privacy policy | |
| 4.7 | Field drop-off measured | You know which field causes the most abandonment | |
| 4.8 | Scheduling offered | Qualified submitters can pick a meeting time on the thank-you page |
For more on form design, see our post on the B2B lead capture form.
5. Speed and Technical (6 checks)
| # | Check | Pass condition | Status |
|---|---|---|---|
| 5.1 | Core Web Vitals pass | LCP, INP and CLS are in the "good" range in PageSpeed Insights field data, or lab data if field data is missing | |
| 5.2 | Hero renders first | Headline and CTA appear before videos, chat widgets and carousels finish loading | |
| 5.3 | Third-party scripts audited | Every tag in the tag manager has an owner and a reason to exist | |
| 5.4 | No layout shift on swaps | Personalized or test variants reserve space, so content does not jump | |
| 5.5 | Broken links fixed | No 404s on pages linked from navigation, ads or email campaigns | |
| 5.6 | Cross-browser check | Key pages and forms work in current Chrome, Safari, Edge and Firefox |
6. Trust and Relevance (5 checks)
| # | Check | Pass condition | Status |
|---|---|---|---|
| 6.1 | Company is identifiable | Company name, location and contact route are visible without searching | |
| 6.2 | Claims are sourced | Every statistic on the page links to or names its source | |
| 6.3 | Security info exists | Certifications, data handling and hosting region are documented for enterprise buyers | |
| 6.4 | Content fits the segment | Key segments (industry, company size) see examples relevant to them; see B2B landing page optimization | |
| 6.5 | Content is current | No outdated dates, discontinued features or old screenshots on key pages |
7. Follow-Up and Handoff (5 checks)
| # | Check | Pass condition | Status |
|---|---|---|---|
| 7.1 | Response time measured | Median time from form submit to first human contact is known | |
| 7.2 | Routing rules documented | Each lead type has a named owner and a fallback when that owner is away | |
| 7.3 | Context passed to sales | Reps see the pages viewed and source before the first call | |
| 7.4 | Confirmation email sent | The submitter gets a confirmation with the next step within minutes | |
| 7.5 | Lost leads reviewed | Leads rejected by sales are reviewed monthly to fix targeting or form questions |
Scoring Sheet
Count the Fails per group, then score each failed check on three questions. The combined score decides what you do with it.
| Question | 1 point | 2 points | 3 points |
|---|---|---|---|
| Impact: how much pipeline passes through this? | Minor page or segment | One key page | All demo requests or all paid traffic |
| Confidence: how sure are you it hurts conversion? | Hunch | Supported by one data source | Supported by data and recordings or customer quotes |
| Ease: how quickly can it change? | Needs engineering sprint | A few days of work | Can change today |
- 8 to 9 points: fix now, no test needed (broken tracking, a form error, a dead link).
- 5 to 7 points: run an A/B test if the page has enough traffic. Check with the A/B test calculator first.
- 3 to 4 points: research more before acting (surveys, sales call notes, recordings).
Illustrative Example
Illustrative numbers, not a real client. A B2B software company gets around 9,000 visits a month and 40 demo requests. The audit finds three Fails worth acting on:
- 1.2 double counting (score 9): the thank-you page fires a conversion on every reload, so the real demo count is closer to 32. Fixed the same day, which also resets the baseline for every later test.
- 4.1 field count (score 7): the demo form asks for nine fields including budget and timeline. Sales confirms they only use five. This becomes the first A/B test: five fields vs nine.
- 7.1 response time (score 8): median first contact is 26 hours. The team adds routing rules so demo requests reach a rep within the hour. Our post on speed to lead covers why that window matters.
Notice that none of the three are headline or design changes. When tracking and follow-up have never been checked, that is where we would expect a first audit's cheapest fixes to be.
Frequently Asked Questions
What should a conversion rate optimization checklist include?
At minimum: tracking accuracy, funnel and segment data, page messaging and layout, forms, speed, trust signals and lead follow-up. Checklists that cover only page design miss the tracking and follow-up problems that often cost the most pipeline in B2B.
How often should I run a CRO audit with this checklist?
Run the full checklist once or twice a year and after any site redesign, tracking migration or form change. Groups 1 and 7 (tracking and follow-up) are worth a quick monthly check because they break silently.
Is this the same as a website personalization audit?
No. This checklist asks whether your current site converts. The website personalization audit checklist asks whether you are ready to show different content to different segments. Run the CRO checklist first; personalizing a page with broken tracking only makes the measurement problem harder.
Can I use this checklist for a non-B2B site?
Most checks apply anywhere. Groups 2.5 (ICP split), 6.3 (security documentation) and 7 (sales handoff) are specific to B2B sites with a sales team behind the form.