# What I Look for in the First 30 Minutes of a B2B Website Audit

> My 30-minute B2B website audit: the five timed checks, the free tools I keep open, and what the findings tell me about a site's deeper problems.

Canonical URL: https://adityamallah.com/blog/b2b-website-audit-first-30-minutes
Author: Aditya Mallah
Published: 2026-10-05T00:00:00.000Z
Updated: 2026-10-05T00:00:00.000Z
Topics: B2B marketing, website audits, conversion optimization, SEO, analytics, SaaS

A full B2B website audit takes days. The first 30 minutes decide what that audit eventually finds, and most people spend them on the wrong things.

In half an hour you can't crawl a whole site or rewrite your messaging. What you can do is check the four things that kill most B2B sites: an unclear promise, missing proof, a buried or weak call to action, and a data setup that hides the truth. Everything else (redirect chains, schema gaps, design taste) can wait an hour.

I run this check with free tools and a timer.

## The Four Things That Kill Most B2B Sites

The first 30 minutes are for evidence, not opinions. And the evidence has narrowed: the buying committee researches without you. [Gartner's survey of 646 B2B buyers](https://www.gartner.com/en/newsroom/press-releases/2026-03-09-gartner-sales-survey-finds-67-percent-of-b2b-buyers-prefer-a-rep-free-experience) (published March 2026) found 67% prefer a rep-free experience, and 45% had used AI during a recent purchase.

The site, not a sales rep, carries the first conversation, so before any tool opens, I hold four suspicions. Every one of them has ended more B2B pipelines than a slow server ever has.

- **An unclear promise.** A stranger can't say what the company does.
- **Missing proof.** Claims with no names, no numbers, no logos.
- **A buried call to action.** The demo button exists somewhere, but a buyer can't find it in three clicks.
- **A data setup that hides the truth.** Conversions aren't tracked, so nobody can say what's working.

One rule for the half hour: look as a stranger, in a private window, with free tools only.

## Minutes 0–5: The Stranger Test

**The stranger test asks one question: could someone who has never heard of you say what you do, who it's for, and why they should care, from the top of the homepage alone?**

The classic Carleton University study by Lindgaard and colleagues (2006) found people form a durable visual judgment of a page in about 50 milliseconds, before reading a word. Usability practitioners formalized it as the 5-second test: show someone a page briefly, then ask what the company does. If the answer is mush, problem found.

What I actually click, in order:

1. Open the homepage in a private window, on my phone first.
2. Read only the headline and subhead. Out loud: what do they sell, for whom? No answer (or one that fits any company in the category) is a fail.
3. Jargon check. "Empowering intelligent workflows for modern enterprises" is a fail. "Pipeline forecasting for manufacturing sales teams" is a pass.
4. Find the primary call to action above the fold. Is there one obvious button? Does it name the action ("Book a demo") or hide it ("Learn more")?
5. Scan the navigation labels. Can I guess where pricing, product, and proof live without hunting?

What bad looks like: three competing buttons, a headline about the company instead of the customer, and a "Learn more" pointing to the page I'm already on. I note the exact wording I couldn't parse, that quote becomes the brief.

## Minutes 5–10: Walk One Visitor Path End to End

Messaging fails in the wild, not on the homepage. So I follow one path a real buyer would take, start to finish.

For context: [Unbounce's Conversion Benchmark Report](https://unbounce.com/conversion-benchmark-report/) (2024 edition, 41,000+ landing pages) put the median landing page conversion rate at 6.6% across industries. B2B lead-gen pages usually sit below that median, and path friction is one fixable reason.

1. Search Google the way a buyer would, a category phrase, not the brand name. If the company runs ads, click the ad. That's their best version of the first impression.
2. Land, and compare. Does the page continue the promise from the ad or search result, or dump me on a generic homepage?
3. Click the page's main CTA. Count the clicks to the actual conversion point. Two is fine; five is a leak.
4. Count the form fields, then fill the form in and submit it on desktop and mobile. The test isn't done until the entry lands in the CRM or inbox, as [OuterBox's practical checklist](https://www.outerboxdesign.com/articles/digital-marketing/how-to-do-a-website-audit) (June 2026) rightly insists.
5. Look at the confirmation state. A real thank-you page with a next step, or a dead end?

What bad looks like: an ad promising one thing and a homepage selling another, an eleven-field form asking for my budget, and a "Thanks!" leading nowhere. Every field is a question the buyer didn't come to answer.

## Minutes 10–15: Audit the Proof, Not the Claims

Buyers shortlist during anonymous research, and most later validate what AI tools told them with a human (69%, per [Gartner's May 2026 survey](https://www.gartner.com/en/newsroom/press-releases/2026-05-20-gartner-survey-finds-sixty-nine-percent-of-b-two-b-buyers-turn-to-sales-reps-to-validate-ai-generated-insights)). Your proof section does the selling a first call used to do. Weak proof is killer number two, and it takes five minutes to grade.

1. Find the customer logo bar. Recognizable, current, legible, or a gray blur of companies I can't identify?
2. Read one testimonial. Does it name a person, a role, a company, and something specific? "Great product, love the team" is filler.
3. Click into one case study. Is there a number, a timeframe, and a customer name? Or 500 words of quotation marks around nothing?
4. Check any big claims. "Trusted by 10,000+ teams" should trace to something, if I can't tell where the number came from, I write it down.
5. Look for risk reducers near the CTA: security and compliance mentions (SOC 2, GDPR), pricing transparency, a guarantee.

What bad looks like: stock-photo testimonials, case studies with no outcomes, and a logo bar that hasn't been touched since the rebrand. Vague proof doesn't just fail to persuade, it raises the exact doubts it was meant to settle.

## Minutes 15–22: Technical Quick Hits

I'm not doing a technical SEO audit here. I'm checking three things: does the site load fast enough, does it work on a phone, and is it actually in Google? Each check has a single tool and a single verdict.

**Core Web Vitals are Google's field metrics for real-user experience: Largest Contentful Paint (LCP) for loading, Interaction to Next Paint (INP) for responsiveness, Cumulative Layout Shift (CLS) for stability.** [Google's documentation](https://developers.google.com/search/docs/appearance/core-web-vitals) (updated December 2025) sets the bar at LCP 2.5 seconds or less, INP 200 milliseconds or less, CLS 0.1 or less, judged at the 75th percentile of real page loads. INP replaced FID as a Core Web Vital in [March 2024](https://web.dev/blog/inp-cwv-march-12), any checklist still citing FID is out of date.

1. Run the homepage through [PageSpeed Insights](https://pagespeed.web.dev/). Read the field data first, real Chrome users, past 28 days. The Lighthouse lab score runs 0–100: 90+ good, 50–89 needs improvement, under 50 poor. Low-traffic pages may show whole-site (origin) data instead.
2. Run the money page too, the product or pricing page, not just the homepage. Homepages get redesign love; money pages get plugins.
3. Open the site on your phone, the device carrying about 61% of worldwide web traffic as of October 2026 ([StatCounter](https://gs.statcounter.com/platform-market-share/desktop-mobile/worldwide)). Does the hero render, does the menu open, does content jump around while you read (that's CLS you can feel)?
4. Type site:yourdomain.com into Google. Is the site indexed at all? Does a staging subdomain show up? Treat the count as an estimate (Search Console is the source of truth) but a missing site or an indexed staging copy is a five-alarm finding.
5. If I have access, open Search Console's Core Web Vitals and Pages (indexing) reports.

What bad looks like: a mobile LCP of six seconds, no field data on the pages that matter, and a homepage title tag that just says "Home". None of this requires a crawler to discover.

## Minutes 22–30: The Data Check

The last block answers one question: could anyone running this site tell what's working? If not, every decision is a guess, the fourth killer.

1. Open GA4 → Reports → Engagement → Landing page, set to the last 90 days. Which pages start sessions? Sort by key events, not sessions, sessions flatter the wrong pages.
2. Add "Session source / medium" as a secondary dimension, the way [Google's own documentation](https://support.google.com/analytics/answer/12931766) suggests, to see where each landing page's visitors come from.
3. Check Admin → Events for the form-submit or demo-request event marked as a key event. Not marked (or marked but showing zero) means the site's most important number is missing.
4. Then submit a test conversion and watch it fire in DebugView's real-time stream. A form that "works" visually but never reaches analytics is quietly common.
5. Open [Search Console → Performance](https://support.google.com/webmasters/answer/7576553) and compare the Queries tab with the Pages tab. Which queries earn clicks? Which pages get impressions but no clicks, a title or intent mismatch? One quirk: query data aggregates by property, page data by URL, so chart and table totals won't always match. That's normal, not broken.

What bad looks like: 90% "direct" traffic, "(not set)" landing pages, and a Search Console property that was verified in 2022 and never opened since. Inflated direct traffic usually means broken UTMs, redirect chains stripping parameters, or consent settings eating attribution.

## The Free Tools I Keep Open

Everything above runs on free tools. Here's what each tells you during a quick audit.

| Tool | What it tells you in this audit | Cost |
| --- | --- | --- |
| PageSpeed Insights | Real-user Core Web Vitals for a single URL (past 28 days of Chrome data) plus a Lighthouse lab score | Free |
| Google Search Console | Queries that earn clicks, indexing status, the Core Web Vitals report | Free |
| GA4 | Landing pages, traffic sources, and whether key events actually fire | Free |
| Google site: search | A ten-second gut check on indexability | Free |
| Screaming Frog SEO Spider | Free tier crawls up to 500 URLs for titles, meta, redirects, and broken links, the natural step two | Free; [paid licence](https://www.screamingfrog.co.uk/seo-spider/) removes the cap (€245/year, June 2026) |
| Your phone | Mobile rendering, tap targets, form usability, the audit tool you already own | Free |

These free tools are enough for the first pass.

## What I Deliberately Skip in the First 30 Minutes

I skip three things on purpose, not because they don't matter, but because a 30-minute version of each produces a 30-minute conclusion, and those are usually wrong.

- **Full technical crawls.** A crawl of a few thousand URLs returns hundreds of "issues" and zero priorities. It's the right step two, with a scope and a budget attached.
- **Brand and design critique.** Taste debates can't be settled by a glance, and the stranger test already catches the messaging failures that design conversations usually orbit.
- **Content inventory and AI-search visibility checks.** Both are worth a dedicated session each, the AI-search one especially, since it needs its own method before you spend anything on GEO.

Skipping is the discipline. The clock forces you to trust the four killers instead of wandering wherever the scroll takes you.

## What 30 Minutes Predicts About the Deeper Problems

A 30-minute audit rarely tells you the fix. It reliably tells you where the real problem lives, and how big the eventual project will be.

- A vague hero headline means positioning drift, the homepage is the symptom; the fix starts in product and sales conversations.
- Missing, generic proof means no case-study engine, expect the same gap in proposals and decks.
- A buried call to action means a committee-run site where nobody owns conversion, expect slow approvals.
- Broken analytics means decisions run on vibes, the first battle is convincing people to measure at all.
- Slow money pages mean CMS and plugin debt with no performance budget, so fixes cost more than they look.

That's why I do the 30 minutes at all. It doesn't produce a report; it produces a bet on where the real audit should dig first, and it's right often enough that I won't start a serious engagement without it.

## FAQ: Quick B2B Website Audit Questions

Short answers, based on how I actually run these half hours.

**What should I check first in a website audit?**
The homepage promise, as a stranger, then one conversion path end to end. If a first-time visitor can't tell what you do in five seconds, everything else you fix is decoration.

**What are the most common B2B website problems?**
Jargon headlines, testimonials without names or numbers, demo CTAs buried or vaguely labeled, forms that dump leads into an unmonitored inbox, and analytics where the main conversion was never marked as a key event.

**How long should a quick audit take?**
Thirty minutes for triage if you use the timed blocks above. A full audit runs days to a couple of weeks depending on site size. Cadence: a quick pass monthly or after major changes, a deeper audit once or twice a year on key pages.

**Which free tools do I need?**
PageSpeed Insights, Google Search Console, and GA4, plus the site: search operator, cover most of a first pass. Add Screaming Frog's free 500-URL crawl when you're ready for the deeper technical pass.

**What can a 30-minute audit catch that a full audit misses?**
What a first-time visitor actually experiences. A crawler can't feel a vague headline, notice that a form asks for a fax number, or sense that the demo button reads as a trap. Full audits quantify; the 30-minute pass humanizes.

## Sources

- Google Search Central (Understanding Core Web Vitals and Google search results) https://developers.google.com/search/docs/appearance/core-web-vitals (updated December 2025)
- web.dev (Core Web Vitals) https://web.dev/articles/vitals/ (updated October 2024)
- web.dev (INP is a Core Web Vital) https://web.dev/blog/inp-cwv-march-12 (March 2024)
- Google Analytics Help ([GA4] The Landing page report) https://support.google.com/analytics/answer/12931766
- Google Search Console Help (Performance report) https://support.google.com/webmasters/answer/7576553
- Google (PageSpeed Insights v5 about) https://developers.google.com/speed/docs/insights/v5/about
- Gartner (Gartner Sales Survey Finds 67% of B2B Buyers Prefer a Rep-Free Experience) https://www.gartner.com/en/newsroom/press-releases/2026-03-09-gartner-sales-survey-finds-67-percent-of-b2b-buyers-prefer-a-rep-free-experience (March 2026)
- Gartner (Gartner Survey Finds 69% of B2B Buyers Turn to Sales Reps to Validate AI-Generated Insights) https://www.gartner.com/en/newsroom/press-releases/2026-05-20-gartner-survey-finds-sixty-nine-percent-of-b-two-b-buyers-turn-to-sales-reps-to-validate-ai-generated-insights (May 2026)
- StatCounter Global Stats (Desktop vs Mobile Market Share Worldwide) https://gs.statcounter.com/platform-market-share/desktop-mobile/worldwide (October 2026)
- Screaming Frog (SEO Spider) https://www.screamingfrog.co.uk/seo-spider/
- Unbounce (Conversion Benchmark Report) https://unbounce.com/conversion-benchmark-report/ (2024 edition)
- OuterBox (How To Audit A Website: Practical Checklist) https://www.outerboxdesign.com/articles/digital-marketing/how-to-do-a-website-audit (June 2026)

## Related reading

- [Your SaaS Website Gets Traffic but No Demos](/blog/saas-website-traffic-but-no-demos)
- [The First 90 Days of SEO for a B2B SaaS Startup](/blog/first-90-days-of-seo-for-b2b-saas-startup)
- [The AI Search Visibility Audit I'd Run Before Spending Money on GEO](/blog/ai-search-visibility-audit-before-geo)

Source: https://adityamallah.com/blog/b2b-website-audit-first-30-minutes
