CRO tools: building a stack that earns its cost

Most CRO tool lists are a pile of logos with a paragraph each, and half of them still recommend a product Google switched off years ago. This one is organized by the job you are trying to do, and it only quotes a price when we could open the vendor's own pricing page and read it.

That last part matters more than it sounds. A lot of software pricing quoted around the web is stale, guessed, or invented outright. Where a vendor does not publish a number, we say so instead of making one up.

The four jobs a CRO stack has to do

Strip away the category names and every conversion tool is doing one of four things.

One: see what people do. Heatmaps, scroll depth, click maps, session recordings. This is the observational layer, and it is where you notice that nobody scrolls past the second section, or that people keep rage clicking a heading that looks like a button.

Two: hear why they do it. On-page surveys, exit questions, post-conversion questions, actual conversations with customers. Behavior tells you what happened. Only people tell you what they were thinking.

Three: change something. A/B testing platforms, multivariate tools, page builders with testing baked in. This is the layer everyone shops for first.

Four: prove it worked. Analytics, conversion tracking, and enough statistical discipline to tell a real lift from a good week.

Here is the pattern we see constantly: someone buys a tool for job three, skips one and two entirely, and then runs tests on ideas pulled out of a meeting. That is not optimization, that is gambling with a dashboard. Jobs one and two are what turn a test into a hypothesis, and they are the cheapest part of the whole stack.

Behavior analytics: heatmaps and session recordings

Microsoft Clarity is free. Not free trial, not free up to a session cap. The vendor states "Free forever" with "No limits on traffic." You get session recordings, heatmaps, AI summaries and mobile app analytics, and it is GDPR and CCPA compliant.

We are going to be blunt about this: a free, uncapped, genuinely good behavior analytics tool exists, so "we do not have budget for CRO tools" is no longer a real sentence. Install Clarity today. Watch twenty recordings this week. You will find something worth fixing before you finish the second cup of coffee.

Pricing verified August 2026. Software pricing drifts. Every figure on this page was read off the vendor's own site on August 13, 2026. Confirm current pricing on the vendor site before you buy.

Hotjar is now part of Contentsquare. This is worth knowing before you go shopping, because hotjar.com/pricing currently redirects to contentsquare.com/pricing. Hotjar has been the default recommendation in this category for a decade, so a lot of advice still treats it as a standalone product. We could not verify Hotjar's current pricing, so we are not going to quote any.

Mouseflow covers the same territory with recordings, heatmaps and friction scoring designed to surface problem sessions rather than make you watch everything. Lucky Orange bundles recordings and heatmaps with live chat and on-page polls, which suits smaller teams who would rather run one subscription than three. Neither price is quoted here because neither was verified.

Whatever you pick, the discipline matters more than the logo. Watch recordings in batches with a specific question in mind, write down what you saw, and turn the recurring problems into a list. A tool nobody opens is an expense, not a stack.

A/B testing tools, and the Google Optimize hole

Let us clear this up first, because it sends people in circles. Google Optimize was sunset and no longer exists. It shut down in 2023. Plenty of CRO tool roundups still list it, usually near the top, usually described as the obvious free starting point. It is not there. Google did not ship a direct replacement either. At sunset it published a list of A/B testing partners and pointed people toward third party platforms.

So the free-testing-tool-from-Google slot is empty, and everything below is what people actually moved to.

VWO has a free tier the vendor labels "Explore for Free," with paid tiers called Growth, Pro and Enterprise. Growth covers A/B testing with both visual and code editors, plus heatmaps, session recordings and form analytics, which makes it one of the few products spanning jobs one and three. Pro adds multivariate testing and scrollmaps. Enterprise adds a dedicated account manager. VWO does not publish starting prices on its pricing table, so we are not going to invent one.

Optimizely sits at the enterprise end, built for organizations running many concurrent experiments across web and product with real governance around them. Convert targets teams that want a dedicated testing platform with a strong privacy posture and a lighter footprint. AB Tasty combines experimentation with personalization and feature rollouts. None of the three publish pricing we could verify, so none is quoted.

GrowthBook and PostHog are the open source options. Both put feature flags and experimentation in the same place, which is a natural fit if your changes ship through code rather than a visual editor. Self-hosting trades a subscription for engineering time, so it is a good deal when you have developers and a poor one when you do not.

The constraint nobody wants to hear

A/B testing needs meaningful traffic and conversion volume to produce a result you can trust. Significance is driven by conversions per variant, not pageviews, so a page generating five leads a week is not going to resolve a test this quarter. Run one anyway and you will get a number that looks like an answer and behaves like a coin flip.

If that describes your site, the honest advice is to skip formal testing for now. Qualitative research and obvious fixes will beat it comfortably. Watch recordings, ask customers why they hesitated, fix the slow page, add the missing proof, cut the field nobody wants to fill in. Ship the change, watch the trend, move on. You can graduate to structured experimentation when the volume shows up.

Landing page builders with testing built in

For most small and mid-sized advertisers, this is the practical middle path. You get the page builder and the testing in one subscription, and you skip the part where a developer has to implement every variant.

Unbounce publishes its pricing plainly. Monthly and annual, with annual saving 25 percent: Starter $29 monthly or $22 annual, Build $99 or $74, Experiment $149 or $112, Optimize $249 or $187. Concierge and Agency are custom quotes. Traffic limits run 500, 20k, 30k, 50k and 100k or more by plan, and all plans include unlimited conversions.

Leadpages splits into two plan families. HTML Pub plans are Starter $10, Pro $29 and Business $49. Leadpages plans are Grow $99, Optimize $199 and Scale $399. Every plan carries unlimited traffic with no caps. Manual A/B testing starts at the Grow tier. Annual billing saves 20 percent, and there is a 7-day free trial.

The structural difference is the one to pay attention to: Unbounce meters traffic by plan, Leadpages does not cap it. If you are running paid media, a good month can quietly move you up a tier on one and change nothing on the other. That is a real cost difference, not a feature checkbox, and it is why we wrote a whole comparison. Full breakdown here: Unbounce vs Leadpages.

Voice of customer: surveys and feedback

This is the cheapest input in CRO and the one almost everybody skips, presumably because installing a tool feels more like progress than asking a question.

Three methods do most of the work.

On-page micro-surveys. One question, fired on a specific page after a delay or a scroll trigger. "What almost stopped you from filling this out?" is worth more than a month of heatmaps, because it names the objection in the visitor's own words. Keep it to one question. A second one halves your responses.

Exit surveys. Triggered on exit intent or after a period of inactivity, aimed at people who are leaving without converting. Expect low response rates and read every answer anyway. Ten honest sentences will reshape a page.

Post-conversion questions. The most underrated of the three. On the thank you page, ask what nearly stopped them from buying, or what they were worried about. These people just said yes, so they are willing to talk, and they will hand you the exact objection your page barely survived.

The vendors here matter less than the habit. Several behavior tools include polls, most form and survey products can do it, and a rough implementation that actually ships beats a polished plan that does not. What you are building is a running list of real objections in real language, which then becomes your test backlog and, frankly, your best source of headline copy.

Analytics and measurement

GA4 is the baseline, it is free, and it is where you will look to see whether anything you did changed the trend. Nothing else in this article means much if this layer is wrong.

And it is wrong more often than people expect. Forms that submit without firing an event. Thank you pages counted as conversions while the form errors out silently. Duplicate tags double counting. Phone clicks tracked on desktop and not mobile. A conversion definition that counts a pageview rather than an action. Any one of these turns your test results into fiction.

So before you optimize anything, verify. Submit your own form and watch the event arrive. Check the conversion count against the leads that actually landed in your inbox or CRM for the same period. If the two numbers disagree, fix that first, because every decision downstream inherits the error.

Two more measurement habits worth having. Decide your primary metric before the test starts, so you cannot go hunting for a segment where the losing variant won. And give a test a fixed duration covering full weeks, because Tuesday traffic and Saturday traffic do not behave the same way and stopping early on a good day is the most common way people convince themselves of a lift that is not there.

What to actually buy, at three budget levels

$0 per month

Microsoft Clarity for heatmaps and recordings, GA4 for measurement, a free survey tool for one on-page question, and manual before-and-after comparison instead of formal testing. That is a complete stack covering all four jobs, at no cost, and it is genuinely enough to find and fix the biggest problems on most sites. Run this for a month before you spend anything. The list of obvious fixes it produces will be longer than your budget anyway.

Under $150 per month

Keep everything above and add a page builder with testing, so you can ship and compare variants without a developer in the loop. This is exactly where Unbounce Build at $99 monthly or $74 on annual billing and Leadpages Grow at $99 monthly sit. Unbounce Experiment is $149 monthly or $112 annual if you want the higher tier and can stay inside the traffic allowance. Leadpages Grow is the tier where its manual A/B testing begins, with no traffic cap to watch.

At this level the constraint is not software, it is you. A builder subscription only pays for itself if pages actually change. If nothing on the account will be edited for six months, stay at zero.

Scaling

A dedicated experimentation platform, a behavior tool with the retention and segmentation to investigate what a test surfaced, structured voice of customer research on a schedule, and someone whose job includes running the program.

The honest note: this tier only pays off above real traffic volume. A dedicated platform earns its cost by running many tests that each resolve quickly, and that requires the conversion volume to resolve them. Buying enterprise experimentation for a site doing modest numbers gets you a very expensive way to fail to reach significance. If you are unsure which side of that line you sit on, count conversions per week per page and be pessimistic about it.

FAQs

The questions people ask most often when they sit down to pick a conversion stack.

Where to go from here

Start free today: install Microsoft Clarity, confirm your GA4 conversions are accurate, and watch twenty recordings.

Pick a builder: the Unbounce vs Leadpages comparison has the verified pricing and the traffic cap difference laid out side by side.

Get the pages right first: tools do not fix a weak layout. Landing page design covers the structure, and PPC landing pages covers the campaign side.

If you would rather have someone else run the program, you can request a CRO review. No obligation, and we will tell you if testing is the wrong move for your traffic level.