CRO agency: what a conversion rate optimisation agency does

A CRO agency is a specialist that helps online shops turn the traffic they already have into more revenue. Not with more advertising. Not with more traffic. With systematic conversion rate optimisation of what happens to the visitors who are already on the site.

Since 2012 our CRO work has generated €245.3M^¹ in additional revenue for clients. The gap between a good CRO agency and a bad one is huge. This article shows you how to spot a serious agency, and which traps to avoid.

What is a CRO agency?

Definition and core tasks

A CRO agency works systematically on raising the conversion rate. The conversion rate is the share of visitors who complete a defined goal: a purchase, a sign-up, a download, an enquiry.

The CRO agency combines methods from data analysis, user research, UX design and A/B testing. It works empirically: every change is validated with data, not with assumptions. Its core tasks are analysing web analytics data, running user tests, writing hypotheses and implementing experiments.

A CRO agency acts as an outside adviser that sets up a structured process. It brings method, tools and experience from many industries. Many companies hire one because they lack the know-how in-house or their people are tied up elsewhere.

How it differs from a classic marketing agency

Classic marketing agencies focus on getting traffic: SEO, paid search, social media and content marketing. A CRO agency optimises what happens to the traffic you already have.

A marketing agency asks: “How do we get more visitors?” A CRO agency asks: “How do we turn more visitors into customers?” The two disciplines complement each other, but their goals differ.

CRO agencies often work with specialist tools such as Optimizely, VWO, Kameleoon or Hotjar. They have experts in statistics, experiment design and front-end development. The focus is on quantitative results. Every measure is judged by its numbers.

When does working with one pay off?

Working with a CRO agency pays off when a company already gets significant traffic but its conversion rate sits below the industry average. A typical sign is a high drop-off rate in checkout or on landing pages.

Companies with complex products, long decision cycles or high average order values benefit most. In these cases even a small rise in conversion rate can produce large absolute revenue gains.

A CRO agency is no substitute for a working product or a clear value proposition, though. The basics have to be right before you optimise. The Academy track eCommerce Fundamentals ↗ gives you a structured start.

The biggest mistakes when choosing a CRO agency

Looking at price instead of process

Many companies pick a CRO agency mainly by hourly rate or monthly fee. That is a mistake. The quality of the work depends on the process: How are hypotheses generated? How is testing done? How are results interpreted?

A cheap agency without a structured approach can cost more than an expensive one with a proven method. The ROI of a CRO agency comes from the quality of its insights and its implementation rate, not from its hourly rate.

No transparency about method

Another mistake is relying on agencies that do not explain their method. Serious CRO agencies document every step: data collection, hypothesis building, test design, analysis of results.

If an agency only talks about “best practices” and never shows its own data, be careful. CRO is not a checklist discipline. Every shop, every audience and every context is different.

Unrealistic expectations of quick results

Some companies expect their conversion rate to double within a few weeks. That will not happen. CRO is an iterative process. The first weeks go into collecting and analysing data. The first tests run for several weeks until they reach statistical significance.

Realistic gains are in the single-digit percentage range per quarter. Over a year they add up to significant growth. If you end the engagement after one month without results, you have wasted the money you spent.

No internal support

A CRO agency cannot work in isolation. It needs access to analytics data, developer time to build tests, and decision-makers who act on the results. If internal teams ignore the agency’s recommendations or delay implementation, the project fails.

So choosing a CRO agency should always go hand in hand with settling internal responsibilities. Who owns the project in the company? Who prioritises the test backlog? Who frees up resources?

What works: services with measurable impact

Data-based hypotheses

Successful CRO starts with a qualified hypothesis. A good CRO agency does not build hypotheses on gut feeling but on data. The sources are quantitative analytics data (bounce rates, funnel analysis, heatmaps) and qualitative user research (usability tests, surveys, customer interviews).

A typical hypothesis reads: “If we cut the form from three steps to two, the conversion rate will rise by at least 5 percent, because we remove friction from checkout.” Hypotheses like this are testable, measurable and easy to rank.

Good agencies use frameworks such as ICE (Impact, Confidence, Ease) or PIE (Potential, Importance, Ease) to prioritise. Not every idea gets tested. Resources go to the tests with the highest expected ROI.

A/B testing and multivariate tests

A/B testing is the core tool of CRO. Two or more versions of a page run in parallel. Visitors are assigned at random. The result shows which version achieves the better conversion rate.

Serious agencies pay attention to statistical significance, sufficient test duration and representative samples. They do not stop a test early just because one version is ahead for now.

Multivariate tests (MVT) test several elements at once. They need far more traffic than A/B tests. A CRO agency decides case by case which test type fits the traffic volume and the question.

Checkout and form optimisation

Checkout is the most critical point in the ecommerce funnel. Every extra step, every unnecessary field and every unclear detail causes drop-offs. CRO agencies analyse checkout funnels and find exactly these friction points.

High-impact measures include fewer form fields, guest checkout, transparent shipping costs and trust elements such as security seals. Removing an optional phone field from checkout can lift the conversion rate by 2-5 percent. Adding a progress bar makes the process feel less complex.

Landing page optimisation

Landing pages are the entry point for many users. A CRO agency optimises these pages for one clear goal. Elements with proven impact are concise headlines, clear value propositions, prominent call-to-action buttons and fewer distractions.

This is not a one-off job but an iterative one. A landing page is tested and improved continuously. What works today can stop working tomorrow because user behaviour or competitors change.

Mobile optimisation

In many industries mobile accounts for more than 60 percent of ecommerce traffic. Even so, mobile conversion rates are often neglected. CRO agencies analyse mobile behaviour separately. Touch optimisation, load times, simpler navigation and mobile payment options are the main levers.

What does not work: hyped methods that eat resources

“Best practices” without validation

The internet is full of CRO best-practice lists: “Red buttons convert better.” “Shorter forms are always better.” “Social proof builds trust.” Without context these claims are worthless. What works in one shop can backfire in another.

A CRO agency that rolls out best practices without testing them is not working from evidence. It swaps one assumption for another. In most cases the colour of a call-to-action button has no significant effect on the conversion rate. Contrast with its surroundings and a clear label matter more.

Over-optimisation and analysis paralysis

Some companies and agencies get stuck in an endless loop of small tests. A new test starts every week, while earlier results never get implemented. The organisation is busy but not productive.

Analysis paralysis is just as bad. Data is collected, heatmaps are made and user interviews are held for months, and not a single test goes live. CRO needs a balance of analysis and action. The 80/20 rule applies: 20 percent of the analysis delivers 80 percent of the insights. After that, test.

Isolated measures without strategic context

CRO is not a standalone tactic. If you optimise checkout but neglect product pages, you only get half the gain. If you test landing pages but leave the ad copy unchanged, you create inconsistencies.

Good CRO agencies look at the whole funnel. They understand how traffic source, landing page, product page and checkout work together. Isolated tweaks to single pages produce local wins that are cancelled out elsewhere.

Technical quick fixes without the user’s view

Load time, Core Web Vitals and technical SEO matter. But they are not CRO measures in the strict sense. A page that loads in 0.5 seconds but does not communicate a clear goal will not convert better than a slower page with convincing UX.

CRO agencies that argue only on technical grounds ignore user psychology. The best technology is useless if the offer is unclear or unconvincing.

The process that works

Phase 1: Data collection and analysis

Every CRO project starts with a thorough analysis. The agency collects quantitative data from Google Analytics, the CRM and ecommerce tracking. At the same time it gathers qualitative data: user interviews, usability tests, surveys, session recordings.

The goal is a full picture of user behaviour. Where do visitors drop off? Which questions stay unanswered? Which concerns block the conversion? This phase usually takes two to four weeks. The result is a prioritised backlog of hypotheses.

Phase 2: Hypotheses and prioritisation

Hypotheses are derived from the insights of the analysis. A hypothesis always follows the same structure: “If we implement [change], [metric] will rise/fall by [value], because [reason].”

These hypotheses are then ranked. The criteria are expected impact, statistical power (enough traffic for a valid test) and implementation effort. Ranking stops resources flowing into unimportant tests.

Phase 3: Test design and implementation

The chosen hypothesis is turned into a test. For an A/B test, the variants are designed, the target metric is defined and the sample size is calculated. The technical build happens in a testing tool or directly in the code.

Before going live there is a quality check. Variants A and B are checked for function, display and tracking. A broken test is worse than no test. It distorts results and undermines trust in the method.

Phase 4: Evaluation and implementation

Once the test has run long enough, it is evaluated. The agency checks statistical significance, practical relevance (effect size) and segment effects. Sometimes a variant wins on average but loses with mobile users.

The winning variant is implemented. The test documentation goes into a knowledge repository. Every test, successful or not, adds to what the organisation learns. The results are shared with all stakeholders.

Phase 5: Iteration and scaling

CRO does not end with one successful test. The findings feed new hypotheses. A winning checkout test leads to the question: what happens on the product page before it? The process is cyclical. With each round you understand your audience better and the conversion rate keeps rising.

Do CRO yourself or hire an agency?

Criterion	CRO in-house	CRO agency
Method	Has to be built up internally	In place, proven across many projects
Tools	Licence costs and training needed	Usually included in the service
Speed	Depends on internal capacity	Dedicated team, fast start
Objectivity	Internal blind spots possible	Outside view, unbiased
Cost	Lower direct cost, more time spent	Higher direct cost, productive from day one
Scalability	Limited by team size	Can be scaled flexibly
Building knowledge	Know-how stays in the company	Knowledge transfer possible, but not guaranteed

The decision depends on internal resources, traffic volume and strategic priorities. Companies with high traffic and limited in-house know-how usually benefit from an agency. Companies with less traffic and analytical talent in-house can build CRO themselves.

A hybrid is also possible: the agency sets up the process and the internal team takes over after a training phase. In practice this is often the most efficient route.

Key points at a glance

A CRO agency is a specialist partner for data-driven conversion optimisation. It differs from a classic marketing agency by focusing on user behaviour instead of getting traffic. Choose one by method, transparency and process quality, not by price alone.

What works: hypotheses from data, validated by A/B tests. Checkout optimisation, landing page tests and a separate look at mobile. What does not work: untested best practices, over-optimisation without implementation, and isolated measures that ignore the funnel.

The process that works has five phases: analysis, hypotheses, test design, evaluation and iteration. CRO is not a one-off project but an ongoing discipline. If you understand that and choose the right agency or internal setup, you will get measurably more revenue from the traffic you already have.

Frequently asked questions about CRO agencies

What does a CRO agency cost?

We work in sprints with concrete named changes. The price depends on the scope and is quoted on request after the Dropoff Analysis.

How long until the first results show?

The first step is the Dropoff Analysis: we look at your shop and name the concrete leaks. From there we work in sprints, each with named changes and a measured result. We report what the data shows, not a promised timeline.

Can you do CRO in-house?

Yes. You need internal resources for data analysis, test implementation and UX design. Tools such as Kameleoon, Hotjar and Analytics are partly free or low-cost. Building internal skills takes time, though.

Companies without experience in experiment design and statistics risk misreading results. A hybrid, where the agency sets up the process and the internal team takes over, is often the most efficient route.

How do you measure a CRO agency’s success?

The main indicator is the rise in conversion rate. Secondary metrics are the number of tests run, the implementation rate of winning variants and the ROI of the measures.

A good agency reports these numbers openly. It documents every test and its results. If you do not get clear figures, question the engagement. Success in CRO is either measurable or it does not exist.

What is the difference between CRO and UX design?

UX design focuses on the user experience and usability of a website. CRO focuses on measurable business results. The two overlap but have different goals.

A UX designer asks: “Is the page clear and pleasant to use?” A CRO expert asks: “Does the page lead to more conversions?” The best solution combines both views. Good CRO agencies employ UX experts or work closely with them.