Beyond A/B Testing: Practical Conversion Rate Optimization Strategies for Low-Traffic Sites
Conversion rate optimization (CRO) is a data-driven process of continuously improving interfaces and experiences to turn more website or app visits into customers. In other words, maximizing the value of your site’s traffic.
Often, CRO is mistaken as being synonymous with A/B testing. Or, businesses might think that CRO requires tightly constricting all possible user actions on a page. Even more businesses think that CRO is a one-time practice that can be done via “quick hacks” from blogs, ebooks, or podcasts.
These aren’t bad starting points for exploring CRO for your business! However, it’s limiting to think that CRO is just about testing, and risky to expect that one-time hacks will drive up your revenue.
We offer CRO services and data-driven, iterative optimization practices that drive scaling value — but if you’re new to CRO, concerned about its effectiveness for smaller brands, or curious about first steps, where do you start?
This quick guide explores CRO and tactical ways to build an ongoing optimization practice for your business, even if you receive relatively low traffic volumes. We’ll cover:
- The Basics of CRO
- The Challenge of Sample Size
- 5 Steps for Low-Traffic CRO
The Basics of CRO
Regardless of whether you’re working with an agency or doing CRO in-house, the core CRO process will include three basic steps:
- Diagnosis
- Problem identification
- Problem solving
Depending on who you consult, there could be many variations in how these three basic steps are executed. But, in essence, it all comes down to understanding what’s happening on your website based on the data you see, defining the problem you need to solve to help visitors (and prospects) complete the expected action, and then finding a solution to that problem.
Here’s a closer look at each step:
- Diagnosis
- It’s difficult (and not recommended) to optimize blindly, so this crucial first step involves making sense of the data you have within reach — analytics, heatmaps, session recordings, CRM data, etc. As much as possible, use at least two data points to help you build a better understanding of what, how, or why things are happening.
- Look for patterns and, following Twyman’s Law, be skeptical of any good, interesting, or different data you see. “Any figure that looks interesting or different is usually wrong.”
- Go through the website as if you’re a customer trying to solve a problem or find information. Compare the real journey to how the UX was designed and how actual users engage with the site during session recordings. Was there a striking difference?
- Problem Identification
- From your diagnosis, list down all the difficulties and issues you encountered. Are there issues that can be segmented together, pointing to root problems? Is it a functional problem with the site or a clarity problem with the layout, language, or design?
- When you identify a problem, also categorize whether or not it’s something that must be fixed immediately (like functional issues or incorrect information) or something that should be tested. This will help you prioritize the actions you need to take.
- Remember that some tests can’t happen concurrently without diluting the other running experiments and that immediate fixes would still take up web development resources, so plan accordingly.
- Problem Solving
- The most common way businesses solve a concern is by immediately shipping out a change or testing a potential solution first. This step is the reason people most often confuse CRO with A/B testing, since these tests are a common method for identifying best-fit solutions to identified problems.
- Regardless of how you want to approach your problem solving, be sure to tag the changes made and/or ideas tested based on their purposes. Document all key learnings from them. This way, you’ll get a sense of the direction you’re going for and you’ll always gain something actionable even if the test ultimately fails to solve that specific problem.
The Challenge of Sample Size
These three basic steps are easier to put into action on a website with a decent amount of traffic. This is because more traffic makes it easier to establish patterns, test ideas, and confidently say that a change did or did not make an impact.
What if you’re concerned
For niche businesses, including mission-driven B2B brands, average monthly traffic is likely quite lower than what a medium to large B2C brand might see. And for any smaller brand, site traffic is expensive and can be a luxury.
You might not have the budget to pull in more people to your site through paid campaigns, or you could be relying solely on organic efforts, which still costs money and resources to support.
Unfortunately, for a standard fixed horizon A/B test to be trustworthy, it needs enough traffic volume to achieve statistically significant and high confidence results. Here’s a quick illustration of why this is the case:
Example A: Pre-test analysis with an ideal test page (enough traffic to get statistically significant and high-confidence results within 6 to 8 weeks)

Minimum detectable effect (MDE) estimations tells us how much impact our tested page variation needs to have to get a significant result. It essentially represents the size of the improvement your new page variant needs to achieve for you to mathematically prove it was the cause of the success. Commonly, this should be less than 20% depending on the scale of the change you plan to run through a 6- to 8-week test.
Example B: Pre-test analysis from a low-traffic web page tells us it’s not viable to do an A/B test based on the MDE and the duration.

In this case, the MDE is too high and would risk generating false negative or false positive results.
But while it’s more challenging to execute the core steps of CRO for a low traffic website, that doesn’t mean low-traffic sites can’t do CRO. The best time to do CRO isn’t when you have the traffic, but as early as you possibly can.
Building a high-converting site can pay dividends as your traffic grows over time.
Think of it this way: Waiting until you’ve already grown your traffic to start optimizing means you’ve likely left a ton of conversions and value on the table along the way. It’s possible to adapt your approach to your current circumstances and iterate as you go.
In the CRO space, the standard rule of thumb for minimum monthly traffic is around 20k-40k sessions per month to yield high-confidence test results.
However, we’ve found 10k sessions per month to be feasible and still drive meaningful benefits when paired with a tailored approach.
The B2B High-Value Factor
When starting with CRO, it’s helpful to remember the outsize impact that improvements can make for B2B businesses.
Remember, you aren’t chasing millions of clicks.
If you have low traffic but high contract values, a single extra conversion from the five ideal buyers visiting your site this month could completely fund your marketing budget for the quarter. Every single visitor can count exponentially more.
5 Steps for Low-Traffic CRO
1. Understand your data.
You don’t need a sophisticated tool to start your CRO analysis.
If all you have is Google Analytics, a heatmap tool like MS Clarity (which is free!), and a spreadsheet with all your leads, start with those data sources. You’ll just need to understand how they relate to each other because they have different data collection methods and purposes.
With your available data, try these steps:
- Take an overall website health assessment.
- Look for the top 3 pages with the most traffic, and where most of the traffic is coming from — referral, organic, email, etc.
- Consider any nuances with intent and behavior. Why these pages, and why these sources? This will help you more effectively analyze your MS Clarity session recordings and heatmap data and look for patterns.
- Note any expected and unexpected behaviors in your MS Clarity data. Then, determine how your observations could be shaping the number and quality of leads that you currently see from these pages.
2. Talk to your customers.
This is the most direct way to get feedback, and the insights can be invaluable.
One of the best data points to collect is what almost stopped customers from doing business with you. If a pattern emerges, add it to the problems identified and find a way to solve it.
Here are three questions that can help reveal the reasons prospects aren’t converting:
- “What persuaded you to purchase from us?” — This reveals your strongest message and brand characteristics.
- “Which other brands or options did you consider before finally choosing us?” — This reveals your true competitors. Your perceived competitors might be different from a consumer’s perspective, so this becomes an opportunity to review (but not copy) what they’re doing.
- “What almost stopped you from buying our services or product?” — This reveals the biggest friction point that customers actually encountered in their journey.
If directly asking your customers is challenging, make the most out of your sales and any client-facing teams. They talk directly to your customers and have firsthand experience of their confusions and frustration points.
Document all the identified pain points along with how your team usually counters them to still close the sale or appease the client. These pieces of information are incredibly valuable for improving copywriting and clarity on your site.
3. Explore alternative and non-A/B test designs.
Testing a solution is thankfully not limited to just A/B tests — if it were, CRO might be impossible for low-traffic sites.
However, there are plenty of alternative data collection and testing methods that can work for lower-traffic sites.
Other common approaches to CRO for low traffic sites are user testing, surveys, focus group discussions, and method marketing. These can all help you learn more about your clients and their buying journeys straight from the source.
If you’re still keen to conduct technical testing despite low traffic volume, you have options. I explored this topic in-depth for Experiment Nation in this presentation:
To summarize a few of the most actionable routes you could take to run CRO tests on a low-traffic site:
- Limited A/B tests constrained to upper level conversions or key events. Measuring your test based on form starts instead of form completions will likely increase the volume and power you have. This could be helpful for low traffic sites that follow a multi-step conversion funnel.
- Before and after testing. This approach allows you to ship changes based on your problem-solving idea while using your pre-existing data as the baseline. You’ll see the difference between the experiment results and what you’d expect using CUPED analysis (a statistical method that filters out background noise to show true performance). Just be ready with a backup of the old page in case something goes wrong so it’s easy to switch back.
- Go-for-it or do-no-harm tests. This is probably the most straightforward CRO test for any site. The experiment design assumes that there is really no downside in doing the test and there is no detrimental effect if you lose. In using this as your experiment design, you simply calculate your sample size and test duration then decide based on raw conversions rather than the typical conversion rate
4. Set up guardrails.
Guardrails are metrics that act as protective boundaries or early warning systems to your test. This will tell you when it’s time to pause or stop and redesign the test because you’re doing more harm than good to your user experience or backend performance.
Sample guardrails could be exit rates should not exceed 80% in doing this test or page load time should not slow down to 10 seconds. If you hit your guardrail metric, pause and rethink your approach.
5. Don’t stop testing and learning.
Because you’re running on low traffic volume, it’s difficult to be sure that the results you get from tests are truly repeatable. Even after you have completed a test and decided that it made a positive impact enough to make the changes permanent, always monitor the performance.
Repeat the test setup as needed to help you make more strategic decisions. More importantly, document your learnings. It’s possible to see a different outcome over time, so it’s vital to know what made the test perform that way.
Whether a test shows positive or negative results, there is always something to learn to inform your business strategy and decisions.
At the end of the day, our goal is to find the middle ground between what your business wants to achieve and the problems your customers want to solve or a pain they want to alleviate.
Your traffic volume should never stop you from improving your website and overall experiences to serve customers better. You just need to be more creative with how you approach it.
Need help with your conversion rate optimization program? Nexus Marketing offers CRO services focused on helping mission-driven brands that serve niche audiences (and often see lower-than-average traffic volumes) turn more of their website traffic into actual leads and sales.
Ask us how we’d approach your website’s CRO program, and we can provide some preliminary observations and recommendations.
Already a Nexus client? Reach out to your account team to learn more about our CRO services.
Happy optimizing!
About the Author

Carla Quizon, CRO Lead, Nexus Marketing
Former Head of Client Success Management and CRO at a performance marketing agency, Carla democratizes conversion rate optimization for B2B websites with specialized ICPs and lower traffic. Over the last five years, she has been developing frameworks that prove smaller organizations can also leverage professional data-driven experimentation to inform their strategies.



















































