Latest posts by Shanelle Mullin (see all)
- A Step-by-Step Guide to Optimizing a Low Traffic Site, Without AB Testing - December 25, 2016
Though, sometimes it can seem like testing and optimization are synonymous… It’s discouraging for those who are just starting out and don’t have enough traffic to start AB testing yet.
The truth is that there’s much more to optimization than just testing, so low traffic sites can absolutely still be optimized.
But let’s start from the beginning.
Why You Can’t Test a Low Traffic Site
Before you can run a test, you need to calculate your sample size. There are a number of calculators you can use to help you with this process. Here’s a simple example…
So, you have a conversion rate of 3% and you want to be able to detect a 5% effect. Each of your variations will need to be visited by 204,493 people. Now let’s say your conversion rate is even lower, at 2%…
Even more visitors will need to see each variation.
One last example. Let’s say you want to be able to detect a smaller effect…
Again, you’ll need even more visitors per variation.
Testing is not a matter of opening your testing tool and waiting for that little “Statistically Significant” marker. You need to reach your required sample size before concluding anything or else the insights will be invalid.
That’s where it gets difficult for low traffic sites.
Unless you’re detecting an incredibly large effect, you’ll need to run the test for months. The longer a test is running, the more vulnerable it is to sample pollution, which can rear its head in many ways.
Sample pollution: Different external influencers that affect or skew your tests that result in invalid or incorrect data.
The longer a test runs, the more likely it is that it will be impacted by: holidays, technical issues, campaign launches, visitors arriving via multiple devices and browsers, visitors deleting their cookies (thus, potentially re-entering the test in a different variation), etc.
There are no absolutes, but two full business cycles (2-4 weeks) is a common test length that mitigates those pollution factors. If you’re a low traffic site, it’ll be pretty difficult to get 204,493 visitors to each variation in that amount of time.
So, How Can You Optimize a Low Traffic Site?
Just because you don’t have enough traffic to test doesn’t mean you can’t optimize your site. In fact, there are many other optimization options to consider.
We’ll focus on five: heuristic analysis, mouse tracking analysis, technical analysis, qualitative research, and user testing.
#1 Heuristic Analysis
Opinions have no place in optimization, but heuristic analysis is another story. It’s when you conduct an experience-based assessment of your site based on a number of conversion-focused factors.
Of course, this isn’t an exact science or a perfect method, but it is quick and can often be “good enough”, especially when you’re unable to test. Note that the more optimization experience you have, the better your heuristic analysis will be.
Heuristic analysis is done page by page for…
- Clarity: Is the content, next step and offer on this page as clear as possible?
- Distraction: Is anything distracting visitors from the next step or offer?
- Relevancy: Does the page meet visitor expectation?
- Value: How well is the value communicated? Does it motivate?
- Friction: What is causing doubt, anxiety, hesitation or frustration?
This process is done best in a group environment, but be careful to avoid random comments. Heuristic analysis should be a systematic, repeatable process with rules and guidelines to follow.
Remember that what you identify at this stage is not to be accepted as fact. Instead, you have to validate or invalidate the findings through qualitative and quantitative research.
Tip: Use Notable to collaborate and organize your thoughts to help with your heuristic analysis
#2 Mouse Tracking Analysis
Mouse tracking analysis allows you to record what people do with their mouse and quantify the data using heat maps.
As with testing, you need a large enough sample size to conduct analysis. Fortunately, 2,000-4,000 per variation is a lot easier to achieve. There are two types of heat maps that you’ll find especially interesting…
- Click Maps: Shows where people click (and where they don’t). This is helpful for understanding attention and noticing elements that visitors falsely believe to be clickable.
- Scroll Maps: Shows how far down the page people scroll. This can help you prioritize messaging and calls to action. You might also identify some false bottoms and opportunities for visual cues.
During “Emotion Sells: The Masterclass” course enrollment week, Talia and her team used a scroll map to identify how far people scroll down the landing page and find any specific sections that need optimizing. These insights (for example, over 62% of people scrolled half way through the page) helped optimize the page and strengthen certain messages on the page.
Once you’re done with the heat maps, you may want to consider two related factors…
- Session Replays: Records your actual visitors using your actual site.
- Form Analytics: Shows how people interact with your forms. Which fields produce a lot of error messages? Which fields do people leave empty most often?
Here are some tools that will help you with mouse tracking analysis…
- Hotjar for click maps, scroll maps, session replays, form analytics and more.
- Clicktale for click maps, scroll maps, session replays, form analytics and more.
- Inspectlet for click maps, scroll maps, session replays, form analytics and more.
- Formisimo is, as TNW once said, like Google Analytics specifically for forms.
There’s a visitor using a really old version of Internet Explorer who can’t use your form properly or there’s a visitor getting a 404 page or there’s a visitor waiting 8 seconds for your site to load.
Here’s what to look for…
- Speed: Slow sites don’t convert. The price of a slow load time will shock you, so it’s important to learn how to make your site faster, both on desktop and mobile.
- Broken Links: 404 pages are not just frustrating, they’re disappointing and bad for search. Open Google Analytics, search for “Page Not Found” (or whatever your 404 page is called), click the page and then behold a prioritized list of 404s to fix.
- Cross-Browser: Your site needs to work properly in every version of every browser. Even though you don’t use that browser version and consider it a relic, I promise someone somewhere is still using it. Use the browser report in Google Analytics to help you prioritize.
- Cross-Device: Your site also needs to work properly on every version of every device. Again, I promise someone somewhere is still trying to access your site from a “ridiculously outdated” device. Use the device report in Google Analytics to help you prioritize.
When conducting cross-browser and cross-device quality assurance, be sure to only compare within the same device family or browser family. For example, you want to compare versions of Chrome to one another, not a version of Chrome to a version of Safari.
Here are some tools that will help you with technical analysis…
- Google Analytics for finding broken links, prioritizing cross-browser and cross-device quality assurance, etc.
- BrowserStack for live, web-based mobile and desktop browser testing.
- PageSpeed Insights for improving your site speed.
#4 Qualitative Research
You can learn a surprising amount by asking the right people the right questions at the right time.
Qualitative research takes two different forms for low traffic sites…
- On-Site Surveys: Hit your visitors, either while they’re still engaging or as they’re leaving, with a simple question to understand their hesitations and identify points of friction you might’ve missed.
- Customer Interviews: Pick up the phone and call your customers, new and old. Ask questions that help you understand their buying process, how your product or service fits into their lives, etc. You don’t just want to know how satisfied they are, you want actionable insights you can use.
The insights you gain from qualitative research can be used to write better copy, refine your value proposition and more. Pay attention to the specific words and phrases they use.
Note that you can also consider customer surveys, but you need 200-250 responses from recent, first-time customers. So, if you’re a low traffic site, you likely don’t have the conversions necessary.
Here are some tools that will help you with qualitative research…
- Hotjar for click maps, scroll maps, session replays, form analytics and more.
- Qualaroo for on-site surveys.
#5 User Testing
User testing allows you to watch and listen as real people who represent your target audience interact with your site and comment their thought process out loud.
Typically, you’ll want to include three types of tasks…
- A specific task. Example: Find a four player Nintendo Wii U game for $50-60.
- A broad task. Example: Find a video game you like.
- Funnel completion.
If you have mass market appeal, you can use a user testing tool. If not, you can turn to relevant forums to find candidates. Usually all you’ll need to offer is a $25 gift card to get people interested in giving you 20 minutes of their time.
Whatever you do, don’t ask for opinions or try to guide them in any way. Instead, stick to the three tasks and pay attention.
Here are some tools that will help you with user testing…
- UserTesting for getting audio, visual, or written feedback on your site or app.
- TryMyUI for getting audio, visual, or written feedback on your site or app.
Don’t bury your head in the sand if you have a low traffic site. Just because you can’t test yet doesn’t mean you can’t optimize yet.
- Conduct heuristic analysis to evaluate the site based on clarity, distraction, relevancy, value and friction.
- Use click maps, scroll maps, session replays and form analytics to improve: visitor focus, message hierarchy, and points of friction.
- Conduct technical analysis to improve site speed, fix broken links, and ensure your site works in every browser and on every device.
- Use on-site surveys and customer interviews to better understand your ideal customer, identify points of friction, and improve your copy.
- Conduct user testing to identify and improve points of friction you might’ve missed.
Shanelle does content and growth at ConversionXL and ConversionXL Institute, conversion optimization training for mid-market to enterprise level marketers. She’s a jill-of-all-trades marketer with a background in PPC, SEO, content marketing, analytics and PR.
Powered by Facebook Comments