Top 5 CRO Tips That Actually Get You Results

Five things decide whether a conversion rate optimization program produces results: audit before you assume, build the page for the awareness level of the people landing on it, offer one path, fix mobile and speed, and run the whole thing as a scientific process. Everything else is a tactic. These five are the foundation, and without them your tests are built on quicksand.

Here is the analogy I keep coming back to. If you were an astronaut on the International Space Station, you would have to exercise two and a half hours every single day. Not because it is pleasant. Because without it your bones weaken and your muscles atrophy, fast, and you become useless up there. There is no way around it. CRO has the same kind of non-negotiables. Skip them and the program quietly stops working, whatever clever tests you run on top.

1. Start with a CRO audit

You are not your target customer. You will not find their wants and needs inside your own head. You have to get those from them, and the way you get them is with a CRO audit.

Optimizing a page means optimizing the experience of the people using it, and you can only optimize that experience by looking at their current behavior. Otherwise you are deciding on gut feel. As Henry Winkler put it, "Assumptions are the termites of relationships." And make no mistake, you are in a relationship with your customers. If you want it to work, do not assume you know what they want. Let their behavior do the talking.

Two things to do first.

Install a heatmap and watch some recordings. Heatmaps and session recordings answer questions you cannot answer from a dashboard:

• Do visitors scroll and read the page, or do they bounce almost immediately?
• Which parts of the page hold attention, and for how long?
• Are they clicking things that are not links, which usually means the design promised an interaction it does not deliver?
• Where exactly do they stop on a long page, and what is sitting at that point?

Dig into your analytics. There is a lot of information about your visitors sitting there unread. Compare conversion rates for the same page across desktop and mobile. If mobile is much lower, you have found an opportunity rather than a mystery. Do the same across traffic sources, new versus returning visitors, and your top landing pages.

Add one more source if you can: an on-site survey. A single question at the right moment ("What almost stopped you from buying today?") often explains the drop-off that the heatmap only points at. If you want the full method, we wrote it up in the guide to creating a CRO audit.

Do this and every campaign starts on solid ground, optimizing pages against your customers' needs instead of your own taste.

2. Know where the visitor is in the buyer's journey

Timing is everything. It is the difference between an excited yes and a flat no, and when you are building pages meant to convert, it is not optional.

Your visitors are at different points in the buyer's journey. Some are avid fans, ready to take anything you offer. Others found you through a search ten seconds ago. You cannot sell to both the same way. The words you use to explain quantum physics to a ten-year-old are not the words you use with a physicist. Pages work the same way, which is why every page should be tailored to the awareness level of the people who reach it.

So before you build, identify who lands on that page, then build for where they are. That is how you signal that you understand what they want, and that trust is what makes the rest of the page work.

How do you find out their awareness level? Start with the search terms and sources that bring people to the page. Someone further along uses words like:

• buy
• compare
• brand A vs. brand B
• price, discount code, shipping

Someone still in research mode uses words like:

• why do I have... (a symptom)
• how can I get rid of...
• what can I use for...

Once you have that, the language, the value proposition, and the path you set out can all be aimed at that specific audience. Same product, different page, much better conversion rate.

3. Offer only one path

People are busy. When they arrive, they want to be guided. So give them one predetermined path that leads to one action, remove the distractions, and steer them towards it on purpose.

A few concrete things you can do.

Use visual cues. Direction matters: arrows, images of people looking towards the button, contrast on the CTA so it stands out from everything around it. And make white space your friend, because a button with room around it is a button people see.

Resist the urge to offer choices. On a dedicated landing page, strip the navigation menu. Cut the secondary links, the "you might also like" blocks, the third offer in the sidebar. Keep one call to action, repeated as often as the page length justifies.

One caveat worth stating, because "remove the navigation" gets repeated as gospel. It is true for dedicated campaign landing pages. It is not automatically true for a category page or a product page inside a store, where navigation and search are how people find what they came for. Removing navigation is a hypothesis to test on those pages, not a rule to apply.

When the page does have one job, visitors see immediately what they can do. If it also matches their awareness level, they are far more likely to do it.

4. Don't forget the tech: mobile and speed

Open your analytics. What share of your visitors are on a phone? If your site is anything like the ones we work on, mobile is the device of choice for most people who come to you.

Here is the thing. People on a phone are in a different frame of mind from people at a desk. They have less time, they are moving, they are doing three things at once. They experience your site differently, so build for that situation and you set yourself apart from most of your competitors while making life better for your existing customers.

The size of the prize is easy to see. Smart Insights puts the global mobile conversion rate at about 1.8 percent against about 3.9 percent on desktop. Mobile carries the traffic and converts at half the rate. That gap is the most under-optimized ground on most sites, and there is a lot of room to close it.

Design for mobile, do not shrink the desktop site. Dumping every element of the desktop page into a narrow column is not mobile design. Decide what a phone user needs at each point, put the primary action inside thumb reach, and cut the rest.

Then fix speed. We have been warned about this for years, and yet with clients who come to us for CRO work we still see site after site falling short on it. Google research reported by Marketing Dive found that 53 percent of mobile visitors abandon a site that takes longer than three seconds to load. If they leave, they never read the page. And you cannot sell to someone who is not there.

Test speed on a real phone on a real connection, not on your office wifi with a desktop emulator. If you want the longer version of this tip, we go deeper in CRO for mobile.

5. Make it a scientific, data-backed endeavor

This is usually where someone says, "I can't do that, it's too difficult, I'm not good at numbers." And this is where I say: this isn't school, hombre. You don't have to do it by hand.

I was not into numbers at school either. But we have technology on our side. CRO tools implement the strategy and do the arithmetic for you, which means you can learn from experiments quickly and improve tests day after day. That speed of learning is the real product, not the individual win.

Here is what I mean by that. We once tested a new layout on one page of a test site. It worked well, so we took it further and applied exactly the same change to ten other pages on the same site. Six of them increased conversions. Four dropped sharply. Our intuition that it would work everywhere was simply wrong, and we would never have known without the data.

So a CRO strategy has to follow a scientific process and be backed by data. That is how you learn about your visitors quickly, iterate winning tests, and raise the chance that the next test succeeds too.

How to apply all five to one page

1. Audit the page

   Put a heatmap on it, watch ten to fifteen session recordings, and pull the analytics split by device and source. Write down what surprised you. That list is your raw material.
2. Name the awareness level

   Look at the search terms and referrers. Decide in one sentence who this page is for and how ready they are to buy. If two very different audiences arrive in similar numbers, that is a case for two pages, not one compromise.
3. Count the competing actions

   List every action the page offers. Keep the one that matches the awareness level and remove or demote the rest. Repeat the surviving call to action down the page.
4. Open it on a real phone

   Time the load on a normal connection, then complete the whole flow with your thumb. Anything that makes you pinch, squint, or type more than necessary goes on the fix list before any design test.
5. Turn the biggest finding into one test

   One hypothesis, one primary metric, a sample size and duration set in advance, and no peeking-and-stopping. When it ends, segment the result by device and source before you call it.

The table below is a quick way to check that you have actually covered all five on a page, and to know what each one should change if it worked.

Frequently Asked Questions