7 Conversion Rate Optimization Experts Answer the Questions You Were Too Afraid to Ask

2014 was the year “Conversion Rate Optimization” became the new “social media marketing.”

Hype riding marketers published story after story about how these “tiny little tweaks” were producing enormous gains, and it got you fantasizing about double and triple digit percentage increases.

Maybe you dipped your toe in and experimented with some of the “tips & tricks” only to be disappointed, wondering why you ever wasted your time.

But here’s the thing, “Conversion Rate Optimization” isn’t a bible of tactics guarded by a mystical circle of cloaked beings high atop a mountainside.

The truth is, a good deal of the Conversion Rate Optimization process isn’t about applying tactics, but by sometimes asking very basic questions of your customers and your data.

I’ve rounded up 7 of my favorite Conversion Rate experts and asked the questions that most of us would feel are too simple, too basic, and too stupid to ask.

Their answers are surprisingly full of insight and will get you thinking more strategically about improving your conversions.

How Do You Define Conversion Rate Optimization?

Chris Goward - WiderFunnel

“Conversion optimization, which I prefer to think of as ‘Marketing Optimization' is a fundamental mindset shift in itself.

It’s a shift from so-called ‘best practices’ to ‘test practices’ – from 'seeking tips and tricks' to 'asking powerful questions' – from speculating to hypothesizing.

A big mistake marketers make is relegating marketing optimization to a tactical department rather than approaching it as a strategy.

You can only win if you move from tweaking and tuning landing pages to posing important hypotheses to understand your audience segments.”

When design & copywriting teams fall into the “Yeah, but I think it would be better if… “ trap, they unknowingly create a culture where sacrifice, compromise, and butt kissing hold more value than creating real revenue.

Conversion Rate Optimization (CRO) is a mindset shift. It starts with data, not opinions, and gets you paying more attention to your customers, their behavior, and why they do the things they do.

For example, maybe mobile buyers spend more than desktop buyers in certain product categories? Why is that? What data can we look at that explains the behavior, and what is best way for us to connect with those customers to learn more about their shopping experience?

These answers aren’t found in anyone else’s best practice guide, and while internal opinions are valid, only your prospects, leads & customers will ever have the “right” answer - which you can use for future test hypotheses.

Where Do You Look For Testing Ideas?

John Ekman - Conversionista

"We analyze the ecommerce funnel and break it down to four steps.

So from a conversion/persuasion point of view you have four points of influence:

1. Getting people to stay (reduce bounce rate)
2. Increase product views (now available in GA enhanced ecom tracking)
3. Increase add to carts
4. Increase checkout rate

So our major take on ecommerce optimization is that “Overall Conversion rate is meaningless, if you don't know in which commerce step you have your biggest problem!?"

So, make sure you don’t waste your time optimizing in the wrong place. First figure out where in your ecommerce funnel you have your biggest problems - then create your AB-testing hypothesis."

Running an online business without a basic understanding of analytics is exactly like grabbing the pilot’s stick for the first time and expecting to win first place at the air show. You wouldn’t fly a plane without taking a few lessons first, would you?

If you do not understand analytics, or have someone on your team that does, you need to make this a priority immediately.

Analytics isn’t just something only the nerdy kids use and pull insights from, but rather a way to understand without question where your visitors are coming from, what they’re doing once they’re on your site, and most importantly, where they are getting stuck when it comes to buying from you.

Matt Gershoff - Conductrics

"One of the main benefits of analytics is that it provides a framework for impartially evaluating decisions, regardless of the position of the decision maker.

Not surprisingly this can lead to enormous internal pressure to skew results so that they are favorable to the ideas and suggestions championed by the 'higher-ups'. Because of this, it is extremely important that your testing/analytics program be granted some degree of independence so that it can be insulated from the politics of the organization.

Otherwise you can wind up with Testing and Optimization Theater - it's all just for show and a total waste of time."

Even with a basic understanding of analytics, you can know for sure where the problems are on your site. Once you have that, you can create a testing plan that addresses those problems and works towards very specific outcomes.

How Do You Know What To Test?

Brian Massey - Conversion Sciences

"If you expect someone to behave logically, they will let you down. If you ask them what they want, they will lie. If you ask them why they did something, they will invent a story.

If you ask them to spend money, they will be become very truthful. Watching people in the act of loosening their grip on their pocketbook is the most reliable data you can collect."

Once you know where people are getting stuck, and why, you need to combine your own intuition with the information you’ve gained from your customers to create tests.

Gathering customer feedback can be done with user testing, heatmap tools, customer surveys and interviews, and any other method that gets you into a dialog with real customers.

Brian isn’t saying you shouldn’t be doing these things to understand your customer behavior, but rather, that each of these methods, along with your own professional insight, is what will help you to triangulate the truth and find out what really gets them to pull out their credit card make a purchase.

How Long Should You Run A Test For?

Peep Laja - ConversionXL

"If you stop the test early, your results will be invalid. You can't test faster just because you/your boss/VCs want to move faster - that's not how math works.

You need to have the right sample size (pre-calculated before you start the test) and test duration (2 business cycles, so 3-4 weeks for most). If you pick the wrong variation as the winner, you render the whole testing process useless."

Once you know where and what you should be testing, the next question most people have is, “how long should it run for before I know the results are conclusive”?

The answer is going to vary from website to website, and honestly, depending on what your traffic and sales look like right now, could make some people uncomfortable. But like Peep is saying, without patience, you could end up picking the wrong variation and void the entire testing process.

This is an important rule-of-thumb to accept as you dive deeper into CRO, as it will give you more accurate results, and free you up to work on multiple tests while existing tests “bake to perfection.”

Recommended Reading: Master the Essentials of Conversion Optimization

My friend Judah Phillips adds a little more context to what Peep’s saying:

Judah Phillips - Smart Current

"CRO requires a different mindset than other marketing activities. You can't simply "test faster" or confidently make quick decisions based on preliminary data.

Like fine wine, CRO takes some time to mature.

A winning test today could invert results tomorrow and become a losing test, just like a losing test now could shift to become a winning test later.

It takes considerable planning, patience, and customers exposed to tests across business cycles to make CRO work. People must understand this fundamental mindset when approaching CRO for their business."

Also, know that one major misconception about conversion rate optimization is that it begins and ends with A/B testing. In many scenarios, an A/B test might not be the best test to gain the insight you’re looking for.

When Do You Stop Testing And Go Full Redesign?

Oli Gardner - Unbounce

"When it’s clear that you’ve hit a local maxima (you can’t move the needle anymore), it’s typically time for a radical redesign.

Or you can explore channel segmentation. For example, the Unbounce webinar registration landing page converts like this:

• Our email list: 72%
• Partner email lists: 50%
• Organic search: 30%
• PPC: 17%
• Social 7%

We’ve hit the local maxima for our own email list, but there is a completely different level of context/trust associated with the other channels, which opens them up as a bigger opportunity for optimization."

It’s easy to forget that your customers might not be looking at your site every day, and just because you are tired of looking at it, doesn’t mean they are.

Redesigns are tricky, and unless you’re changing your branding, or addressing usability issues that require a major overhaul, may also be unnecessary. Also if the redesign isn’t handled with a great deal of respect for data, can absolutely demolish your conversion rates.

Oli makes a great point about working improving conversions through channel segmentation, and using tools like VWO (formerly Visual Website Optimizer) or Optimizely, you can show a slightly different version of a page to Facebook visitors than you would for PPC visitors.

This allows you to essentially “keep what’s working” for one traffic channel, and make iterative improvements for the others.

The cool part about using strategy is that when you do go full redesign, you’ll have a deep understanding about what’s working for each traffic channel, and can integrate those things into the new design without making blind sacrifices and losing out on tons of revenue.

About The Author

Tommy Walker is the Editor-in-Chief of the Shopify Plus blog. It is his goal to provide high-volume ecommerce stores with deeply researched, honest advice for growing their customer base, revenues and profits. Get more from Tommy on Twitter.