July 12, 2016 | Sarah Danks


This is a recap of a speech given by Oli Gardner, co-founder of Unbounce, at the MnSearch Summit in 2016.

It’s widely known that Oli’s seen more landing pages than anyone else on this planet. Not only that; he’s a quick-draw critic of conversion-centered design. Show him a landing page and he’ll be rattling off what’s wrong with it — and how to make it better — within seconds.

Oli’s MnSummit presentation was on fleek, per usual — this guy practices each speech down to the second and makes public speaking look like it’s child’s play.

OH, and PS: I can’t proceed one second further without giving him photo credit for all his awesome slides (and the header on this blog) … he took them HIMSELF!!!

Here’s what he had to tell us about conversion equations and the future of marketing.

The future of marketing is conversion automation

As someone who spends most of every day thinking about conversion rate optimization (CRO) and how to make everything lead to more conversions, I can imagine Oli has more information than most anyone else on this topic.

3 most common CRO questions

  1. What should I test?
  2. In what order?
  3. What’s a good conversion rate?

Anyone involved in optimizing for conversion rates knows they need to make it better, test different variations, etc. But it’s not always easy!

4 most common CRO mistakes

  1. Testing opinions; not insights
  2. Testing the wrong things
  3. Stopping tests too early
  4. Focusing on surface-level conversion metrics and not lifetime value

You can’t test things “just because” — you need to know what to test, when, and WHY.

Oli says machines will end this with automated megavariate™ testing of marketing interaction models connected to cohort analysis for the highest customer lifetime value.

What’s “automated megavariate testing” even mean?

Basically, machine learning will determine where improvement opportunities exist. Calls to action (CTAs) will be moved to more optimal positions. #CRObots will run scroll map analyses to ascertain how many readers reached CTA, then offer suggestions to improve.

Video chapter links will appear where the highest engagement spikes occur. Forms will respond with a delightful chuckle when you enter 1@1.com as your email address.

That said, it’s all a matter of math.


Conversion equations will show you WHAT to optimize and the scores will tell you in WHICH order to do it.

The most important aspect of the conversion equation?


The clarity equation will show you in which area(s) you need to improve. Readability issue? Immediacy problem?

No worries.

The clarity equation can be minified to an initialization (because Oli LOVES those):


Distraction + Expectation + Readability + Visual Identification + Immediacy + Specificity + Hyperbole

If you’re “low” in any of the 7 areas, you can pinpoint what you need to work on to solve your clarity issue.



Distraction means there are too many things to do on landing page. What’s the goal??? PICK ONE THING.

Immediacy: Can people understand your value proposition within 5 seconds? Ask people, “What do you think this page is about?” It might be really obvious (to YOU), but clearly it’s lost on visitors because despite all your best efforts you STILL don’t just tell them what you want them to do!

Hyperbole: Are you explaining what you do? Or, just saying how good you are at doing it?

Hyperbole is, without a doubt, the single greatest thing in the history of the universe. <ha>

What you’re offering isn’t not unique…

…or maybe it IS. But there’s too much hype. (e.g., Award-winning, #1 marketing platform, fuel your brand, marketing innovation technology…)

Wondering if you’re using too much hype? From the company that brought you landing pages —

— er, from the MARKETING SOFTWARE company that brought you landing pages…

…presenting Unbounce’s Dejargonator, a nifty Chrome extension that helps you “eliminate meaningless marketing speak that makes your landing page less persuasive.”

The future of marketing:

Copy will be analyzed as you type, and words that don’t add clarity will have more impactful alternatives automatically suggested.

Readability: Is it easy to read? Possible to read? Is it an enjoyable read? Is it fast to read?


A quote from Joseph Bentzel about Pagefights (that awesome live landing page critique that is no more):

“I’m glad I saw this on a lazy Saturday when I had time to immerse myself in this lukewarm shower of digital hipster parochialism seamlessly integrated with slightly veiled sado-brogrammer invective.”

Let’s measure the text readability of that. The Flesch-Kincaid Reading Ease score measures how easy your copy is to read. 0 is impossible; 100 is perfect. That quote receives a score of 18.4, which means it’s not very easy to read.

Something else that factors into text readability is use of acronyms. Acronym count — or, what ratio of words in your headline can be attributed to potentially unknown acronyms? (e.g. CFOs opt for Oracle ERP Cloud)

A design element that affects readability is contrast ratio: the contrast between your headline and the background behind it.

Then there’s Motion Reading Facility Ratio (MRFR) — what percentage of the available text can be read before it disappears? Think scrolling carousels. If you managed to read 10 out of 45 words, the MRFR = 0.22. Hint: that’s not good.

Another design element is Typographic Golden Ratio. How close is your paragraph sizing to the ideal for web readability? Using the Golden Ration Typography Calculator, you can learn the ideal type for readability. Ideal ratio is: 16px font size, 25px line height, and 680px paragraph width.

In light of this readability stuff, Oli offers a quick way to help increase your readability score:

Simple To-Do List

  • Turn off slider autoplay
  • Simplify the copy
  • Narrow the paragraph width
  • Title-case the headline

Visual Identification — when viewed in isolation, can people tell what your image means? Or what the purpose of your product/service is?

For example, check out this landing page and see if you can tell — at a quick glance — what product you think this company sells:


In a poll conducted, 6% answered business loans. The rest were all, “Dunno. Viagra? A talk show? A doctor consultation? Something about old men? Meeting a guy? Not sure.”

When Larry King’s (random) image was removed, 36% answered with “business loans.” That’s a 500% increase! Sorry, Larry. You do not help the cause.

Here’s how to fix visual clarity problems — use Jaspersoft 6.2.

Conduct a 5-second visual identification test to ask, “what do you think this software does?” Look at the image with and without a caption and see if it makes sense.

Expectation — when viewed in isolation, does the CTA describe the goal of the page? And, what will you get when you click it?

E.g., a landing page with “Get the details” as a CTA vs “See Moz Products.” Which do you think does better in the clarity equation? Yeah, the Moz CTA, naturally.

PS: Calculating clarity score is easier with a calculator 🙂


What would the learning machines do to improve the equation?

Alter the coefficients, of course…

…but they’ll need help understanding how we react to our analogue senses.

If your design is confusing people, you might need a little A.D.D. (again with the initializations!)

ADD = Attention-Driven Design

When you can identify a design problem, you can use A.D.D. principles to solve it

Because, design matters.


Attention-driven design does many things:

  • Creates joy
  • Reduces confusion
  • Gets people out of the way
  • Reduces germs
  • Removes frustration

Data-informed design accelerates delight.

But what if I’m not a designer, or a developer — how do I create mockups for those tests?

The 3-minute no-permission-required conversion research hack you should all be using: Usability Hub. Find qualitative insights, show stakeholders to get buy-in, start making change.

By the way, MnSearch, Oli went ahead and took your own CTA and offered a little CRO help:


…told you he’s quick to criticize landing pages 🙂

You’ve gotta start somewhere. Probably with the first rule of CTAs:

Have a f#cking CTA!

EVERY website and landing page has conversion opportunities, and EVERY conversion opportunity needs a (targeted, timely, relevant, unique, delightful, appropriate, show-me-once-value-added) CTA.

The video equation:



Whoa. That’s a lot to take in.

But hang on, there’s more:


One thing you can say with certainty about Oli Gardner: he sure loves jacked-up, convoluted conversion equations! Way to try to deduce every single algorithm there is, Oli. Hang on a tic. It’s an Oli-gorithm!

When using video, you’ll want to know which words influence the click-through rate of video annotations. But how to tell? Ask the data.

Want to know which words influence the click-through rate of full-screen CTAs? Ask the data.

PS: Research has shown annotations and CTAs are opposites! CTAs vs annotations:


In the future, CTAs will have autocomplete recommendations of more impactful options, and the expected increase in conversion:


Here’s the thing: Every conversion opportunity is unique. Which is what makes us human.

Which leads us to social proof. But Oli warns to be careful what you put inside quotation marks.


Social Proof — the success of your social proof lies in your ability to demonstrate the transformative effect of the user journey created by your product or service.

Social proof equation: FLATS (MOAR initialization!)

Social proof = Reading Facility + Linking + Authenticity + Transformation + Security (F.L.A.T.S.)



Take these 2 testimonials and their social proof score.

First, Juan Do:


And now, Andrew Miller:



You can see how a vague — albeit glowing — testimonial a) without an accompanying portrait, b) that isn’t results-oriented performs in comparison to the winning type of review.

Future of marketing:

Oli says in the future, automatic image source lookups will determine if an image is stock, and recommend a change, while visual analysis will check for unrealistically white teeth to do the same…


(Image Source: weheartit.com)

…Okay. Ross is my addition; Oli didn’t include him in his presentation. But, I mean — how could I NOT???

Moving on.

Linking & Security

Linking in a testimonial — does the testimonial HAVE a link? If so, does it link to an external page, or a Lightbox?

Trust seals: security vs familiarity. Are security seals being used appropriately for your audience? I.e., if your audience is female vs male, older vs younger, etc. Different “seals” can carry different connotations of safety.

(Security info from ConversionXL)

Future of marketing:

Trust seals will be targeted precisely to the demographics of our visitors.

Pro tip: Interview your testimonial givers to mine every element of the social proof equation rather than using a generic, watered-down platitude.

Think pop-ups suck? That they’re evil? That no one likes them? That you should never use them?

Think again!

Oli reiterates that ANYTHING can be delightful:

  • emails
  • banners
  • and even popups


But, but…

…how can pop-ups actually be successful???

Because: relevant. Valuable. Well-timed. Designed to delight.

“Technology isn’t the problem — we are.”

~ Oli Gardner

Machines will learn our value systems.

And that, my friends, is the future of conversion rate optimization — and marketing overall.

Want to see Oli’s presentation in all its glory? Check it out here.



Subscribe To Our Newsletter

Our newsletter includes industry news, tips, and upcoming events!