Stop optimizing your website's conversion rate and do this instead

Conversion rate optimization and A/B testing might be wasting your time and resources.

Find out why CRO is not the best solution for your early stage product and 3 other alternatives that can make a much better impact on your growth.

By Sumit Hegde

Thinking about spending a couple of hours improving your website conversion rate?

Maybe contemplating changing your CTA? Or perhaps, the images?

Or you're thinking of running some A/B tests on your landing page, possibly?

Hold up!

Because you're going to want to hear this.

Conversion rate optimization is WASTING your time and resources—because those rates do not give you the full picture.

They're just a bunch of numbers that tell you your landing page isn't working the way it should, but the WHY it's not working...that's anyone's guess. 🤷🏽‍♀️

More is not always better.


You could have a 40% conversion rate, but if more than half of your users churned the day after they converted, what does it even matter?

If you're just converting less qualified leads, it's a matter of time before the effects start to show.

Higher customer acquisition costs for lower returns.

Is that something you really want?

Conversion rate is just one small piece in the puzzle that is your business: and a corner-piece at that.

It is a flawed metric that hogs all the spotlight and distracts you from the real stuff you should be doing to polish your website and bring in the right customers.

Before you get into conversion rate optimization, ask yourself this: Do I even know if this will work?

But wait: Why trash talk A/B testing?


Because it's mostly slow, frustrating, guess-work.

And the results are usually dismal.

Look, if you have less than 1000 conversions per month, there's literally no point in A/B testing.

Fine, you improved your conversion rate by 5%. You know what 5% of 400 is? 20.

You got 20 new customers, yay.

This is not going to move your ROI in a big way.

How do you even know those 20 customers converted because you ran those tests?

20 customers is not enough to judge the validity of your tests: these guys could have just come because one of your customers aired their love for your product on social media, you know?

Correlation ≠ Causation.


And all that time you spent micro-managing the A/B tests: designing the versions you're assessing, setting it up in your testing tool, you could have spent understanding how to really, REALLY make bigger revolutions happen.

A/B testing is great, but it's not for everything and everyone. You gotta get other s@#t sorted first.

What about data from analytics?


The fact that 800 people bounced off your site doesn't tell you much.

I always say: funnels are not as straightforward as we always assume them to be.

People act in unreliable ways for insensible reasons—assigning a singular reason to the act of bouncing off your site would be an error.

There's always more to the picture than what is assumed from numbers.

Now, applied right, analytics, A/B testing and conversion rate optimizations can give interesting results.

But these are the finishing touches to your cake—the icing and the cherries.

NOT the cake batter, NOT the sponge.

And without the foundation, it's just not working.

What is the cake part of the cake then, you ask?


This image by Des Traynor explains it well:

A/B testing and perfecting the website through optimization of individual bits is REFINEMENT.

And refinement doesn't allow you to solve the real, damning problem, making you miss out on your best customers.

Exploration, on the other hand, allows you to understand what works best for your business and your customers. Exploration means this:

Identifying your best customers.


Look if you actually want more conversions AND qualified leads, you need to know who your ideal customers are.


Because the wrong customers do not want your product.

They don't need what your product offers.

They are not willing to put in the effort required to reach the outcomes your product promises, they will not contribute to your goals and they WILL churn.


Basically, they are the person in a group project who simply sits back and instructs you on how to do it right, doing nothing, sipping a cold mocha. We all know one of these guys.

The right customers are people who will genuinely benefit from your product—these are people who will stick and bring in the revenue and help you improve the product in ways that bring you closer to your goals.

So how do you find the best customers?

Here are a few clues:

  1. People who are already paying you.
  2. People who might benefit from your product.
  3. People who are using a competitor's product.

Look at Grammarly, for example. They really know the audience they're targeting: Millennial professionals, students writing term papers, generic email-writer, bloggers and that's who they make their ads for.

Translate this to your website, and damn!

Understanding them.


Let's go back to the Grammarly video above.

The advert shows a deep understanding of the customer's problems, the solution the customer needs and the outcome they want.

Without really understanding a customer, you're simply throwing darts around a room while spinning really fast and hoping to hit the bull's eye.

The most you're going to do is provide a very free and very unnecessary acupuncture to your dear friend. I speak from experience, do NOT do it.

INSTEAD do this:

  • Learn what their problem is.
  • Why they need your help to fix it.
  • Why your solution is perfect.
  • Find what outcome they wish to reach.

Go over user session recordings, do customer interviews, read customer support chats, MAKE IT HAPPEN!

And while you're at it, segment your users!


There's no one size fits all.

You probably cater to different types of people—just like Grammarly does.

So why send them the same generic message when you can divide and rule? Worked for the British.

Imagine you have a customer support tool: this would be useful for many types of businesses! E-commerce websites, other SaaS start-ups, educators and consultants.

You can't sell them the same generic messaging to all of them.

Create landing pages specific to these segments' needs, their problems and the solution they require!

Look how awesome-ly Webflow does it!

Another kind of segmenting you can do is based on where your users are coming from.

  • If a general Google search or an Instagram ad led them to you, the visitor probably needs a bunch of educating and nurturing since they do not know anything about your service!

    —Get them to commit smaller first, maybe. Read a blog, sign up for your newsletter!
  • If a visitor reaches your site from a email-campaign you have been running, they sorta know what you do and how. Now you need to reinforce the benefits of your product and voila!

    — You can usually have a bold CTA asking them to purchase the product at this stage!
  • If they're coming from the referral of an influencer in the field, they probably know a bit about the service you provide. Now you just have to convince them why you should be their choice!

    — Ask them to book a demo call, or direct them to case studies!


Basically, you need to create landing pages based on WHERE the visitor is in the funnel.

Are they aware of your product? Are they actively researching your product? Are they almost sold, but are comparing you to other options? Have they already decided?

Now, for the actual building:

Rather than toggle with multiple versions of a mediocre site, build a strong site that takes all the above into account.

A website that talks about the problem the specific section of customers is facing, the reason they need to fix it, how your solution comes into picture, and drive it in with some social proof.

With a site like that, you're ready with a cake that makes mouths water. (Okay, these cake metaphors are making me really hungry!)

With a site like that, you get customers who are invested in your product and care about it.

THIS is what your company needs to be aiming for: not some arbitrary numerical goal.

Once you have this in place, your conversion rate will automatically go up and then...only then, should you work on A/B testing and optimizing your site. Because...

And when you do run tests, don't analyze things that don't matter. The color of your CTA won't change much, but your value proposition can.

BE SMART—test things that make a difference!

If you want a PDF of the website template I use to bring CRAZY results for my clients: head over here.

I promise it'll make a huge difference in the number—and the quality of customers you get!


Thoughts? Questions? Feedback?
Connect with me on Twitter:

By Sumit Hegde

I help SaaS businesses make more money by creating badass websites. And by badass I mean high-performing, flawlessly positioned and customer-centric.
Learn more about what I do
PSST!
Want the exact process to convert 2x more customers on your SaaS website?
Thank you! Your submission has been received!
Oops! Something went wrong while submitting the form.
What to read next
Breaking down the conversion secrets of Slack's new website
read more

About time you join the Unicorns!

I like you! Wanna get on my secret list?
I’ll send you the best SaaS strategies, teardowns, case studies and insights I get working with other SaaS clients.
Thank you! Your submission has been received!
Oops! Something went wrong while submitting the form.