A When, Where and How Guide to Conversion Rate Optimization For Startups
Sure, you came up with an outstanding idea for a product, but don’t pat yourself on the back just yet.
This is only the first of the 5 phases of a startup lifecycle. And while it’s important that you get it right, what and how you chose to do next can make or break your business.
To state the obvious, growing a business takes time, dedication and not least some fundamental know-how. It’s the latter we’re going to address in this article.
Knowing when to optimize conversion rates is almost as important as knowing how to do it.
When Should You Start Optimizing?
Rush in and you’ll just be burning money and wasting time with nothing to show for it.
Do it too late and you’ll miss out on a lot of opportunities and will be left with an awful lot of things to patch and fix.
According to Morgan Brown, the perfect moment to start optimizing is in the second stage – when you already have an MVP (Minimum Viable Product).While there is no doubt you should put conversion optimization into your strategy as user growth starts to scale out, it’s crucial to remember that what works for a company that has hundreds of thousands of visits per month, won’t work for a startup that barely has 2000.
With this in mind, let’s explore the most effective and accessible ways startups can boost their conversion rates.
Don’t waste time on A/B testing if you don’t have traffic
During the past few years, A/B testing became the most popular Conversion Rate Optimization method among marketers and even though all the hype it received is obviously not for nothing, I have to say that it has been proved many times to be unfit and inefficient for startups.
This happens due to the fact that startups often don’t have enough resources to run split tests.
For one, the procedure can be rather costly, and then there’s also a painful lack of time and traffic.
It’s not a secret that you need a large enough sample size in order to conduct an A/B test, but why is that?
Both variations should be seen by a sufficient number of users, which varies from case to case, before you can attribute your experiment’s success or lack of it, that is, to your hypothesis rather than sheer coincidence.
Keep in mind that it’s best to aim for a statistical relevance of minimum 95%.
So the question you should be asking is:
“How long does it take for one of the variations to outperform the other by 20% ?”
For a startup, it can take months for a single A/B test to reach statistical relevance and get validated, so now is not the time and place to experiment with button sizes and colors.
What you can do instead is go for the big impact changes, like testing two completely different landing pages.
While it can seem like a bold move, it will save lots of time you’d otherwise spend waiting to see which button converts better.
The good news is, A/B testing is not the only way to optimize your funnel.
Set up your web analytics tools right
“If you can not measure it, you can not improve it” – Sir William Thomson (First Baron Kelvin)
Setting your Google Account properly and tracking your data is crucial if you want to know where your sales come from and which marketing channels bring the best return of investment.
But these are not the only benefits you should be after.
Other dimensions you should be watching closely are your exit and bounce rates and on which of your website’s pages they peak. This will allow you to identify your pain points and focus on them while optimizing.
Watch out for these bad guys!
At this point, it’s worth mentioning a few of the most common mistakes people and companies make when setting up a Google Analytics account:
- Not filtering out internal visits
For more accuracy, you should make sure to create a filtered view out of which you exclude your company’s IP address. Just think how many times a day you check your company’s website, and then imagine what impact several people doing that regularly can have on your monthly report.
- Forgetting to set up goals
Depending on your type of business, your website can have different goals.
Take, for example, a blog. Its conversions would mainly be time spent on page, engagement or maybe signing up for a newsletter.
Things are not the same for an E-commerce website, where the main conversions are purchases or completing a form.
Setting up these goals can give you a much clearer, much needed overview over how your visitors engage with your website and how they progress down the sales funnel.
- Failing to segment your users
Segmenting your users can give you priceless information about their behavior, demographics and geographical info, things that will come in handy when building your buyer persona.
- Not making the best use of the UTMs
UTMs (Urchin Tracking Module) codes are tiny bits of text you can add at the end your links in order to individually track your campaigns.
Each one of them contains information, such as Source, Medium, Content (optional) and Campaign that it communicates to Google Analytics (or other analytics tools).
This way, you are able to see which ads, campaigns, content and marketing channels bring you the best bang for your buck.
Collect qualitative data from the website’s visitors
There are two main types of data you can collect and make use of:
Data obtained through quantitative research
- focuses on numbers and quantifiable insights
- transforms those numbers into usable statistics
- it’s thoroughly accurate
- it’s rigorously structured
- you can mostly get you hands on it through online polls and peeking into web analytics and heat-maps.
Even though it sounds like an enticing basis for website optimization, it has one major drawback: it needs a larger sample of population, a thing that we already established you don’t have yet.
Data obtained through qualitative research
- is what you should go for
- It’s harder to interpret
- focuses on who your users are and what they expect from your product
- It’s indispensable to conversion rate optimization
- helps you build buyer personas
- helps you adjust your unique value proposition and copy to address your visitors using their own vernacular
Qualitative Research Methods
But how can you actually conduct qualitative research? Below, I listed the most popular and efficient methods.
When it comes to qualitative data, surveys are your best tool in the shed. Used right they can tell you (almost) everything you need to know about who your customers are.
Like with everything in marketing, you should decide what your goals are and craft your question in a way that brings you closer to meeting them.
You can trigger surveys on load, on exit or on scroll and while it’s generally better to keep them short and sweet, there are special cases when this won’t work.
A buyer persona survey is a good example for that.
In order to have a complete grasp on all the aspects surrounding your users, from age and sex to income and lifestyle, you need to address between 10 and 15 questions.
So how do you keep your respondents engaged? The best way is to offer small rewards, like an ebook, a voucher or a discount.
Among the things you should look for are:
- Friction (what stopped or made the buying process more difficult)
- What drives their purchase (what problems your product solves for them)
- What influenced their decision to chose your product (always good to keep an eye on the competition and their answer could give you a complete different vision on your Unique Value Proposition
Some examples of questions you should make sure to ask:
- What made you chose us?
- What would you improve at our website if you had the chance?
- Did you consider buying from any competitors before?
- What stopped you from purchasing?
- What would determine you to finish your purchase?
This is the closest thing to the real deal you can get. Though it’s highly effective and it saves you lots of precious time, this one, can prove to be a little bit over your budget.
Even so, done right, it can have a significant impact on your conversion rate and it’s likely to shed some light over your users’ behavior.
The whole process consists of crafting a list of specific, but not too detailed tasks and observing someone interacting with your website for the first time, as they comment along.
Hopefully not your users…
There are two type of user testing:
- Over the shoulder user testing – real-time sessions
- Remote user testing- observing the user remotely via audio and screen capture technology
Here’s a couple of tools you can work with for this purpose:
Live chat transcripts
If you don’t have live support on your E-commerce website, stop whatever you’re doing and make sure you implement it now!
While most strategies in marketing have promoters and detractors, the benefits of live chat support leave little to debate.
Additionally to topping all other customer service communication media, a mere glance over statistics shows an increase of 38% in purchases and a decrease of 30% in cart abandonment rates.
Pretty swell, right?
The trick is to know how to get the most out of it, once it served its initial purpose. Getting the transcripts, going over them and identifying patterns will provide some precious info on the most encountered:
- sources of confusion and friction
- information that might be lacking from your website
- frequent feature requests
You can even use the information from these transcripts to create FAQs or compelling content for your blog.
So, get your hands on this secret gold mine and start digging for cues! What you’ll find might surprise you.
Collect leads using interactive overlays & surveys
As the Lean Marketing Framework (AARRR) shows, your work doesn’t stop after the users made a purchase.
Stop for a moment and imagine you meet a great girl or a charming guy and you walk over to say “hi”. They engage in conversation and seem very keen on getting to know more about you, so when you suggest going for a date, they accept straight away.
Now imagine having an amazing dinner by candle light, enjoying your date’s company and not taking their number at the end. Besides it being rude, you also miss out on a lot more fun dates in the future.
Much like in the date analogy above, it’s great you managed to convert your users, but if you want to get the most out of this newly sprung relationship, the ball is in your court.
Acquiring customers is the expensive part. It is much cheaper to spend your efforts on retaining them.
After the initial purchase, it’s up to you to measure their satisfaction through a Net Promoter Score, phone call and prepare the ground for the next purchase.
So how do you get leads?
One of the best ways to do that is by implementing a pop-up (also know as overlay) or widget lead collector on your website.
This is an example of a lead collector survey on the Marketizator website.
But in the age of information, users know their data’s worth and aren’t exactly looking forward to giving it away.
That’s where 3 very important principles come into play:
Be honest and straightforward about what your visitors might get if they leave their email address e.g : news about the latest updates, exclusive promotional offers, tutorials.
Offer discounts or vouchers in exchange for their personal details.
Not only do books and PDFs help you position yourself on the market as an authority in your domain, but they can also represent a powerful incentive for your users. Ask them for their name and email address and send them your “How To” guide, white papers or “Tips and Tricks.”
Here’s another example of how Marketizator collect leads on the blog:
This is a lead collector overlay that pops out when visitors want to exit the page.
What you should take away from this
There’s much more to CRO than A/B tests.
As a startup you should get creative and use the the little time and budget you have wisely, to ensure that when growth starts you will be able to scale.