Conversion rate optimization is a terribly sexy idea.
You jump into Optimizely, highlight your main features, make sure everything is above the fold, and test to see which tweak will give that little percentage a lift.
Because of your masterful split testing, your company becomes rich, your boss kisses your feet, and you ride off into the mother lovin’ sunset.
Slow down, cowboy.
There are plenty of ways to carry out your testing strategy. But before you even start thinking about which tests you’ll run, how you’ll set them up, etc, you need to have a few essentials with you: customer understanding, goal clarity, and goal values.
It’s not an overstatement to say that you should become absolutely obsessed with your customer.
It’s also not an overstatement to say that “making it work” with just demographics is lazy.
Think about your favorite uncle and your least favorite college professor or high school teacher. There’s a good chance that they’re in a lot of the same demographics, and there’s a terrible chance that they’re actually anything alike.
Demographics is like getting someone’s A/S/L (age, sex, location) on instant messenger back in the day. It gives you a vague idea of who you’re talking to, and you really have no idea what that person is like or how they’ll react to things.
If you want to run campaigns that touch a nerve and write subject lines that your readers can’t wait to open, you’re gonna need to become their best friend.
You need to really understand their personality. Who they are, what they want, and what makes them tick. You need to know if their perception filter is blocking the message you’re trying to tell them.
You might think that your audience can relate to your stock image of a smiling couple, but a lot of them are probably thinking, “That woman is too damn happy. Something’s up.”
Running usability tests right off the bat is tempting. Submit your website and have people go through your funnel while commenting on their journey. Definitely useful.
But most likely, the people giving feedback aren’t who you’re after — they’re testers.
Even if you can choose the demographics and supposed interests of your testing group, they’re not the ones interacting with you and your website because they don’t have the actual, real life need.
One of the coolest things about customer insight is that even though they’re all supporters, no two connections are the same.
Some customers have been with you for years, some a few weeks, and some maybe just a few hours. They’ve all been through different situations that led them to you. If given the chance, a few of them probably have something that they’d like to say.
Create a survey and ask the questions that make you nervous.
If it’s been awhile, do it again. There are plenty of survey tools out there. Qualtrics is robust and powerful. Survey Monkey is easy to use and has some nice features. Here’s what their builder looks like:
They’re both pretty intuitive with helpful support if you get stuck. However, my favorite’s gotta be Google Forms. It’s simple, free, and you can do some cool things with its script editor.
You can fire off customized auto-responses, or have answers automatically sent to other members of your team. If you’re like me and code like a 5-year-old, the good people at Stack Overflow or GitHub will help you out with whatever you need done. Just be nice.
With the majority of these survey tools, getting started is a breeze. Make a free account, name your survey, and you can start entering your questions right away.
Make sure that your questions are clear and straight-forward. And don’t you dare rely on just radio boxes and multiple choice.
Open-ended questions are the thoughtful, often dramatic, insights you need.
There’s even value in some of the ridiculous answers if you can try to understand the motivation and the personality behind them.
Here’s a winner that I’ve seen a few times:
But what this says is that one of your customers wasn’t even really looking for the service or info that you provide, but there was something of yours that caught their eye, resonated, and made them stick around.
A follow up question like “What content on our site are you most interested in?” can make the connection more clear.
Any answers are better than none, but to get the most possible participants, offer an incentive. Here’s a great example from Conversion XL of an email you can send to your list:
Keep it simple.
Let them know how important their feedback is and offer something in return for their time and thoughtful answers.
You don’t want the incentive to be so valuable that anyone will take the survey as a quick come-up. It should be just valuable enough to the people who really care about what you offer and what you have to say as a company.
In my experience, an ebook does the job just right (and it’s cheap/free). Just repurpose some of your more popular content into a PDF and a design a cover. Boom — you’ve got an ebook.
You’ll start to see common trends at around 100 responses, and it will start getting repetitive at 200. If you don’t have an email list or customer base that you can survey, tools like Qualaroo let you survey your current website visitors with a small fly out box like this:
Whatever tool you decide to use, create your survey with care.
Conversion XL has one of the best guides for creating killer qualitative surveys and getting useful analysis from them.
You’re going to want to set aside a good amount of time (maybe days) to get into the answers, categorize them, and create your buyer personas.
Bryan and Jeffrey Eisenberg, widely recognized as pioneers of online marketing and conversion rate optimization, created Buyer Legends (highly recommended!) as a way to understand the relationship your customers have with your brand.
Give your supposed customer a name, a job, an income, a face, and a quote that sums up his reason for caring about your company. Here’s one that Munro Shoes did:
You’ll be surprised at how much easier it is to optimize for Brandi Tyler rather than a female, aged 35-45, with an income of 35k-70k, whose interests are shoes and fashion.
Actually you won’t at all be surprised. But you’ll be happy. And you’ll get better results from your marketing efforts.
When you and your customer get to be best friends, then you can start to think about how you’re going to impress them.
Quite simply, if you can’t identify your goals, you shouldn’t even be thinking about what tests to run.
Here’s a fun exercise: Tell a coworker you have a quick question for an email you’re sending to the marketing director/department head/boss/. Reassure them that it’s simple question, you just need help with the wording.
Ask “How do we make money, and what will help us make more?” Proceed to watch them squirm.
Often times, employees know what their company does, but they don’t think about what drives more money. They don’t think about the business goals. Your conversion rate optimization strategy is about making more money, not proving your testing prowess.
You could be getting higher conversion rates at the front-end, but what if your new tests are shortening your customer lifetime value?
Make sure you know the business goals for the short, medium, and long term. Lay them out.
Got ‘em? Good. Now open up your homepage. Do your website goals reflect your business goals? Maybe not as well as you think.
Here’s a local paddle board shop.
For someone like me, who’s not very good at paddleboarding, but wants to get to a respectable skill-level, I think I’m at the right place.
It seems like they have a lot of solid information.
The most noticeable thing is definitely the logo, which is cool, but takes up half the page.
There’s an event/classes calendar on the right, but it doesn’t seem like anything’s going on. There’s a huge blank space in the prime real estate right in the middle of the page, with two random social media buttons flanking the left. Oops.
If the business goal has to do with getting lots of rentals and eco-tour appointments, they probably need to more prominently feature their phone number, not to mention their lack of a call to action.
Also, if you know that eco-tours and rentals are the moneymakers, you don’t want to crowd them among all of the other tabs. Have them stand out.
Now, I’m not a quote guy, but I was at the Conversion XL conference in Austin when I heard Oli Gardner from Unbounce say something very simple, very powerful, and very sticky:
“One page, one purpose.”
Even with your homepage, everything on a page should be centered on getting the user to take one specific action that supports your goal.
Since your customers are already used to navigating around the pain points of your website, usability tests can be a great survey follow up to get design/layout opinions.
Usability Hub is a “fast user testing” service where you can submit a snapshot of your page with your desired action.
It helps you see the pain points that you can reasonably expect from a first time visitor.
Remember that these people probably don’t know or have a need for your brand. You can get around this problem by recruiting your own customers with UserTesting, but you’ll pay a premium for this premium insight.
You can have your marketing goals, but make sure that you can directly point to the business goal that they’re trying to accomplish.
To align marketing with your business goals, go with some questions like these:
- What journey do they take to make a purchase/decision? How many steps are there?
- What type of information is used to get new leads and eventually turn those leads into awesome clients?
- Are you satisfying an impulse or are you a solution to a business need?
- Are you a pressure-free brand or do you make money from aggressive sales campaigns?
Don’t waste marketing on campaigns that don’t fit the makeup of your business goals or help your prospects take the next step in your funnel.
In the same light, don’t optimize something unless it needs to be optimized.
A common optimization whoopsie has to do with a KPI (key performance indicator, found in your website’s analytics) that almost all of us have tried to fix, the bounce rate.
Our lovably wrong common sense tells us that reducing bounce rate is the way to get people to spend more time on your site which means they’ll have more time to get seduced by your expertly crafted website and, in turn, throw their money at you.
So what do you do? You turn off the lower quality traffic sources, your bounce rate goes down, and you sit back smiling in preparation for an influx of sales.
Four weeks in with the improved bounce rate, and sales have dipped by 13 percent. Uh oh.
By shutting down the “low quality” traffic, what you just did was reduce the number of people that see your page. Regardless of these people possibly being quick to leave, you just lost a bunch of potential customers.
What if those people were just taking note of your site so they could come back later to buy?
Your goals weren’t in line.
When it’s time, remember to plan your conversion rate optimization strategy around your goals. Start by making sure that what you’re looking to optimize will not handicap any higher-level goals.
Ask yourself, “How does this affect revenue?” when deciding between tests. How much more money will your company make because the number of people clicking through to your Google+ page increased by 15%?
Yup, you guessed it. Pretty close to zero.
There’s nothing wrong with having goals like higher click-through rates and more time spent on site, but please make sure they’re not getting in the way of making more money.
Testing is always contextual, so there are no concrete examples of things that you should always test, but here’s a great article from Hubspot that tells you which KPIs you can start to look at.
So you’ve got your customer, and you have an idea about what goals are the most important. Now it’s time to monetize them.
Knowing how much your goals are worth will make prioritizing way easier. It’s not the only factor of course, but seeing the dollar impact can help you (and others) see the bigger potential of your testing.
First, make an ordered list of your goals, relative to your business priorities.
For the first go around, assuming you’ve done your homework, go through and rank them yourself. Try to keep your list to the top five or six goals, so you’re only focusing on the real revenue influencers. Your initial list could look something like this:
If you’re working with a client, or if there are any other departments in your company that have a stake in your testing program, now is the time to loop them in.
Let them know how crucial their insight is to this. Make sure that everyone is on the same page and address any differences now. It’s much easier to move around priorities now, before you create your conversion rate optimization strategy.
Next, estimate your values. Don’t worry about being completely accurate — the numbers by themselves aren’t the most important.
It’s the relative magnitude of each goal that you should be more concerned with. If you don’t know any of the exact values, just start in the middle and work outwards.
Aside from product sale, these numbers might seem pretty abstract. That’s ok.
You can definitely make use of these estimated values for your strategy, but if you have the time and resources, getting the actual values is pretty simple and (obviously) more accurate.
You can do this by getting friendly with your CRM (customer relationship management) tool. The most popular tool is Salesforce, but others like Insightly and Pardot do a great job as well, if you don’t already have one.
If there’s no CRM, get friendly with the sales team.
If there’s no sales team, just dig into accounting yourself. All you’ll be doing is dividing your revenue by the number of each goal that has converted. You can look at last year’s numbers, or do YTD (year to date).
Let’s work off my example. From January 1st to today, let’s say I made $50k in revenue.
I sold 500 products, there were 1200 free trial signups, 8,125 ebook downloads, 25,000 blog comments, and 100,000 social media engagements. With some simple math, here are my actual values:
Not all of your top goals are going to be directly tied to numbers like these, but it’s still important to understand the value.
You can dig deeper into these numbers by looking into things like lifetime value and attribution models (figuring out which part of your funnel gets credit for the conversion), but for your initial conversion rate optimization strategy, don’t worry about it yet.
These relative weights are simple but can show you a lot.
Here’s why you need to know these numbers:
Let’s say that you’re testing two different landing page formats.
You see that variant A is converting traffic at a higher rate for the landing page conversion (say, a lead magnet like an ebook download) than variant B.
However, following the funnel shows that variant B is converting free trial signups at a higher rate than variant A (which is the next step in the funnel).
If you didn’t know the relative values and how they affected your revenue, you’d immediately declare variant A the winner.
If the number of additional free trial signups makes a bigger revenue impact, you’ll choose variant B all day.
So there you have it!
By understanding your customer’s perspective and your goals, you have your foundation.
You’re ready to think about how you want to create your testing strategy to impress the daylights out of your clients, your boss, and yourself.
Remember, conversion rate optimization is a tool to make more money. Plan your approach, and the results will come.
What kind of conversion rate optimization strategy have you found to work the best?