You can learn a lot from a company’s landing pages.
It’s one of the best measurements of how well their marketing is run.
First of all: Do they have any landing pages at all? Even a few years ago, only 67 percent of medium-sized companies used any landing pages (!). And only 48 percent built a new landing page for each marketing campaign.
Even now, you only have to click a few online ads to find a company that’s still sending traffic to a generic page. You can see this all-too-common mistake on social media, in search ads, even in emails.
Many companies and marketers still don’t use landing pages anywhere near enough. And so they send their expensive traffic to a confusing page with 24 options for their visitors to pick from… and get a terrible conversion rate.
We hope you’re a long way ahead of that. But almost every company can benefit from optimizing their landing pages. Especially when you consider how much work and time and money and resources goes into driving traffic to those pages in the first place.
When you weigh all that investment against what it would take to optimize, improving your pages starts to look pretty attractive. This can be a project with an extremely high return on investment.
But to get that lofty ROI, you’ll need to understand some principles of landing page optimization upfront.
Here’s what you need to know:
1. Think hard about what you want to optimize for.
This is basically picking the right target to aim at. Without the right target, even the best planning and execution won’t result in much.
Hopefully, if you’re reading this, you already know what you want to optimize for. But just for reference, here is what your fellow marketers are hoping to get from optimizing their pages:
No big surprise here: Most people want to increase conversion rates. And that’s smart. But flip that over for a second: One in three marketers aren’t optimizing their landing pages for more conversions.
Perhaps instead they want to increase visitor traffic (half of all marketers do). Or maybe they just want leads that convert into customers more quickly, thus shortening the sales cycle. Or they want less expensive sales-qualified leads (SQL).
Whichever KPI they’re after, it should be tied to business goals. Otherwise you’ll end up with beautifully optimized pages that don’t get you any meaningful results.
2. Optimizing your landing pages may reduce sales at first.
This principle is particularly slanted towards split-testing, but it can apply to any optimization change.
Not everything you do is going to work.
Not every change improves results. In fact, when split-testing pages, most of what you test will not perform as well as your control.
This rarely gets talked about, but it’s the first thing you’ll run into if you start a comprehensive rehaul of your landing pages.
Testing requires some backbone in this regard. When I had one of the biggest wins of my career with landing page testing, we were actually on the fifth test of a particular page. I had to plead with the owner just to get them to test the page in the first place. They were sending $12,000 a day of AdWords traffic to this page… and they had never ever tested it.
By test five, I was starting to get not-so-subtle messages from the owner that he was tired of my fancy ideas about landing pages.
But I persisted. I bet on one last test. A super-simple button copy test. A test of changing three words on the button – the call to action.
By the end of the next week, one version of the new button copy had doubled the conversion rate. One little test – set up in two hours, changing just three words – saved the client over $1 million in ad spend per year.
But we lost about $25,000 worth of sales in the process of finding those magic three words.
So know this: Landing pages optimization is going to probably cost you some business before you find even a 10-30 percent lift. But if you can keep the powers-that-be at bay, and hold steady, and get your statistically valid results, you can absolutely increase conversion rates by 30-50 percent over time (like over 3-6 months).
You can only do that if your boss doesn’t freak out because they got 20 percent fewer sales last week because you ran a test that didn’t beat the control.
There is one way to address this problem: If your landing page has enough traffic, you don’t have to send 50 percent of your traffic to version A and version B. A sophisticated split-test tool will let you direct, say, 10 percent of your traffic to your test page. This lets you test, but minimizes the damage if the new page is a dog. But it may also extend how long the test has to run for.
If you have enough traffic to do multivariate testing (where you test multiple variables – the headline, the CTA, the image, the bullets, the offer all at once), you can often find that magic variable faster. But still be ready for a dip in conversions until the test is done.
This is optimization, after all. Not magic.
3. The early results of a test do not equal the final results of a test.
Landing page optimization tests are kind of like horse races.
Early on, you may see one version look really hot. Your boss may be thrilled. And that same variation may be going strong when you check in the next day. Your boss, always hoping to move things along, may start bugging you about ending the test early. Especially if you hit day 3 and still have an outsized winner.
Don’t do it.
Don’t give in to testing impatience. It’s all too common for a particular variation to look really good early on. But over time, they often fall away.
This is particularly disappointing when one variation looks like an early, strong winner. It is really tempting to just call it a winner, and then run through the office proclaiming the good news.
Again, don’t do it. Because every test – every single test – must run at least one week. No matter how awesome one version appears to be performing early on.
You may have to fight to keep that test running. So be it. Come to think of it, most conversion rate optimizers actually are kinda scrappy. They need that backbone to make sure all their tests are statistically valid.
Ending a test early means you’re not a doing statistically valid test. That means you may be picking what looks like a winner, but isn’t actually a winner.
And that means you could be picking a version that will actually reduce your conversion rate.
Bonus tip: Don’t run tests in weird weeks, either. Running a test during a weird week will throw off your results. “Weird weeks” are any major holiday, or during some massive, remember-for-a-generation type of news event.
4. Keep your messaging consistent.
Do your landing pages match the messages from what people saw just before? In other words, do they see the same phrase on a landing page that they saw in the ad? Do they see the same message in the landing page as they saw in the email?
It’s critical to keep your messaging consistent through those steps. Otherwise your visitors can feel disoriented. Or worse – they’ll think they’re on a page that doesn’t match what they want. And so, they’ll leave.
Here’s an example of an AdWords ad and a landing page that doesn’t use the same core phrase. When someone is zooming through pages, scanning to find exactly what they want (and filter out the rest), the inconsistent messaging here could make them think this landing page isn’t what they want.
And here’s the landing page for the bottom ad:
Another spin on this principle is to test different ads or emails before people get to the landing page. A different email may change the conversion rate of your landing page. A different ad might, too.
5. Think about the end of your sales funnel – not just the beginning.
“What? Aren’t we doing this to get more conversions?”
Sure, you can just aim for more conversions. But it might not work out the way you want it to.
As you probably know already (especially if you’re in sales), not all leads are created equal. It is possible to double the conversions for a landing page, but to end up with less business.
We talked about optimizing for KPIs besides conversions at the beginning, but let’s take that a step further here: Make sure that any landing page optimizations you do aren’t hurting more durable, meaningful metrics like lifetime value.
Sometimes, the page that gets barely a 5 percent conversion rate ends up being the golden gateway to your highest value clients. You wouldn’t want to “fix” the landing page to get 30 percent conversions up front, only to end up with low-value, one-time clients, right?
6. You can increase conversions without ever touching the landing page.
Case in point: A landing page for an old client of mine – a family practice law firm – was getting a 5 percent conversion rate. The traffic coming to the landing page was from AdWords ads, including the broad match keyword, “New Mexico adoption”.
After some poking around in the AdWords account, I discovered that the person who set up the AdWords account never added any negative keywords. So the ad – and the landing page – was getting clicks from people who had been searching for phrases like “New Mexico dog adoption”.
After the negative keyword problem was fixed, the conversion rate for the landing page tripled.
Lesson learned: Traffic matters for landing pages. A LOT. In fact, one of the favorite tests of CRO professionals is to send two different traffic sources to two identical landing pages.
For example, page A gets social media traffic. Page B gets AdWords traffic. Or one of my favorites: Page A gets mobile traffic. Page B gets desktop traffic.
It can be an educational exercise. A profitable one, too.
7. Don’t test more than one thing at a time.
This is one of the most frustrating parts of optimization, but it’s a rule you need to follow: Don’t run two tests at the same time.
Know why? Because if you’ve got two tests going at once, and something does change in your results, how will you know which test caused the change?
Honestly, this is a rule that people break routinely. And it mucks up all sorts of testing and optimization work. Almost no marketer I’ve met is willing to test only their landing pages at one time. That would mean no testing of ad copy while that landing page test is running. And no testing of emails. And no testing of lead nurturing follow ups, and no… you get the idea.
So don’t expect to be perfect on this point. But at least try to minimize the damage.
8. Speed converts.
Want to do just one thing to your landing page that almost guarantees improvement?
Cut the load time in half.
Seriously. People bail on content that takes too long to load.
One way to do this? Use AMP for your landing pages (and everywhere else, too). See more about how to do this here.
Oh yeah: And just having your pages zip to load isn’t enough. Those forms have to work, too. Or you’ll lose just as many visitors as you would if the page was slow.
9. Most tests don’t blow the doors off.
If you read marketing sites a lot, you might be forgiven if you think that most landing page tests result in massive improvements. Like going from a 2 percent to a 27 percent conversion rate, for example.
Unfortunately, that’s just not true. But if you start a brand new landing page optimization program and don’t get a couple of big wins right away, some people may start to question your work. They’ll wonder if testing and optimization just doesn’t work… or if you don’t know how to do it correctly.
Just keep calm and test on. If you follow good optimization practices, your rewards will come. Expect at least a 20 percent improvement in conversion rates within 3 months.
10. Think hard about your visitor intent – and about your value proposition.
Often, copywriters are the people tasked with defining a value proposition. But it is arguably a better job for the marketer. Value propositions are a mindset – they’re what your company offers people. They are the benefits of what you offer, not the features.
You need to align what your customer wants from their click with what you are offering on the page. If you can get those two things to intersect – the visitor’s intent and what you’re offering – your conversion rates will soar.
Just witness the 66 percent increase in event registrations that The Global Leadership Summit got by applying this tactic.
They did it simply by zeroing in on the value proposition of their landing page, and changing it from what the marketers thought it was to what some smart market research told them it actually was.
Whew – that’s a lot to know about landing pages! But don’t let all the information get you down.
Instead, 80/20 it.
In other words, look for optimization tasks that are likely to get you the highest return for the least amount of work.
For some ideas about how to achieve that, consider how other marketers ranked landing page optimization tactics according to how effective they were and how easy they were to execute.
Let that chart inform your decision, but don’t just blindly use a tactic because other marketers say they’re using it. Best practices are generalizations. They’re helpful, but they may not necessarily apply to your situation.
The best area to focus on will depend on where your landing pages are. Maybe you’ve already made sure every landing page you’ve got is mobile-friendly and easy to use. Maybe you know every last page loads in a blazing two seconds or less. Then it might be time for some split-testing.
Or maybe you’re like the one out of three marketers who doesn’t necessarily want more conversions. You want better conversions – leads that close faster and bring in more business.
So once again, before you get too specific about your marketing to-do list, remember your business goals. Every piece of content you have – including landing pages – should be designed and managed to optimize those priorities.
Otherwise, you run the most dreaded risk of any conversion rate optimizer… optimizing for the wrong things.
Ready to get started, but not sure what to test? See Derek Edmond’s post, 11 Key Factors In Building Landing Pages for B2B Search Engine Marketing Initiatives. It walks through the basic elements that every landing page should have and how those elements should be presented.