Most of us do not have enough traffic to A/B test every color and font on our lead magnets. Instead, we’re forced to make do with a trickle of traffic, where we would like a torrent.
Heck, the very reason we’re optimizing is to turn that trickle into a torrent. If we had a torrent, then we wouldn’t be so desperate to optimize, would we?
To conduct an A/B test, you need traffic. But to attract traffic, you need to conduct a number of A/B tests. It feels like a classic catch-22.
Fortunately, there is hope. There are ways that you can use the little traffic that you actually have to optimize your landing page(s).
Today we’re going to explore the what, where and how of it.
Ready? Then let’s jump in!
What does the data say?
The first step is to simply look at the data you are already collecting through such tools as Google Analytics and programs like it. Sometimes you can learn a great deal before you start A/B testing different designs.
For example, you might find that there is suddenly a big drop off from one step to the next. This alone shows you that something is wrong. So discuss it with your marketing team to brainstorm solutions.
Note: your team may not agree on what the problem is, let alone how to solve it.
If you’ve got several dissenting opinions, consider testing a few different strategies and seeing which works best. Just make sure you give each test enough time to run before starting on a new strategy.
What are your visitors doing?
Alternatively, use software like Inspectlet, which actually lets you see what the visitors are doing on the page. Sure, they won’t exactly be telling you why they don’t fill in a field or click away when they read something, but with a bit of logical thinking, you should be able to figure out what it is that’s turning them off.
- Are you asking the right questions?
- Is there a hidden cost that is suddenly revealed?
- Does something inspire distrust?
- Is a new condition revealed that they didn’t know about before?
- Is the form far more complicated than was originally implied?
These are all possible reasons that you might be losing customers. Try tackling them and changing them and then seeing if your conversion rates improve. Transparency and directness will also help you to improve customer loyalty, resulting in more conversions down the line.
What have others said?
Another strategy is to simply read what other people have written about what you should expect at certain steps in your funnel process. Examine your website to see whether you are meeting certain goals, whether this means conversion rates or average sale.
If you find that your numbers diverge widely from what other people have found at that step, with you converting far fewer people than expected, focus your attention on that particular funnel “leak.”
You won’t be able to test what will work better easily, but here too you can get a lot of help from what others have done in the past. Simply consider the different choices that other people have made and then try applying those results to your own site. Then see what that does to your numbers. And after that, you can look at improving another aspect of your sales and conversion funnels in the same way.
Naturally, this is not a very scientific way of doing this, but as long as you look at percentages rather than looking at numbers, you should still be able to conclude if your changes are having a positive effect.
Can’t do Quantitative research? Do qualitative instead
Yes, it’s all about big data nowadays. That doesn’t mean, however, that there are no other ways to conduct research. And so, if you can’t do quantitative research because you don’t have the numbers, do qualitative research instead.
What do I mean by that?
Well, just get a few people in front of your platform, have them move through the process and ask them what’s going on.
You can do this in quite a few ways. You can either get people to come in and talk to you as they navigate your site. This will give you an insight into the different steps they take, particularly if you can give them to give a stream of consciousness account, where they comment and critique your text, your interface, and your visual content.
Alternatively (and far more cheaply) consider asking them in questionnaires that you send their way when they’re on your site. You can ask questions of people that do actually make purchases, asking them about any of the procedure that they didn’t like and where they were not particularly impressed.
Of course, the problem with that is that these people actually did buy, and so they’re not exactly the people who you want to ask questions. Instead, if people do click away, consider using a simple pop-up window asking them why they’re leaving. Most people will still click this window away, but the few answers you do get might be very enlightening.
But wait, you say, won’t that annoy the visitors? Yes, it very well might. The thing is, the fact that they’re clicking away probably means that they aren’t going to buy your product anyway. For that reason, asking they why they’re leaving – while it hurts you a little in the short run – might well pay long-run dividends that outweigh these costs.
Want to bet on it?
If you don’t have the visitor population to test every small little thing, instead try using it to test something big. Something risky.
Big risks can result in big rewards . . . or big failures. And so that’s where you want to concentrate your A/B testing instead. In this way, you’ll know if it’s going to work and can easily switch back if it doesn’t.
Feel uncomfortable gambling?
I understand. It is your business you’re putting on the line. The gradual growth idea would seem to work best. The thing is, that the gradual growth idea is also what the big boys are using. But because they’ve got so much more visitors and so many more clients, when they grow by 2% and you grow by 3% they’re still galloping ahead of you in raw numbers. 2% of a million is a hell of a lot more than 3% of a thousand.
And beyond that, you are more likely to see dramatically different testing results if you make major changes to your landing pages. Yes, you’ll see a slight differentiation by testing different lead magnet sizes and different photos. But a big change will tell you definitively which design you customers prefer.
In today’s marketplace, which is saturated with big fish, sometimes you need to gamble and try and win big, in order to take market share before the big boys have you on their radar. And that’s where you want to direct your A/B testing.
In many ways, A/B testing is a brute force approach to finding out what you should do with your page. And that’s great if you have a huge number of visitors coming through that you can apply this tactic to. But a lot of sites don’t have that. In that case, it’s a matter of being smart, a matter of using other methods and using the research that other people have already done.
So don’t A/B test what color your buttons should be. Instead, simply use the research that is already out there, or ask a bunch of customers to tell you what they think, or go with your gut.
Instead, focus the few A/B tests that you can do on places and things where you can really make differences – places where you can show the soul and character of a brand, places where you can really change the name of the game.
In other words, if you don’t have the audience, don’t look for optimization, but look for revolution.
How have you revolutionized your sales funnel? Tell us in the comments!