Testing 1, 2, 3: the forgotten key to your Internet marketing success

Testing is the hidden key to your Internet marketing success

I’ve worked on a lot of online marketing initiatives over the past decade or so, working with everyone from 1-person law firms, to big marketing agencies, to some of Canada’s largest corporations. And many have been very successful. But even the most successful campaigns could have been better.

Here’s where I see marketers and agencies missing a huge opportunity for better results from their Internet marketing.


I’ve seen marketers invest literally hundreds of thousands of dollars – sometimes millions of dollars – in campaign design and development and media buys, but very rarely do I see them invest in testing their online marketing campaigns to make sure they’re effective.

I’m not talking here about Quality Assurance (QA) testing, where you check to make sure links work and Web pages and emails display the way they’re supposed to – that, thankfully, most marketers are on top of.

And I’m also not talking about big budget market research and usability studies, although both of those things have their place and time.

Here’s an example of what I do mean.


Split testing for email marketing


Let’s say you’re planning on a sending an email campaign to your list of 1,000 people (you should be sending targeted messages to specific segments of that list, but that’s a blog post for a different day) and you want to maximize the number of people who open that email.

You know that the day of the week, that you send on, makes a big difference to open rates you’ve seen some generic market research that suggests that Tuesday mornings are the best time to send marketing emails and you’ve seen some contradictory research that suggests Mondays are better.

But what’s the right day for your list? Here’s how to find out.

Take 20% of your list – 200 subscribers chosen at random – and send the campaign to them on Monday morning. Take another 20% of your list at random and send the campaign to them on Tuesday morning. Then compare your open rates for the two sample groups. Send to the remaining 80% of your list on the day that performed best (you are tracking your open rates, right?).

And then keep repeating these kinds of tests across multiple campaigns. You can test other days of the week, different times of day, different subject lines… All of the different factors that influence your success.


Huge increase in Internet marketing performance


If you can constantly, incrementally squeeze an extra 2 percent, 3 percent, 5 percent improvement in performance out of your campaigns with each round of testing, by the end of a year you’re going to have a huge increase in performance compared to your original baseline.

Tip: tools like Hubspot and Mailchimp make it easy to automate your split tests for email marketing campaigns. 

You can do the same kind of thing with the Calls To Action (CTA) and landing pages on your Web site, too. Want to know which headline drives more conversions? Show one headline to half your audience and another headline to the other half. Compare results after a while, weed out the poorer performing one and test another factor.

Tip: Hubspot makes it super easy to test different landing pages and Calls to Action. But with some extra elbow grease, Google Analytics’ “Experiments” feature can be useful for this, too.


Pre-launch testing


You can also test your designs and content before putting them out there for the general public to see. The Internet has opened up a whole new set of opportunities to test your creative with focus groups for a fraction of what it once would have cost. Testing simulation tools, too, can complete computer generated tests that are remarkably accurate when compared with much more costly tests with real people.

Tip: Tools like Five Second Test, Feedback Army, EyeQuant and Loop11 can provide you with useful pre-launch insights.


Best practices for marketing optimization

It’s important to remember to only test one factor at a time (same goes for your email testing) so you know which factor is influencing the results.

But the one thing that will put you ahead of at least 80% of your competitors is to just do it. Just test. Most marketers aren’t testing and optimizing their campaigns at all. So here’s your opportunity to distance yourself from the pack.