Inbound Marketing Experiments, Of Course
Like the 80 million other folks who managed through the blizzard of 2016 this weekend, I had a lot of time on my hands. So, what did I do? I ran some inbound marketing experiments. I thought it might be fun to share a few of those with you today.
We won’t have the data on the experiments yet. I usually let them run for between one week and four weeks, depending on the data sets. But, you’ll at least get a glimpse into the mind of an inbound marketing scientist, something very few people ever have.
Here are a couple of my favorite experiments.
Landing Page Headline Swaps
I don’t know why, but as Oli Gardner said, “Most of the subheads are better than our actual headlines.” I see this all the time. I think it’s because we (marketers) start thinking cute headline and then cut to the chase with the subhead. One of the best tests I run is the one in which I switch the headlines. Make the subhead the headline, and make the old headline the new subhead. Make the site live, and let it run for as long as you need to get a decent sample size.
Don’t forget to benchmark the old page’s performance before you start messing around with it. I’ve made that mistake a couple of times, and it’s very frustrating.
Back to sample size, I don’t think you need to be a statistics professor. I’m not sure you need a 95% probability that something is statistically accurate. Just make sure you’re working with a reasonable amount of data. Two or three visits to your new page is not enough, but two or three hundred should probably be fine – at least enough for you to draw some conclusions.
CTA Button Redesigns And Placements
This is a good one. If I find a piece of content that (in my opinion) should be performing better, I’ll try a different CTA button for the same content on the same page with the same CTA placement. Remember to not change more than one variable from test to test. If the conversions increase, we have a winner. If not, we might have a content problem. You’ll learn as much or more from failed tests, so don’t be discouraged. The metrics won’t always go up.
I’ll also try moving the same CTA for the same piece of content to a new spot on the same page or to a different page. Again, sometimes people don’t connect with the content options, depending on their disposition on a particular page. Moving it does open this up.
Form Field Reductions And Additions
This is a good one, too. People love asking for a ton of information on their forms. Data and research show that, typically, three fields perform about the same as 10 or more fields, but most forms have requests for information in between three and ten. So, I’ll take an underperforming page and lower the form fields to just one or two fields. I’ll watch the conversion numbers, and if they improve, I’ll slowly start adding fields back in until the numbers plateau. Then I know exactly how many pieces of info I can ask for without impacting results.
Guest Blogging Placements
This does wonders for driving new visitors to your site and even turning them into leads, but not all guest blogging opportunities are created equal. So, you have to test these, too. For every off-site SEO opportunity like this, you need to set up a dedicated landing page with unique tracking code, and you need to keep an eye on which of these are working and which are underperforming.
Every tactic in an inbound marketing program has to be worth the effort it takes to create. So, one guest blog that produces 15 visits and one lead over the course of a month is nowhere near as valuable as another guest blog that produces 1,500 visits and 56 new leads over the course of a month. Eventually, you should stop doing the first guest blog and start doing more with that other site or finding similar sites that will allow you to guest blog. It’s the power of data that enables you to make decisions like this.
Blog Title Performance
Finally, blogging is so core to inbound that you have to run a few tests here, too. You have to look at which titles garner the most views, the most inbound links and the most social shares. When you find one or two that are far superior to the others, focus on creating more in that genre. One of our most successful blog articles in the history of Square 2 Marketing is about home pages versus landing pages. That triggered us to do more on this topic.
Look back at past performing blogs and use that data to create some new potential blogs. Run those and see if you can beat your best performing blog. Do this enough times, and you're likely to see increased views month over month, as well as increased subscribers, two very positive outcomes from a test like this.
The more you do this type of work, the better you get at creating the experiments and trying to execute ones that actually produce valuable results. Remember, not all experiments have to be successful to be positive. I’ve learned more from our failed experiments than our successful ones, hands down.
As you look at your inbound marketing program and start to think about some experiments you want to run, just remember to change only one variable at a time, benchmark your current performance first and don’t make the experiment too complicated. If you stick to these guidelines, there are thousands of tests you can run that will give you valuable insight and help you improve the performance of your inbound marketing.
Start Today Tip – Keep it simple. Just start with something easy. No matter what you’re thinking about, the simpler the test, the easier it will be to run the experiment and the easier it will be to see if it produced any interesting results. Like the headline test above, benchmark current performance, change the headline and then come back later to see if performance has increased. Just make sure you have enough data points to make the test valid. Now, you’re an inbound scientist, too.
Square 2 Marketing – Inbound Results Start With ME!