The logic behind this was, rather than making visitors click through to a landing page in order to subscribe to our blog, we'd allow them to subscribe right there at the bottom of the post -- saving them the extra click, and reducing friction for conversion.
Makes sense, right? And it worked! Our post view-to-submission conversion rate increased 4X once when we switched over to the forms.
It got us thinking: If we can increase our subscription rate that much on the post level, what would happen if we replaced our standard lead generation CTA with a full lead generation form right there on the post, too? Could we increase lead generation from our blog by 4X? I'll admit the thought got me a little giddy.
I got even giddier when I read about how one of our partners, IMPACT Branding & Design, reported a 71% increase in leads when they used a full lead gen form at the bottom of a post compared to their standard CTA.
But before we just assumed it would work just as well for us (because we all know what happens when we assume), we decided to test it out for ourselves.
Here's what we did, how it turned out, and what you should learn from it ...
Putting It to the Test: Standard Blog CTA vs. Full Form
Because we can't A/B test a CTA vs. a form within a blog post, we tried to control our variables as scientifically as possible. Here was our methodology ...
Here's how the CTA looked at the bottom of the post ...
Blog Post B: Full Lead Gen Form
We published our second post the following Tuesday morning -- the same day of the week/time slot as Blog Post A. For this post, we styled and embedded our full lead gen form, mimicking the image and copy used in our standard CTA for the same Content Mapping Templates offer.
This post was entitled, "Why You Need to Stop Neglecting Your Old Content" -- a subject fairly similar to that of Blog Post A (so in this case, "content"), with a similar "why" angle and post structure. We tried to control other variables as well -- both posts included a secondary subscribe form as well as a slide-in subscribe form.
Here's how the full form looked at the bottom of the post ...
So ... which performed better? Below you'll find the results I recorded from HubSpot's CTA App and HubSpot's Forms App. To control for the different number of views each post received, I compared the view-to-submission rate for both posts exactly one week after each post was published.
For the standard CTA, the view-to-submission rate was 2.8%.
For the full form, the view-to-submission rate was 2.4%. I know, right? Definitely not what I expected.
Here's a summary of the results ...
I also used HubSpot's Attribution Report to determine how many new leads (compared to raw submissions) each post generated during those same one-week time frames, since new leads is one of the primary metrics our team is measured by. Even when looking at leads generated from each post, the post with the standard CTA outperformed the post with the full form.
Note: While these results weren't technically statistically significant, to us they were directional enough to decide against scaling the use of full forms on our blog. Since we would've had to style custom forms for every one of our offers, the performance of the forms would've needed to be overwhelmingly better than the performance of our standard CTAs.
So as it turns out -- contrary to my hypothesis -- our post with the regular CTA outperformed the post with the full lead gen form.For us.
What Can We Make of This?
Here are my theories ...
1) Our lead gen forms are too long for something like this to be effective for us.
To be honest, having that 14-field form at the bottom of a blog post is probably extremely overwhelming to our visitors.
My boss, Joe Chernov, worded it perfectly: "The 12 fields scares people off -- it's like sticker shock. But when someone clicks the CTA, they've emotionally committed to consuming the offer -- thus an increased willingness to fill out the form."
Considering the opposite is true of our subscribe forms, which perform better than their CTA counterpart and only consist of one field, I have a hunch that a shorter form would change the results of this test dramatically. But reducing the number of fields on our lead gen form is probably not something we'd do here at HubSpot. Our forms are long for a reason: More fields allows us to collect more information about our leads in order to better qualify them. It also weeds out crappy leads from our funnel.
2) Landing pages may be more important than we think.
I'd always assumed that people took a quick glance at the title of the offer and the image to decide whether it was worth filling out the form to download. This may prove that theory wrong. It seems people may actually need that extra validation the copy on the landing page provides before they decide to fill out our forms. A subscribe form doesn't need as much explanation for a visitor to understand what they're getting, which is another reason why subscribe forms probably perform better than lead gen forms.
What Can You Make of This?
Notice how I used the phrase "for us" about a million times in this post? Sure, full forms didn't perform better than our standard CTAs for us. But does that mean it won't work for you? Not necessarily. Just consider that example from IMPACT.
While it's interesting to read about how this didn't work for the HubSpot blog specifically, the biggest takeaway for you should be that you need to test things out for yourself. Question so-called "best practices," and get your hands dirty by testing stuff for yourself. Just because something didn't work for someone else, doesn't mean it won't work for you. Don't be afraid to apply this lesson to all areas of your marketing, either.
What do you think about our form test? What other tests have you run that generated results challenging "conventional" marketing wisdom? Share them in the comments!
Originally published Sep 2, 2014 6:00:00 AM, updated February 28 2018