In order to make sure your lead generation strategy is the best it can be you need to be constantly testing and optimizing your campaigns. Dayna will talk you through the different things like programs, vendors, design and emails you can test.
- [Narrator] To get the most bang for your buck with your lead generation campaigns, you need to be constantly testing and optimizing all of your different programs. In order to make your lead generation campaigns better over time, testing is an ideal way to know what performs well and what doesn't perform quite as well. So what can you test when it comes down to your lead generation strategy? Of course, you can test your programs, so determining what works best when you're thinking through different channels, different events, different offers.
Vendors are a fantastic thing that you should always be testing, especially if you're using these paid vendors. You will tend to get different performance from vendor to vendor, so make sure that you're keeping track of that. Email subject lines are a great thing to test, as well, if email is a large part of your lead generation strategy, as it should be. Whether they are paid or database emails, you do always want to be testing subject lines. Email copy is something that you can absolutely test. Calls to action, so not only the look and feel of the CTA, and the language, but also actually the offer itself and the placement.
So this is a great way to continuously increase conversion. Design, design is a fantastic thing to test, not only on your website, but also within your other programs. So think about social media and the different types of imagery you might use there, these are very easy ways to test design throughout these various programs. So your website and website navigation. So with your website you want to make sure that you're always testing different conversion paths, you're always testing different copy, and that you're also testing different imagery on your website, itself, in order to not only increase conversion, but also increase engagement and time on your site.
You can test form lengths, so for your various landing pages what is the ideal amount of different fields that you want people to fill out. You can certainly run a test to see how many fields that somebody is willing to fill out. Social media messaging, so whether it's Twitter, Facebook, LinkedIn, be sure that you're testing out different copy, different tone, is it funny, is it serious. For content, different content topics will perform better. And also think through testing content in the various channels as you're considering your offer.
So then you want to also test the times of days that you are posting things. So email, social, as well as your blog, will generally have some best practices as far as what days that you send or post these things and the times at which you post or send them. Definitely be sure to test some different times out, and you might find that you're surprised by some of the results. But first, let's start with A/B testing. These types of tests are often called split tests, and what they do is they compare two elements to determine which element has a better conversion rate.
And, of course, this is easy to execute since you are only looking at two main elements. So, typically, a simple 50/50 send or a traffic split is how you can determine success. So if you're sending an email you could send 50% to one email, 50% to another. If it is kind of a more out of box type of subject line or test, what you can do is you can actually partition off a certain amount of your database and send 50/50 to maybe 10% of your database to try and see if you see any additional results.
Same with traffic splits, you can set things up that the test is running to a smaller amount of traffic, but either way, again, A/B testing, simple split tests, you are testing just two elements. So multivariate testing typically shows more complex information, however, you should use a software solution to build your multivariate tests. There are a few tools out there like Optimize Lead for the website that can help to this. You can also do some multivariate testing if you have a marketing automation platform like Marketo or Eloqua, those tools do allow you to use multiple elements.
One of the downsides of doing multivariate testing is it can be a little bit more complicated to figure out what is causing the change, whereas an A/B test is fairly simple and you know it's one over the other. So how do you set up a test, how do you begin thinking about testing in your lead generation programs? So, number one, is you want to formulate your question. What are you testing and why are you testing it? There's no point in testing just for testing sake, you do want to make sure that you are testing things that move the needle, and that might be by looking at different programs that aren't particularly performing well and then coming up with a second version to test.
So how do you define success? That's definitely the next step that you need to think through. So are you defining success by increased downloads, increased conversion rate, more MQLs, more time on site? So, for instance, if you are doing a website test you might be testing something like engagement which would be time on site. So if you're switching around different images, you want to see how people are spending time on your page, and particularly maybe if it's a product page, you can test stuff like that, you just need to make sure that you have a solid understanding of how you define success based on your particular test.
And you also need to make sure that you construct your hypothesis. So based on the research and previous experience, what do you think will happen? Generally, when I set up a hypothesis, I do have an idea of one way or another what I think will happen, if it's two subject lines, if it's two images, I do have a best guess, and that's what you want to have, you want to be able to have a stake in the ground one way or another. So how do you conduct your test? So once you kind of have set your test up, the next step is actually conducting your test, and you might be doing this through a marketing automation platform, you might be doing this off of a website platform or a testing platform, something like Optimize Lead.
So the number one thing is to isolate your variable, so you need to determine what you are testing and you will need a control and a test. So, generally, if you're testing something there will be a version of it already created, so that's your control. So if you have one email that you have sent out multiple times without the type of results that you want, that would be your control group, and then you have your test, the new version. You want to use a large enough sample size. Your sample size can't be too small, but also be careful that it's not too big.
So this is very much dependent on your database size, on the traffic that you're generating to your website. You'll have a good kind of gut feel on how you want to test things, you might want to start smaller and then go bigger. So you want to eliminate confounding variables, if possible. If you are doing an A/B test, you want to keep everything the same except for your testing variable so that you can ensure that the process is consistent across both elements. So what do I mean here? So, in addition to just changing something like an email banner and that's it, you also want to make sure that you're sending these emails to the same amount of people, that you are sending the both emails at the same time, on the same day, to the same audience.
So, everything that you can keep consistent will help really focus your test. So then, of course, look at the results. You need to determine if your test has reached statistical significance, which means the probability that the results are meaningful and not by chance. So this generally means, especially if it's a website test, that these tests have to run a little bit longer in order for it to reach some sort of statistical significance. You might be looking at a website test and seeing things change very, very slowly or only a little bit, and that means that you have to run the test longer.
Most of these testing platforms will let you know when you've reached significance. The other example is if you are sending out something like an email blast and you are noticing that the differences between your two variables are very, very minimal, you might not have reached significance at that point, and that you might have to run that test again, or tweak things even further. All in all, the more testing that you do, the more optimized your lead generation campaigns can be. There's such a wide variety of things that you can test across all of your different programs.
I definitely recommend that you do take time to do some of these tests so that you can optimize things going forward. Again, the more optimized your lead generation programs are, the more successful that they will ultimately be.
- Identifying funnel stages
- Defining key goals and metrics
- Deciding on an approach
- Building a lead generation plan and a team
- Aligning with stakeholders
- Choosing a lead generation technology
- Building attractive content
- Generating leads with blogs and social media
- Creating an SEO strategy
- Tracking ad performance
- Increasing visibility through events
- Using paid ads and direct mail
- Qualifying leads
- Measuring campaign effectiveness