One of the greatest challenges facing nearly every nascent or small organization is how to drive visitors and leads to its website. However in most cases, driving all of the people in the world to a website won’t mean a thing, if those visitors do not convert as the website would like. SEO, one of the core offerings of Distilled, is certainly an option for some websites, but as a matter of prioritization, sometimes it should not be the primary focus as results are often in a very short space of time.
So what can be done to drive visitors when time isn’t your friend and your dreams of ranking number one in the search engines look bleak? There are lots of different options out there including PPC ads from Google and Bing, or advertising on Twitter, LinkedIn, and Facebook, among others, but this article will focus on Facebook as an acquisition channel. But once you get people to click on your Facebook ad, where do they go? What action is supposed to be taken? How do we measure (and learn) from what’s going on? How can we re-engage with people once we’ve paid for their visit? All of these are the right questions to be asking and will be addressed in this blog post.
Disclaimer: I am not being compensated in any way for highlighting services used in the post they just happen to be services that work well for accomplishing what we’re looking to do.
Why did I choose Facebook to be the acquisition channel highlighted in this post? Facebook is one of the largest channels and is one of the most cost effective you can possibly find, but apart from these reasons, it is the ability to target users at such a granular level that really sets it apart from the competition. By compiling data from its 1.591 BILLION monthly active users, its acquisitions of companies like WhatsApp and Instagram, along with partnerships with data mining companies like Acxiom, Epsilon, and Datalogix, it’s highly likely that Facebook knows a good thing or two about you.
The simplicity and ease of its set up is both a blessing and a curse, as it allows people who don’t know what they’re doing to waste ad dollars on targeting the wrong people or not crafting a good ad, but allows those who know how to use it the great advantage of highly targeted ads for cheap.
This post won’t go over setting up your ad account (more on that here), but we’ll begin as if you already have a payment account linked. Firstly, you’ll need to make sure you have a Facebook page for your business or organization and once you have that, you can click “Create Campaign”. The other data seen in the screenshot below is from a separate campaign I have running at the moment.
Doing this will bring you to a page presenting you with numerous objectives for your campaign. Each one certainly has its own use case, but a very useful one is “Send people to your website”, as we’ll have complete and total control over what information we present and ask for.
Next, input the desired landing page URL for your ads and we can get to the fun part (you’ll also be prompted to select a tracking pixel, but you likely won’t have this yet, so skip this for now). At this stage we’ll be able to leverage what makes Facebook ads so special, the granularity of targeting. Depending on what your ideal audience is, you’ll need to specify things like location, age, gender and language, in addition to the “detailed targeting”, which is really powerful.
Let’s assume we work for Vail and want to sell Vail-related clothing and/or skiing equipment to people. Of the many things we can do, we can target people who have recently been to Vail (by changing the button that is by default set to “Everyone in this location” to “People recently in this location” and selecting the location as Vail, Colorado. We can change the age, gender, and language requirements. however. we’d like and then get to the “Detailed Targeting” where we can focus on behaviors like Purchase behavior > Sports and outdoors > Winter sports. You can then click “Narrow audience” just below this for the ability to target people who fall into the first category AND this second one.
A common mistake would be to target people by throwing many different behaviors/interest/demographics all into that first pool because Facebook will include all of the people who fit at least one of those. By narrowing it further, you can elect to tell Facebook you want people who have multiple interests in common, which makes your audience smaller and more qualified. For example’s sake, we can now include people who have an interest in skiing OR snowboarding.
You can slice and dice this any way you want, the options are so numerous and it just takes some exploring to figure out exactly the type of people you’re looking to target. To get a few more ideas, here’s a list of some of the potential ways one could target an audience.
It’s not strictly necessary, but you could also add a connection type by (Facebook pages, apps, and events) either with people who like any of them, friends of people who like them, or excluding people who like them.
Set the budget you feel most comfortable with, though I’d recommend going as big as you can stomach at the beginning when your testing is most important and then dialing it back as you learn what works and what doesn’t. Keeping the bid amount as “automatic” over “manual” allows Facebook to set the appropriate bid for your ads and they normally do a great job at this so long as your ad copy is relevant to your landing page and people respond well to your ad (more on this later).
When you’re done with this, move on to “Choose ad creative” and you’ll find yourself on the next page. You can choose either one or multiple images (I like to use one) and then you can get into the copy section of the ad. Facebook has a rule that an image can’t have more than 20% text across it, so you’ll need to select one that abides by this. You can upload your own images if you’d like, but you’ll also have access to free stock images from their library. This aspect is vital to a good campaign because if the quality of your headline, text, and expanded description is low, you’ll receive negative feedback, which will drive up the cost of your clicks by lowering your relevance score and hide your ads from being seen. Conversely, if your ad receives positive feedback such as likes, shares, and comments, your relevance score will increase and your cost per click will go down.
Important things to keep in mind here:
The headline, text, and news feed link description all have character limits that you’ll need to fit within otherwise the ads will be cut off and it won’t help drive visits to your landing page.
You’ll have the ability to select where you want your ads to be shown (bottom right quadrant) and this is important because depending on who your audience is, they may not be active on mobile or Instagram, so displaying ads there can be a waste of budget.
Adespresso ran a fantastic test on how relevance score affects costs and was able to produce this graph. Notice how as relevance score improves (10 being the best and 1 the worst), the cost per click dramatically decreases.
Once you place your order, your ads need to be reviewed by Facebook to ensure it meets their guidelines before it can be served to the desired audience.
If you use Google Analytics to track and measurements, there is often misattribution when using Facebook ads. To make sure that the integrity of tracking is high, use the Google Analytics Facebook URL builder to add the proper tracking parameters to your landing page URL. Once you have your ad set(s) set up, you can go into the “ad settings” within “edit ad” to change the URL of the one specified.
Scrolling to the bottom of this page (ad settings), you can also enable conversion tracking to measure how people perform from a given ad or ad set (there is, at least, one ad within an ad set). You’ll likely need to create a pixel, follow the instructions and add it to the <head> tags of your desired pages. Once the pixel is created and placed on the desired page, you can test if it is set up properly with the Facebook Pixel Helper. Enable it in ad settings and then you’re on your way.
One of the advantages of digital marketing (Facebook included) is the ability to monitor and test hypotheses with data. If you buy an ad on a bench, you’ll have no idea how people react to it or if they prefer a different image or copy.
By using multiple ad sets, you can test different audiences or messages and see which performs best. Using our Vail example from before, we could in theory test people who have an interest in skiing versus those who have an interest in snowboarding to see who will perform better. This test would require creating another ad set, removing snowboarding from the original so both are isolated. Ideally, the ad copy and image(s) are the same for both ad groups because if there are multiple variables, we won’t know exactly which ad set performs better. Within an individual ad set, you can test different ads (by image or ad copy) to see how people react to each ad.
Where do visitors go? (The landing page)
For those with less than professional coding and design skills (myself surely included), Unbounce is here to save the day. Unbounce is a fantastic landing page design site that will handle all of your landing page hosting and allows for great A/B testing as well as 3rd party tool integration all for an affordable monthly rate. That all sounds like a lot, so let’s break that down a bit.
Unbounce has loads and loads of templates from which you can choose for professionally designed landing pages. Once you select a template, you can easily edit it with no programming knowledge required. You can easily create, modify and preview pages for both desktop and mobile so your landing page always looks great regardless of a visitor’s device.
Aside from a pure page development and deployment perspective, Unbounce also works great for A/B testing. A/B testing is the practice of comparing two slightly differing features (in this case, web pages) to establish which one works better for a particular audience. It allows us to empirically test and let the data decide as opposed to having a single person guess and hope the version he or she prefers works best. It’s a very important part of the process of optimization and ultimately allows us to improve our conversion rate.
If one were to code a page themselves or use a design team, you may be able to have a landing page, but it would be much more difficult to accurately track the results of testing and getting changes put into the dev queue for testing often can take a long time to be implemented. Unbounce makes for producing changes and testing ideas easily and quickly.
Though this isn’t science class, you will be conducting experiments – there will be a control and variants. When you test different ideas on your landing pages, you’ll have “champion” page and at least one variant. The variant(s) should be near identical except for 1 change (i.e. different copy or images) and Unbounce allows you to drive a percentage of traffic to whichever page you like. The screenshot below shows the A/B test center and a few of the variants I was working with.
3rd party integrations
Let’s say that a visitor decided to click on your beautiful Facebook ad, arrived on your optimized landing page, and decided to convert, providing their name and email. How exactly do you capture that information, what do you do with it and when do you act on it? Ideally, you’d want to respond as fast as possible to someone who just provided their information, but what if you’re unable to check your email or respond at a given moment?
Unbounce allows for integrations with other companies so that every bit of information you capture is handled automatically and fed to other software you may be using, like MailChimp, which is an email service provider.
This ability to integrate with email marketing software brings us to:
Email marketing for re-engagement
Wouldn’t it be great if every single time a lead came in from your landing page, the visitor who provided their information automatically received an email thanking them and welcoming them to your organization? This is precisely what MailChimp (and others) are capable of doing, so when hundreds or thousands of leads start pouring in per week, ensuring that the proper emails are being sent at the appropriate time is key; it would be nearly impossible to respond to each email manually.
Unlike Unbounce, MailChimp currently offers a “freemium” version of the services, so depending on the number of emails being sent per month, you can actually use their product for free, with some caveats. One of those caveats is that you won’t have the ability to send automated emails (emails automatically generated and sent to those who come in through services like Unbounce) if you are on the free service of their site. This fact isn’t always very obvious and the lack of customer support doesn’t help this clarification, but I hope this article does.
Once you sign up for their services (there are several tiered levels depending on your needs), you can set up automated emails. From here, you can choose a template and modify it to include custom copy, images, videos, buttons and more to best provide value to your target audience. The ability to automatically send emails is nice, but MailChimp really provides value when you use their analytics and reporting for A/B testing, just like your landing page. In this case, however, you can test things such as subject line, sender, and the actual copy or design of your email. I recommend that you only test one thing at a time such as different subject lines, but same email copy, or vice versa, because if there are too many variable changes, it will be impossible to understand which version is truly better.
Another really cool feature of MailChimp is the ability to send follow-up emails. We’ve already covered how you can automatically send an email when someone converts on your landing page, but what I haven’t mentioned yet is how you can actually add subsequent emails to your “drip campaign” to increase engagement and reminder your subscribers of your service. An example of this would be to send a follow-up email to people who haven’t responded two days after having received the first email. But you can get even more specific than this. You can send emails to people who have opened you previous ones, clicked on links and more. Now it’s possible to engage with specific people in a way that best meets their expectations and can guide them further down the funnel.
It’s hard to overstate the importance of the information that can be gleaned from your reporting that you have in MailChimp. You can see which people most often interact (or not) with your emails, where they are, and as you collect more information, it makes creating new hypotheses and testing all that much easier.
As I mentioned at the beginning of this post, I’m not being compensated to say anything about the services mentioned in here. The reason I do is only because I have first-hand experience with them and trust them to get the job done. It’s important to remember that when you’re thinking of user acquisition, you’re also thinking of their lifetime journey and the experience that they’ll have from the moment that your ad is first spotted to the moment when they convert.
You do not need to be particularly technical or have much knowledge of programming to successfully execute these steps, but using Facebook, Unbounce, and MailChimp will allow you to control nearly all of the user journey and ensure that the path is seamless and professional-looking. Do not be afraid to test, in fact, one of the greatest assets these tools have is the ability to put you in the driver’s seat to examine hypotheses, fail often, and look like a stud from your successes. Have the confidence to speak to developers, superiors or anyone else and say after iterative testing, you were able to optimize your funnel to attract more visits and create more qualified conversions than before.
Happy testing and please share below any other tools or advice you use to optimize your funnel from acquisition to conversion. For more information on the software mentioned and this topic in general, please visit:
via Distilled http://ift.tt/1ZkjPzt