Sign up for weekly recaps of the ever-changing search marketing landscape.
A Completely Different Kind Of Landing Page Optimization
What exactly is landing page optimization? For most search marketers, optimizing landing pages means A/B testing and multivariate testing (MVT). It means using tools such as Google Website Optimizer to experiment with different arrangements of a page – or different variations in the content of a page – to maximize its conversion rate.
We may ask questions about many different elements of a page. Which headline works best? Is the offer button better with “complimentary shipping” or “free shipping”? Does a picture of the product compel more people than an image of a friendly customer service representative, or the other way around? What about a blue color scheme versus a green one?
This is landing page optimization as most people know it. And for what it does, it is a good approach. Let’s call it content optimization – finding the best presentation of content on a particular landing page.
But content optimization is not the only way to optimize landing pages. In fact, when it comes to boosting your conversion rate, it may not even be the most effective method.
Segment optimization offers a new approach
Segment optimization is about determining how many different landing pages are optimal for a given campaign, and determining how each should be different from the other.
Instead of stretching one page to try to please everyone, which is quite hard to do – segment optimization breaks out several specialized landing pages that each focus on pleasing a particular segment of your audience.
For example, say you’re marketing language learning software. Although everyone who clicks on your ads wants to learn a language, there are different motivations among them. Students hope for better grades in their classes. Vacationers crave more authentic trips. Business travelers are most concerned about efficiency.
It would be difficult to have one page that speaks passionately to all of those distinct audience segments simultaneously.
Just consider the headline. If you were deploying one page to fit everyone, you might try lots of variations (content optimization) to discover that “Learning French is easy!” is the best headline (on average) for all respondents. Let’s say it achieves a 5% conversion rate, not bad, but not earth-shattering either.
But if you had three different landing pages, one for each of these segments – you might find that “Ace your French exams!” works best for students, “Experience France as only a French speaker can!” works best for vacationers, and “Business French in 10 minutes per day!” works best for business people. These might achieve conversion rates of 12%, 11%, and 14% respectively for each segment, a tremendous success that more than doubles your overall conversion rate.
You could never achieve this using one page for everyone. “Ace your French exams!” would perform disastrously for vacationers and business people. If you tried that headline for all respondents, content optimization would throw it out as suboptimal because 2/3 of the audience would think it was awful. But when it’s presented to students (and only students) – it is the indisputable champion.
In that example, segment optimization was achieved by deploying three different pages instead of just one. Content optimization was then used to determine which headline was most effective for each separate segment.
How can you begin using segment optimization in your campaigns?
Start by making a list of possible segments within your audience. Who are the different types of people who look for you online – and why? Don’t restrict yourself to the way you may have segmented people in your database or your business plan. Brainstorm what’s important and relevant from the respondent’s point-of-view, by considering any or all of the following issues:
- the specific “problem” the respondent wants to solve
- the demographic/psychographic “persona” of the respondent
- the respondent’s stage in the buying process
- the role of the respondent in their organization
- the respondent’s geographic location
- the respondent’s industry or the size of their organization
These are your initial buckets into which respondents could be segmented. Don’t worry if there’s overlap between buckets, as these won’t necessarily be either/or choices.
Next, review the keywords and ad creatives you’re running in your search marketing campaigns. For each keyword/creative pair, ask yourself – is there a particular segment that its respondents would clearly belong to? If the answer is yes, add it to that bucket along with the number of clicks per month it generates. If there answer is no, leave a question mark next to it – perhaps with a handful of segments it might appeal to.
For instance, in our example above, the keyword phrases “french exam” and “college french” are obvious candidates for the student segment. Phrases like “business french” and “executive french” fall into the business traveler bucket. But “learn french” can’t be segmented just from the keyword.
Now, look over your segment buckets and see which ones have the most number of clicks per month. These are your best targets for segment optimization.
For each one, create a dedicated landing page that is focused on the needs, wants, and characteristics of that particular segment. Here you can use content optimization such as A/B testing or MVT to find the best headline, imagery, layout, etc. for each page.
You can almost be guaranteed that these segment-specific landing pages will outperform your more generic ones. With your first few big segment wins in place you can move further down The Long Tail, to less popular, but still easily identifiable keywords to meet each segment.
What about keywords that you can’t automatically associate with a particular segment? In those scenarios, you can use techniques such as multi-step landing pages and “directed behavioral segmentation”. But that’s an article for another day.
Some opinions expressed in this article may be those of a guest author and not necessarily Search Engine Land. Staff authors are listed here.