I had the recently chance to speak with Peep Laja, founder of conversion optimization agency ConversionXL. Peep shared some great insights with me about ConversionXL’s approach to optimizing sites for continual growth.
Don’t Call Peep Laja a “Conversion Optimizer”
ES: I see the ConversionXL website says you don’t like the term “conversion optimization.” Why not? I mean, “conversion” is right there in the agency name.
PL: “Conversion” on its own isn’t the relevant metric. The relevant metric is the one that supports your business goal. Often, this is revenue, but not always. In our agency, we start our discussions with clients clarifying what aspect the business is trying to grow. It could be sales value, it could be number of sales, it could be memberships. Meaningful optimization needs to be about growing the business, not just boosting digital metrics.
[tweetthis]Surveys are still the best way to get qualitative reasoning behind website behavior.[/tweetthis]
Uncovering the “Why” Behind the “Where”
ES: Identifying the leaky spots on a website or specific web page is pretty easy. You web analytics reports will give you metrics like cart abandonment, page views, page bounces, and time on page. Heat map tools show how people are engaging with your pages before they leave – when it seems like they’re looking for something they aren’t finding and what they’re ignoring entirely.
But how do you dig behind the “where” to find the “why” people are dropping off a website so you can put together an intelligent A/B testing plan?
PL: Our team uses qualitative surveys to uncover reasoning behind why people are acting as they are on your website. We’ve found that our clients only need to get around 200-250 qualitative surveys done to get statistically significant information that we can use to design a successful optimization testing strategy.
ES: Does that hold for high traffic sites? Wouldn’t they need a larger sample size?
PL: With around 200-250 qualitative surveys, you’ll start to see the same scenarios and explanations repeat over and over. This sample size holds steady regardless of traffic.
ES: Wow. So we’re not as original or unique as we’d like to think we are.
PL: Yeah, people can be pretty predictable.
ES: So that’s great news for small businesses that don’t have the resources to invest in extensive testing. They don’t need to! Many of the heat map tools include instant online survey and polling tools you can use. I’ve spoken with other optimization experts who counsel against using polls and multiple choice questions for their qualitative surveys. What do you think?
PL: Absolutely avoid multiple choice questions for qualitative surveys. When you present multiple choice questions, you limit users into the pre-defined buckets you’ve already created. The point is to get their feedback, not push them into our assumptions. We need users to have the freedom to share their thoughts as they have them.
[tweetthis]For online behavior personalization to return value, a website has to be generating at least 1000 sales a month.[/tweetthis]
Finding Opportunities in Segmentation
ES: We’ve talked a lot about qualitative surveys. Surely there’s role for quantitative analysis as well?
PL: Quantitative information and analysis are critical for audience segmentation. We find a lot of our clients’ conversion opportunities in identifying very narrow audience segments. These narrow segments go way beyond simple persona demographic and interest segmentation, but digital behavior segmentation. Digging into data such as how visitors come to your site, how much time between their first visit and conversion or purchase.
We add online behavior data points in with other persona data to define the segments within our clients’ website audience. These segments are now narrowed by persona interests, online behaviors, and intent to buy indicators. That makes is possible to customize offers and messages that increase conversion of a specific segment.
ES: That granular level of segmentation must really help high traffic sites hone in on their biggest opportunities and not waste efforts for low or no return. But here’s where small and medium businesses are at a disadvantage, yeah? If they could perform the quantitative analysis, is their sample large enough? And even if it were, do they have resources to meaningfully customize offer language and messages for a high number of narrow segments?
PL: Yeah, we find that a website needs to generate at least 1000 sales a month for it be worth their while to research and execute a narrow segmentation strategy. If your website isn’t meeting the 1000 sales per month threshold, the return on the cost of investing in this sort of hyper-personalization just won’t be there.
ES: That doesn’t mean SMBs should forgo segmentation? They just need to do it on a scale that makes sense for them.
PL: Exactly. They should definitely segment offers and messaging as relevant to personalize a visitor’s experience where you can. Just keep an eye out for the point of diminishing returns in terms of investing time and money into the online personalization.
[tweetthis]For online behavior personalization to return value, a website has to be generating at least 1000 sales a month.[/tweetthis]
Speaking of the Point of Diminishing Returns
ES: I also read on your website that you work with your clients until you hit the point of diminishing returns. How do you quantify that point?
PL: We do keep testing and refining with our clients until they reach the point of diminishing returns, which is when we see the same amount of improvement month over month.
For example, you run through a number of A/B test iterations and start seeing a big boost in your growth metric. Let’s just say “sales” here to keep it simple. After an initial big jump, the increases will settle, say 5% per month. In this scenario, once we see a 5% boost in sales month over month, you’ve reached the point of diminishing returns. This is where we gracefully remove ourselves.
But we don’t leave our clients alone to fend for themselves though. Throughout their time working with us, we train clients how to test and refine on their own website to continue to find optimization opportunities. We make these processes and expertise available to businesses that aren’t clients through our ConversionXL Institute.
ES: Well, thanks Peep, for giving us a peek behind the curtain of how ConversionXL operates.
PL: Sure, my pleasure.
Wrapping It Up
You can hear from Peep directly via Twitter (@PeepLaja). Here are some other ways you hear more from Peep and the ConversionXL team:
ConversionXL website (http://conversionxl.com)
ConversionXL Institute website (https://conversionxl.com/institute)
Or join them at ConversionXL Live 2017 – a 3-day event this April in Austin, Texas, featuring a ton of UX and conversion experts.
Elisa Silverman is a B2B content writer with a background in law and technology, who’s spent a career helping diverse groups of people communicate well with each other.
Also published on Medium.
