In order for your company – or clients – to sell better online, you need insights. Not the unreliable ones that come from long-winded surveys, but the insights come from either actively talking with your prospects, or at least passively observing them. Here I’ll focus on the second method, because it’s the most cost-effective way to get good qualitative insights quickly. Most importantly, these Insights will tell you why your visitors aren’t finding what they want to know, and thus why they’re bailing out of your conversion funnel much too soon.

When marketers and analyst talk about analytics these days, they’re generally talking about data analytics, like those you get from Google Analytics or similar analytics suites. Or, if you’re more sophisticated, you may be aware of predictive analytics, data mining, or segmentation analytics. These tools and algorithms are certainly all the rage these days. But unless you collect reams of behavioral and transactional data every week, and have significant dollars to invest, these tools will be out of your reach.

But don’t fear: for $50 bucks a month and a few weeks of patience, you can gather some amazingly insightful visitor analytics, data that will tell you:

  • The top tripping points on your website user experience
  • Your visitors’ top questions and concerns
  • Overall, where your designs are hot, and where they’re not (literally!)

Plug in one or more of these tools, and you’ll be well on your way to knowing – versus guessing about – what makes your visitors click, and ultimately makes them convert.

Since it’s an election year, let’s start with polling.

Run a one-question poll

image_poll_3q_article_markhall

Not sure where your site’s messaging is mismatching your visitors’ expectations? Ask! It’s as simple as that, yet done all too rarely on the web. Early in your web store’s user experience, push a one-question poll that asks:

‘What are you looking for today?’ or ‘Is there anything we can help you find today?’

Subtly push this layer in the bottom right corner after a specified time (10-15 seconds is a good starting point; you can always adjust this timing later). Don’t make the answer multiple choice, as those answer choices can be restrictive. Instead, just include a 5-line comments field.

Make sure the visitor can collapse this layer at any time so it’s not annoying. This layer should just sit there quietly, always welcoming comments.

If your site already does a good job of funneling your visitors to the right section, you can instead ask a different question on your landing, sectional or category listing pages. In this context, a good question to ask is:

‘Any questions or concerns keeping you from continuing?’ or ‘What’s stopping you from buying from us today?’

Responses to questions like this will tell you what key content is missing on your ‘decision tipping point’ pages, questions that, once you know, you can proactively answer in your next design update, so that your visitors will more smoothly slide down your conversion funnel.

Two to three hundred responses is enough to get some meaningful insights. As an example, if your site has over 50,000 visitors a month, you serve this poll to 20% of visitors, and you get a 3% response rate (typical for these polls), you should have this number of responses within a month. If your site doesn’t have that many visitors yet, you should wait to use polls. In the meantime, though, you should scheduling times to speak directly with your most valuable customers (B2C) or clients (B2B).

Finally, I’ll share a word of caution: Resist the temptation to run a survey! Surveys longer than 1 or 2 questions have two fundamental problems: they’re often inaccurate (a lot of surveys in the last presidential election predicted that Clinton would win, and how did that go?), and they’re subject to inaccurate post-action rationalization (they don’t capture the ‘gut’ sentiment or emotions that most drive humans’ decisions). What’s more, they are downright annoying! So do don’t ask a lot of questions; just show them you’re there, you care, and you’re listening. Isn’t that how most people most want to know?

Configure and review some heat maps

image_heatmap_3q_article_markhall

The latest neuroscience and ethnographic research has convinced me that what people do is much more telling than what they say. As University of California at Berkeley neuroscientist A. K. Pradeep succinctly put it, ‘The act of thinking about a feeling changes a feeling.’ Ask an image-conscious friend why they bought that expensive car or tricked-out truck and you’ll see what I mean.

So put most of your research attention on what your visitors are doing – where they’re scrolling, interacting, and clicking. All you have to do is set up a few heat map reports. Then, after your chosen tool has captured enough clicks, you’ll be able to view click and scroll areas on colorized maps. Red and orange mean it’s a ‘hot’ area and is getting a lot of attention; blue and green mean it’s a ‘cool’ area, not attracting many clicks. Of course, your goal is to get interactions where you want them, like call-to-action buttons, and not where they are irrelevant, like on copy or videos that outside of your ‘happy conversion path’.

I’d be remiss if I didn’t mention this caveat: I know from running hundreds of user tests that where your visitors are clicking doesn’t exactly match where they are looking (their visual ‘focus cone’) and what their brains are perceiving (within their peripheral vision). The latest estimates put this click-to-eye correlation at about 80%. But, while heat maps aren’t as accurate as methods like eye tracking, they do give you a good sense of what elements on the page are attracting the most attention, and thus where your design is and isn’t on track.

Watch some session recordings

image_liketowatch_3q_article_markhall

I have a confession: I like to watch ‘fail’ videos on YouTube. It’s cruel, I know. And I feel like I’m doing the same thing when I watch session recordings: videos that show your visitors using your website, at least up to the point where they bail out. Even though there isn’t any audio, you can see clearly what your visitors are seeing. And after watching several sessions you’ll know where most people are hesitating, or tripping up on forms, or otherwise not doing what you want them to do.

I’ve found that you’ll learn something useful from about half of the recordings, so plan to watch 30 or more recordings on both the sales and transactional sides of your conversion funnel. If you own an eCommerce site, this will be your top landing pages, homepage, category pages, product pages and checkout pages. If you run a lead gen site, this will include your landing (form) page and form confirmation page.

Sum up your findings

After you’ve collected enough data, take a couple of hours to review the data and write a short (1-2 page) ‘key findings’ report. Be sure to note the reference numbers of sessions that best highlight common issues visitors are experiencing so that you can easily find them later. Then share your observations and findings with your teammates in person. I can guarantee that you’ll have a good discussion. Best of all, this will be based on real data, not just a trip to opinion-ville.

And don’t turn off your analytics too soon just to save a few bucks. Revisit the data in a month or two to see if you’re still seeing the same trends. Or set up new heat maps on other pages. You will likely find, as I have, that spending money on such a tool will give you a very high return on investment (revenue lift achieved versus dollars spent). It may even be the best research investment you’ve made in years.

So, which tool to use?

Finally, let’s talk visitor clickstream analytics tools. In the last few years I have used two tools the most: LuckyOrange and HotJar. Both include heat maps, feedback polls, session recordings, form analytics, and basic conversion funnels. Only LuckyOrange, however, includes a built-in chat tool. So if you don’t have the budget for a separate online chat solution, this tool may be a better choice for you. Pricing for both tools is around $50/month for a mid-level plan.

If you’re willing to spend more, and want some deeper logic and reporting for your polls, check out Qualaroo. (No, it’s not from an Australian company, and it doesn’t capture bounce rates, but these guys were the pioneers in online polls, and the tool is as polished as they come.) If you use Visual Website Optimizer or Optimizely, I noticed recently that these tools now integrate with ClickTale and CrazyEgg, more high-end click-tracking suites. So you may not need to buy a separate tool.

Start tracking clicks and questions!

If you’ve got data analytics but don’t have visitor analytics, you’re missing out on understanding the key reasons why your visitors aren’t sticking on your shopping or lead-generation website. Especially if you’re on a tight budget, you’ll find that tracking visitor clicks and sessions is well worthwhile, even if you’re not a professional voyeur like me.

So don’t wait! Sign up for one of these tools, set up a few reports, then congratulate yourself! Because you and your team will have taken your first step to gaining a little customer empathy and gathering some great insights, insights that will soon allow you to confidently turn a few website updates into conversion gold.

Leave a Comment


Mark Hall is an eCommerce Expert, Conversion Rate Optimizer, Voice of Customer (VOC) Researcher and Concert Band geek. Having consulted for over 20 years with a wide range of clients including AT&T, American Express, Edmunds, SpyTec and the California Lottery, he understands which marketing strategies and tactics will best boost companies' conversion rates and revenues. Connect with Mark: blog: www.marksconversioninsights.com/subscribe linkedin: www.linkedin.com/in/MarkHallVOC google+: plus.google.com/+MarkHallVOC twitter: @MarkHallVOC