How Personalization Affects Conversion — What We Learned from 500,000 Visitor Journeys

In Part 1 of the series, we discuss what types of notifications visitors are anticipating on your website, while Part 2 covers where to place those notifications. Now, let’s look at how web notifications fit into personalization strategies and what results they yield.

eCommerce marketers have been determined to enhance visitor journeys through data, optimization and analytics for some time. To reach this goal, marketers became usability-driven and directed resources into cosmetically optimizing websites — from testing headers to Google’s 41 shades of blue.

While endless optimization produces positive outcomes, visitors grew more accustomed to well-designed websites. Now, the visitor satisfaction is more about what meets the eye, literally.

Being functional is not enough for consumers — it’s a new norm. They expect websites to be personal and reflect their individual needs.

Companies which recognize this shift are ahead of the game — brands who create personalized experiences by integrating data and advanced technologies are currently achieving revenue increases of between 6–10%.

Lately, we’ve been helping such companies strategize and deliver personalized experiences for their visitors, and here’s what we learned from their visitors’ journeys.

Difference between Optimization and Personalization

Before we dive into strategies, I’d like to draw a distinction between ‘optimization’ and ‘personalization’.

Optimization simply means improving your website by testing and implementing the top-performing design. It’s low-hanging fruit for marketers — elements like the size and position of add-to-cart buttons can be tested, and their impact is easily measured. However, getting tangible results takes time. An A/B test has to reach significance before the learnings can be analyzed and implemented, which prevents a marketer from delivering the enhanced experience in time. Increasing the ubiquity of CRO tools makes us think beyond the page.

In other words, think Personalization. Instead of improving the website, personalization improves the experiences of specific visitors, often leveraging Artificial Intelligence technology. Backed by the power of machine learning, companies can now offer each customer a personalized experience based on all the data that is available from the machine’s first collection up to the present moment.

Optimization maturity stages

We’ve gathered data on how personalization strategies impacted over 500,000 visitor journeys to understand what results to expect from them. Here are our key takeaways.

1. Segmentation fuels performance

Challenge: Every business has products or services that deserve the biggest amount of marketing due to the revenue they produce. So did a B2C brand that wanted to drive traffic to a particular feature on the website. While having healthy monthly traffic on the blog and the website, it was crucial to target those visitors who were not predicted to engage in other key funnels. To do so, we had to rely on narrowly-defined segments.

Strategy: Use notifications to engage visitors who are new to the marketed feature, predicted not to engage in other key funnels, and showed interest in blog posts with tags relevant to the feature.

Results: 2.13% conversion lift

Why it worked: Visitors involuntarily expressed a potential interest in a feature we wanted to showcase. We leveraged that interest and pointed visitors to the part of the website that was not only crucial for business, but more importantly, useful for visitors.

2. Complex funnels benefit from timely personalized notifications

Challenge: Sometimes, complex funnels are necessary. Think of a funnel, one step of which is a long form. You want a visitor to fill it out in order to collect data that is valuable for your business. At the same time, getting a visitor past the form to the final steps of the funnel remains a priority. How do we give a visitor the chance to provide us that information? And what do we do if the visitor shows hesitation / decides to leave?

Strategy: If a visitor has a proclivity to drop-off on the Form step, send a notification saying the form is optional to proceed.

Results: 6.55% conversion lift

Why it worked: Visitors who demonstrated commitment by providing their data were notnotified. We made it obvious that the form is optional, but only for those who were about to leave. In this way, the business has the data it needed, and the visitor is nudged at the precise moment with effective messaging to move towards the final funnel step.

3. Returning visitors appreciate personalization

Challenge: Often, despite your marketing efforts, visitors leave. It’s important to treat them properly once they are back. Initially, we saw a 75% drop off rate for returning visitors who left mid-funnel. On their return, they could have browsed even more pages than during their first visit. Still, none of their actions led to conversion and they left once again. It was crucial to take those visitors back to where they left off, and to where a conversion was about to happen.

Strategy: Send a notification to returning visitors that reminds them where in the funnel they were, by using environmental variables.

Results: 11.21% conversion lift

Why it worked: If visitors left the first time, they were not hooked. With the usage of an environmental variable, like location, we notified visitors that in the end of the funnel they could get in touch with professionals in their area. This message provided more value when visitors were reminded to which area/direction/step they should head.

Personal Notification for a returning visitor

4. Predictive marketing influences visitors

Challenge: Often, your traffic peruses your blog or windowshops rather than has a straightforward conversion path. Your stats may be showing healthy overall traffic, but it doesn’t result in any tangible revenue for your business. We needed to attract wandering visitors into a sign-up funnel, but do so at the opportune time. Each visitor had their own definition of when that time was.

Strategy: Use predictive segmentation to distill the right audience for the funnel.

Results: 4.72% conversion lift

Why it worked: Over time, the predictive model learned what visitor patterns indicated a proclivity towards a certain funnel. By looking at digital footprints of other visitors, the model determined whether a visitor was eligible for this notification, or whether it was better to not expose them to this campaign.

5. Ask for a big commitment after a smaller one

Challenge: Some products or services you want to offer may appear as a big commitment for a visitor — either the price is high or the amount of data required to sign up is vast and personal. At the same time, marketing that particular product or service may be crucial for your business. How do you nudge a visitor at the optimum momentum without turning them off completely?

Strategy: Congratulate visitors on a small conversion they’ve made and guide them towards the new funnel.

Results: 3.8% conversion lift

Why it worked: Visitors like to be acknowledged. If they have successfully progresses from one step in the funnel to the next one, or caused another micro-conversion — let them know with gratitude. Their actions show loyalty which can lead them to the product or service you want to showcase.

Notification regarding another commitment at the end of the funnel

6) Mobile notifications are easy to ignore, unless they are not obvious

Challenge: Ecommerce companies develop apps to compliment their online experience and give users a choice about how they want to interact with the business. A newly built app needs exposure, and your website traffic is your captive audience. Many businesses use this tactic on their websites, going for an industry standard of design. So did we, only to realize that visitors were desensitized to these types of notifications and ignored them.

Strategy: Segment mobile users, and target them with an instant notification — placed on the bottom and uniquely designed.

Results: 0.73% conversion lift

Why it worked: If you hop on the website that wants you to download the app, chances are you’ll see a top inline notification that will push the content of the page down. The messaging was more reflective of what the app can do for the visitor instead of “Download the app”. We exposed mobile users to the bottom inline notification instead an added unique design and messaging to it.

7) Notification overload is real

Challenge: You may notice that some campaigns share an overlap in visitors. Overlap may occur if you use predictive segmentation but? don’t always have visibility into the audiences. Because your visitors may be eligible for different types of campaigns during their sessions, they’re then exposed to multiple notifications. When does the experience shift from pleasant to cluttered because of one notification too many?

Strategy: Determine and set the standard delay between notifications based on the average time on site.

Results: Overall 0.5–1% lift for campaigns.

Why it worked: Visitors are quick to judge if you annoy them. Sending one notification after another is a slippery slope, especially if their time on site is short. By determining the average amount of time a user spent on the website, and setting a standard delay between notifications, we prevented notification clutter. In this case, less is more.

To Sum Up

McKinsey & Company estimated that personalization has the capacity to reduce acquisition costs by as much as 50%, lift revenues by 5% to 15%, and increase the efficiency of marketing spend by 10% to 30%. We witnessed this transformation potential become actualized by personalizing visitor journeys for them. See the case study here.