hero image

Second Funnel Pages

Increasing conversion up to 5x over conventional eCommerce sites

Second Funnel is an eCommerce software startup I worked on. As a member of the founding team, I helped define our first product and got it live with our first customers. Ultimately we turned a hunch into a product that was used by companies like Gap and Sur la Table and visited by millions of online shoppers.

Second Funnel provided a way for marketers and merchandizers to deploy custom eCommerce pages without having to change the code on their site. Rather than try to squeeze more conversion out of an existing purchase path, Second Funnel enabled an entirely new experience that was optimized for inspiration and discovery. As a result, the pages generated by Second Funnel were able to increase conversion by up to 5x over conventional eCommerce pages.

We didn’t have a product when I joined, but we had identified a real problem in eCommerce and had some theories about how to solve it. My role evolved with the company and turned out to be a combination of customer development (figuring out what to build and for who), user research and testing, and interface design.

While at Second Funnel, I helped define our product, validated our early hypotheses, and designed an eCommerce interface that significantly increased conversion rate and time on site for our customers.

We observed a problem in eCommerce

Looking at eCommerce sales, we noticed that referral traffic from social media sources converted about 4X less often than other sources of traffic.

conversion rate bar graph

Meanwhile, people’s time online was dominated by social media and, in particular, visual content (think: Pinterest and Instagram). Over half of time online was now spent leisurely consuming content.

As expected, retail brands adjusted their marketing tactics to align with this new behavior. By some reports, over a third of marketing budgets were allocated for content creation and social media marketing.

Why then, with so much attention directed towards content marketing and social media, were these channels converting so poorly?

We had a hunch…

To change a behavior, you need to change the environment.

While the UX of a conventional eCommerce site works great for people who already know what they want, we believed that the experience totally alienated window shoppers (i.e. people who are interested enough to browse but don’t have a strong intention to buy something).

In our view, a marketer should only send people to their eCommerce site who have a strong desire to buy. We saw a need for a middle layer experience that would allow a marketer to stoke a window shopper’s interest until they were absolutely ready to make a purchase.

Our goal: Second Funnel would be this middle layer experience.

Unknowns we needed to answer

  1. Can we create a new purchase path that ultimately increases conversion rates?
  2. Will marketers adopt our approach and send traffic to us instead of the eCommerce site?

Journey mapping

I started our design process by doing user research to learn more about the status quo. First I talked with our peers, asking them about how they use social media with brands and how they shop online. Then I observed as they navigated the purchase paths that had been set up by most retail brands. For example, I watched people go from, say, a brand’s post on Facebook to a product detail page on an eCommerce site. At each step I noted what people were feeling, thinking, and doing.

journey map

Pain points

It was clear that most people had a negative experience when they got to the the eCommerce site from a social site. There were four main aspects of the user experience that needed improvement:

  1. People feel lost & confused because what they see on the landing page often doesn’t relate to what they clicked on.

  2. People feel frustrated because they can’t quickly and easily find what they wanted to see, which is often an outfit or the products shown in the advertising creative.

  3. People feel overwhelmed because there are too many options and too many decisions to make about where to go next.

  4. People feel pressured because they’re often confronted with aggressive promotions and they aren’t in the mood to make a decision.

Leverage points

I also thought about the levers that influence someone to buy something. In general, a purchase depends on three factors:

  1. Exposure. How much someone is exposed to an item.
  2. Desire. How much someone desires an item.
  3. Friction. The ease (or difficulty) involved with the whole process, from first seeing an item, to getting information to justify the purchase, and finally executing the transaction.

To influence purchase we would need to increase exposure and desire while reducing friction.

Design goals

Drawing on what we learned from the research, we used the How Might We method to collectively decide on some design goals. Our solution would have to answer the following questions:

  1. How might we create a seamless transition between entry-point and landing?
  2. How might we make it easier to discover products?
  3. How might we create a “don’t make me think” shopping experience?
  4. How might we create an environment where people want to spend time browsing?

Design patterns

We explored design patterns that were already working to hold people’s attention in media and retail.

Inspiration came from image-based social media sites and lifestyle blogs and magazines—places where people were already spending a lot of time online getting inspired.

design patterns

We also looked at pure play eCommerce innovators (Amazon, Nasty Gal, Mr. Porter, and Betabrand) that we thought were doing a great job of marrying editorial and commerce.

Finally, we thought about what makes physical retail spaces so enjoyable to browse: In a nice boutique, products are shown in context, so shoppers feel how they are connected to a lifestyle; it’s easy for people to scan around and quickly get a sense of what’s there; it’s frictionless to pick something up and put it back down; and there is almost no pressure to buy.

Sketching a product concept

I combined various design patterns into a product concept that we all agreed would best solve for our design goals.


product concept sketch

Prototypes and early validation

With a product concept roughly defined, our next step was to get working prototypes in front of customers and online shoppers so that we could test our concept.

I reached out to prospective customers and asked open-ended discovery type questions followed by a show-and-tell of our concept. These early conversations with directors of digital marketing and eCommerce helped us validate some of our ideas and got us our first two beta testers.

early prototypes

Our first prototypes showed that we could increase key metrics like time on site, items viewed per visitor, and click-through rates. The data we gathered was enough to help us get our our first paying customers.

Testing and iteration

We iterated on the product and the UI as we worked with more retailers and continued to run A/B tests. Every design decision was measured against our analytics, which tracked the funnel of page-level metrics leading to conversion.

To improve our metrics, I experimented with design details like the hover state on tiles, showing information like price and discount percentages, and changing the style and wording of the CTA. We also played with the ratio of lifestyle imagery to product imagery and tested various grid configurations.

design iterations

For our gift guide product, I designed and tested multiple versions of the category-level navigation, and ultimately increased the percentage of visitors who visited multiple categories from less than 10% to over 25%.

navigation design iterations

And, for our “shop-the-look” interface, I evolved the design and increased click-through rates from single digit percentages to between 10-25%. This was especially important for mobile devices, which accounted for over 50% of traffic.

mobile iterations

mobile shop the look interactions


Ultimately, the product we built was used by retailers like Gap, Samsung, and Sur la Table, and reliably lifted conversion rates by as much as 5X when A/B tested against traditional eCommerce pages.

outcome before and after

See More Case Studies