How Does Granify Work?

How does granify work-100.jpg

Granify increases e-commerce revenue. The next logical question is, how?

There are a few ways to answer this question but the best way depends on your background and the purpose of asking this question in the first place. For example, if you’re a software engineer, you’ll likely want a more technical answer. If you are a UX designer, you’ll likely want an answer that describes how the customer experience is changed.

I wanted to get a well rounded answer, so I asked the team: How does Granify work?

Of course, if you want a more detailed answer, or one specifically for your e-commerce site, you can request a free consultation.

How Does Granify Work? (Answered By Granify Employees)

The short and sweet answer:

Granify acts like an in-store sales associate, online.

— Danielle Malgieri, Enterprise Account Executive


Regarding the user interface:

Our smart messages often present the shopper with information that exists on the site but is not readily accessible. For example: a site may offer free returns in store, but this info may be buried in a footer link. We take the information and make it more accessible to the shoppers who will get the most value from it.

— Mike Buss, Designer


Regarding the user experience:

The smart messages are targeted to gaps in the shopper experience that cannot be addressed at a site-wide level; for example, if you notify every shopper about your returns policy, you will overwhelm and likely aggravate them. However, when the shopper’s behavior indicates concern about the policy, this information is exactly what is needed to overcome their concern.

— Chris Murray, Designer


Addressing how fast Granify works:

Granify works by closing and shrinking the e-commerce optimization feedback loop. Typically, retail and e-commerce companies aggregate user data, and then make a UX change long after those customers left their site. With Granify, the customer concerns are addressed in real-time.

— Lacie Larschan, Content Marketing Manager


How behavioral targeting is valuable:

Rather than target people by bucketing them based on mining profile data (which is usually missing or wrong) we target people using AI that looks at their shopping behavior. It's more accurate, has a higher yield, and is completely anonymous so it’s more private too.

— Christopher Holmok, Senior Software Engineer


Addressing the business need:

Granify is an e-commerce conversion secret weapon. We can observe, predict, and act on each of your customers as they shop your site, in their moment of need. Externally, if our machine learning technology ever notices an opportunity to keep your shopper moving through your conversion funnel, we can introduce stimuli to delight your shopper and keep them focused on buying. Internally, we have the power to make sense of your traffic chaos by running 100s of micro tests to determine best case shopping experiences tailored for each of your visitors.

— Jay Doughty, VP of Growth


On manpower and ROI:

Granify has developed machine learning that can predict hundreds of different behavior indicators on the shoppers you hold the highest influence over and then deliver smart messaging to close those deals. Identification and analysis on how to better convert those ‘on-the-line’ shoppers is a task that companies build teams around and spend millions of dollars on, and Granify can do it in real time with no manpower required.

— Chris Vaughn, Director of Marketing & Sales Operations


The answer fit for prospective employees:

Think of a time when you've gone shopping at a regular brick-and-mortar store and encountered a fantastic salesperson: someone who understood your needs, had all the information, and was compelling but not pushy—someone who generally made your shopping experience that much better.
As soon as you walked into the store they were immediately evaluating your behavior and body language: what you looked at, if you were in a rush or just browsing, if there was something specific you were trying to find. Based on that information, they personalized their approach to what you needed at any given moment.
That's what Granify's messaging does but it’s one system to serve millions of online shoppers at the same time. It analyzes your digital body language—what images you view, your scroll rate, mouse movements, time per page, and hundreds of other tiny micro-actions—to determine what message you should see (if any) and when the best time is to show it to you. This creates a personalized, optimal online shopping experience.

— Heather Whyte, People Operations Manager


The short answer from our top data scientist:

Granify increases sales by improving the shopper experience on retailers' websites. It’s like personal shopping assistant, just better because we don’t ask how we can help, we already know how we can help and do so in the most efficient way.
This is a highly personalized service as Granify tracks each individual online shopper’s behavior at a high level of granularity: how they move their mouse/touch screen, which product they are viewing, what they add to cart, etc. All this data is then analyzed by our servers and we apply machine learning models to predict a shopper’s concerns and needs. Then, our servers analyze the available tool set and make a decision whether to show a message to the shopper or wait because the shopper doesn't need help right now.
All of this happens in real time—we act before shoppers leave the site. We provide excellent results as our machine learning algorithms are trained using vast data on shopper behavior and reactions that we can then match with current shopper sessions.
The great thing about Granify is that we know how much value we generate for a retailer and we only charge based on this additional value. We split all incoming traffic into two groups, one in which we only observe shoppers but do not interact with them (a.k.a. baseline group), and one in which we do interact with shoppers. This is one big A/B test which can tell us how much incremental revenue our AI service generates by comparing performance in each group.

— Marcin Mizianty, PhD, Director of Data Science



Looking for more? Our readers also liked...

Want weekly insights in to the world of e-commerce optimization?

Recent Posts