EPflow is a uniquely scalable display optimization service. It can minimize visual clutter for consumers while increasing profits for retailers and agencies. EPflow brings relevant items to the attention of consumers by automatically generating and evaluating thousands of alternative displays. It is powered by a software technology that predicts visual attention based on cutting-edge neuroscience research.
EPflow optimizations may take seconds, minutes, or hours, depending on the particular circumstances. The optimization process is fully automated and highly parallelizable. In the advanced use case, optimized displays can be delivered in real time, like Google search results. This is made possible by pre-optimizing displays based on historical site use. For example: many ecommerce sites are characterized by high correlations between product searches performed on consecutive hours, days, and weeks. Optimizing all the displays generated by product searches performed during a given hour enables EPflow to deliver pre-optimized displays in real time when the same product searches are repeated during the next hour.
The price of EPflow optimizations depends on several factors, including volume (upfront commitment) and pricing model. We offer flexible pricing plans tailored to specific customer preferences. We can work based on either licensing, pay-per-use, or pay-per-performance.
EPflow automatically generates and evaluates thousands of alternative displays based on optimization criteria and computer predictions of visual attention. Conceptually similar solutions include split (A/B) and multivariate testing, but neither is sufficient scalable to optimize grid displays. In comparison, EPflow is more scalable by orders of magnitude because it relies on computer predictions instead of behavioral measures.
Optimization criteria express the relative relevance of display items. The most effective approach to determine relevance and prioritize items may depend on several factors, such as purchase likelihood and profit margin. Let's consider two hypothetical scenarios:
Your business strategy is to maximize profits by increasing repeat business from individual consumers. You are willing to potentially sacrifice short-term profits if making them would decrease the likelihood of repeat business. In this case, it would be prudent to prioritize items based on their purchase likelihood.
Your business strategy is to maximize short-term profits by increasing sales of more profitable items. You are willing to potentially sacrifice long-term profits if making them would decrease the likelihood of short-term profits. In this case, it might be useful to prioritize items based on their profit margin.
Whatever your business strategy may be, EPflow can optimize displays such that the most relevant items are most likely to get noticed and selected by consumers.
By default, EPflow assumes that display items to be optimized are sorted by descending relevance from top-left to bottom-right. Relying on this default setting minimizes the friction of using EPflow at the cost of not taking full advantage of its potential value. An alternative approach that can generate more value to customers is to send to EPflow a feed of relevance scores per display.
A series of validation studies yielded 326,258 clicks by 7,051 consumers and demonstrated that EPflow can triple click-through rates (CTR) for display items. Earlier validation studies were based on eye tracking and demonstrated strong correlations between our software predictions of visual attention and human fixations in images and videos. Some of these studies were commissioned by 3rd parties, including a behavioral research consultancy and a Fortune 200 customer that later offered to license the underlying technology. Please email firstname.lastname@example.org to receive our validation white paper.
Paying attention to the right thing at the right time underlies the ability of people to interact with their environment. The central role that attention plays in human behavior has led to wide ranging scientific efforts to characterize the underlying mechanisms. Attention studies typically rely on neural and behavioral data that are collected using a variety of techniques, such as brain imaging and eye tracking. Insights gained from these studies enabled EyePredict to develop realistic computer models of visual attention that can be used to optimize product displays.
What is the difference between EPflow and eye tracking?
Both EPflow and eye tracking can solve problems that relate to visual attention, but eye tracking is based on human evaluation, whereas EPflow provides computer optimization. Eye tracking can result in an evaluation report that shows how visual attention is deployed based on a study that typically involves <100 consumers and visuals. If done well, eye tracking reports also include valid design tips based on non-trivial interpretation by human experts. In contrast, EPflow automatically rearranges interfaces to improve consumer response, which highly differentiates it from eye tracking in terms of both actionability and scalability.
Do people always pay attention to certain locations or items (top-left, red objects)?
People exhibit statistical biases, but what they actually notice in a given situation is highly context-dependent. For example, imagine a 5x5 grid containing 25 identical Coke cans. Neither of the Coke cans is particularly noticeable. Now leave one Coke can unchanged but replace all the surrounding Coke cans with Pepsi cans. This contextual change causes the remaining Coke to stand out, even though it did not change in any way. This simple example illustrates that the features of both the target and its surroundings strongly affect attention deployment. Scientific research has demonstrated that attention deployment is determined by complex interactions that cannot be reduced to a simple set of fixed rules. Predicting attention deployment is particularly challenging in the context of real world commercial displays, which are often much more complex than the simple Coke/Pepsi grid described here.
Is EPflow sensitive to differences between consumers?
EPflow can take advantage of consumer-specific information to improve its predictions, but it can also optimize commercial displays without such info. This is possible because different consumers are highly similar in their attention choices, particularly during brief exposures to new displays. If consumer-specific information is provided, EPflow can use this info to increase the accuracy of its predictions and the value of its optimizations.
EyePredict provides automated display evaluation and optimization solutions. We can increase the likelihood that relevant display items will get noticed by consumers, which is a critical step on the path to conversion. However, the EyePredict technology can neither create displays from scratch nor ensure that they lead to a particular emotional impact. Designers can use the EyePredict technology to ensure that display items stand out, but they cannot be replaced by it. The relationship between EyePredict software and designers is analogous to the relationship between CAD software and architects.
Do I need to install anything in order to use EPflow?
Typically, no. Customers need to use a Web browser, such as Firefox, Chrome, Internet Explorer, or Safari. EPflow does not require any plug-ins, but your browser must support standard features, including SSL, AJAX, and cookies.
Secure Sockets Layer (SSL) is a standard Internet protocol that safeguards electronic communications between two or more computers. When transmissions are sent over an SSL connection, data is encrypted and cannot be modified in transit by other parties on the Internet. We provide SSL support at no additional cost, depending on customer interest.