![\"Retail](\"https:\/\/dmsretail.com\/RetailNews\/wp-content\/uploads\/2022\/05\/RETAIL-ONLINE-TRAINING-728-X-90.png\")
<\/a><\/p>\n<\/p>\n
At this point, most brands have a solid grasp of the theory behind user experience (UX). It means making it easy for users, buyers or readers to interact with a site or app. It means understanding user expectations and anticipating their pathways, challenges and goals. It\u2019s all about streamlined functionality.<\/p>\n
As technology has gotten more advanced \u2013 and users have gotten better at using it \u2013 UX has become a business in and of itself. We have proven best practices, data-driven strategies and experienced vendors and consultants. But just as some brands are finally becoming proficient in UX, just when it seems like they\u2019ve cracked the code\u2026UX is on the cusp of fundamental change.<\/p>\n
Soon, humans won\u2019t be the primary users of brands\u2019 websites and apps. AI agents will be. Agents are getting smarter and more effective every day. Right now, many agentic AIs are just one step beyond your average chatbot, but soon they\u2019ll be able to do so much more. We\u2019ll come to use AI agents to make purchases for us; to book hotels and airline tickets; to pay our bills; to conduct research and report back with an informed recommendation.<\/p>\n
From a UX perspective, this means everything is about to change. In time, UX will be more about the agentic <\/em>user experience than the human <\/em>user experience, and brands need to get ready.<\/p>\n In every industry, you need to understand your audience to be successful. But UX is more make-or-break for some sectors than others. Businesses in sectors like retail, hospitality and QSR can improve user retention rates by 15% just by improving their UX.<\/p>\n But here\u2019s the thing: what works for human <\/em>users may not work for agents, or vice versa. Brands need to take time to understand how their agentic users may differ from human users.<\/p>\n Broadly, agents will likely come to a site or app with more of a mission than a human would; agents will be less susceptible to browsing. They\u2019ll have received a very specific directive to inform that visit to the site (e.g., \u201cFind me a dress for a semiformal wedding in Vermont in June.\u201d or \u201cOrder me a supreme burrito with a side of guac.\u201d) This means the tried-and-true upsell techniques that work well with human browsing (e.g., displaying a carousel of shoes that go well with a particular dress) likely won\u2019t be as effective for agentic users.<\/p>\n Brands will need to rethink their tactics and start optimizing their digital platforms for the next generation of users.<\/p>\n Unless agents are given direction about specific brands\/sites to visit to complete their tasks, they\u2019ll likely perform a search function as a first step in their workflow. This means that if brands want to be on an AI agent\u2019s shortlist (i.e., if they want to be found) they\u2019ll need to dial in their SEO and GEO (generative engine optimization) practices.<\/p>\n SEO helps sites rank well on traditional search engines, while GEO helps them rank on AI-driven search engines. It may be hard to predict how, exactly, AI agents will complete their searches, so brands should ensure they account for either possibility.<\/p>\n There\u2019s a lot of overlap with SEO and GEO best practices, such as digestible content and alignment with user intent, but there are some key differences to keep in mind. SEO uniquely favors domain authority, backlinks and keyword matching, while GEO favors EEAT (experience, expertise, authoritativeness, trustworthiness). If brands get a handle on both SEO and GEO, they\u2019ll be well positioned to capture the attention of agentic users.<\/p>\n Of course, once agents are on the site, they\u2019ll need to be able to easily find what they\u2019ve been sent to look for. The product catalog must be enriched with descriptive product attributes (e.g., size, color, material), detailed metadata and alt-text or captions for any imagery. To support easy navigation, brands can create a .txt <\/em>file that lists each page and asset on the site. This sitemap lives only on the back end of the site and will ensure that agents and LLMs are aware of all the site has to offer.<\/p>\n Then comes page navigation. Brands may want to look at the various buttons and menus on the site and consider how an AI agent may interpret them. While informal, casual language (e.g., \u201cCheck out what\u2019s new!\u201d or \u201cFeeling hangry?\u201d) support brand voicing and identity, agents likely won\u2019t appreciate it like humans would \u2013 and they might not even understand it. Prioritizing clarity and simple language is likely the best path forward.<\/p>\n Finally, brands will need a way to determine whether their site visitors are humans or agents. As previously noted, there are nuances that distinguish agentic behavior from that of humans, and the right tools can flag these nuances to identify visitors accordingly. This identification process will prove pivotal when it comes time to start measuring success.<\/p>\n One way brands can evaluate whether they\u2019ve done it right is by looking at sales numbers and metrics like abandoned cart rates. But while these KPIs are undoubtedly useful, they likely won\u2019t provide the granular detail brands need to build superior UX. To go deeper, brands should tap into agentic behavioral data capture and analysis.<\/p>\n Behavioral data capture lets brands track all user behavior \u2013 whether human or agentic \u2013 on their site or app. Analyzing activity like bounces, dead clicks and scroll depth can help identify the best-built pages as \u201cnorth stars\u201d and illuminate opportunities for improvement elsewhere. For instance, if brands see that they have robust visitor volume on a given page but very few clicks, they can flag that as a disconnect worth investigating further.<\/p>\n By delineating human and agentic behavior, reoptimizing digital platforms and leveraging behavioral data analysis, brands can confidently enter the new, agentic-first era of UX. It\u2019s worth noting that this change won\u2019t happen overnight; brand leaders have plenty of time to do their research, refine their approach and implement any new tech tools that might help them cater to AI agents. But the sooner they get started, the more equipped they\u2019ll be to go boldly into our new digital frontier.<\/p>\n As <\/em>Fullstory\u2019s<\/em> Chief Product and Technology Officer, Claire Fang brings more than two decades of product leadership experience to the executive team. With a background spanning public companies and startups, Fang brings a wealth of expertise in delivering innovation in enterprise software, building world-class product and engineering organizations. Before Fullstory, Fang served as the Chief Product Officer at SeekOut, leading the company\u2019s product management, design and marketing functions, and was responsible for product vision, strategy, roadmap and execution. Prior to that, she was the chief product officer for Qualtric\u2019s EmployeeXM business, where she oversaw the product management, product marketing, and product science functions.<\/em><\/p>\n<\/p><\/div>\nHumans vs. Agents \u2013 What\u2019s the Difference?<\/strong><\/h3>\n
What an AI-First Site Looks Like<\/strong><\/h3>\n
Behavioral Data Isn\u2019t Just for Human Users<\/strong><\/h3>\n
\n