©champpixs via Canva.com
Will Third-Party Apps Kickstart an AI-Powered Retail Revolution?
The power of AI is spreading rapidly throughout the retail solution provider ecosystem, potentially democratizing a technology that has so far been limited to the most tech-savvy companies and platforms. Klarna is the latest company to launch advanced third-party AI solutions, such as AI-powered shopping assistants, in a trend that could help retailers create more convenient experiences.
Klarna is updating its app with new features like the shopping lens, which allows customers to take pictures of items and use that information to search for and compare similar products. The AI turns the image into a search term, which is used to find the best deals available on the Klarna app.
The in-app camera feature can “visually identify over 10 million items” and match them with more than 50 million online offers. The tool builds on Klarna’s existing discovery feed, which already makes recommendations based on personal interests.
Klarna is also enhancing its discovery feed with shoppable videos, ranging from unboxings to tutorials, that let customers make purchases from within the player. The solution provider is aiming to capitalize on growing view time for its video offerings, which are up 60%, as well as rising click-through rates, up 25%.
Klarna users can access product information as well by scanning an item’s barcode, including customer reviews and whether there are variants available for sale online. These solutions were created in response to a May 2023 poll that found eight out of 10 shoppers look forward to having an AI shopping assistant.
“Just like the internet gave everyone access to information, AI gives everyone access to intelligence, context and personalization,” said Sebastian Siemiatkowski, CEO and co-founder of Klarna, in a press release.
Klarna isn’t the only solution provider rolling out new AI-powered tools this year. In April, Mercari introduced a ChatGPT-powered shopping assistant capable of helping customers narrow down their top choices in crowded marketplaces. Salesforce has also been adding generative AI-powered tools, with an emphasis on improving the customer experience by empowering retail workers on the backend.
Major tech companies like Google were the first to integrate the potential of generative AI into their search functions, but now retailers of all sizes can access similar capabilities through third-party solution providers. This could mark the start of an AI-powered revolution that defines customer expectations in the coming years.
Discussion Questions
How could the spread of generative AI-powered tools change the face of e-commerce? Do you expect smaller companies to take advantage of AI-powered tools on third-party search apps?
Poll
Loading …BrainTrust
David Naumann
Marketing Strategy Lead - Retail, Travel & Distribution, Verizon
Nicola Kinsella
SVP Global Marketing, Fluent Commerce
It’s interesting that both examples relate to search and the discovery of products. This, along with recommendations, is where I think AI will have the greatest initial impact in retail. There is so much stuff out there and it takes time for consumers to sift through it all and find the right option. AI can do most of the leg work. That said, it remains to be seen whether AI has inherent biases or is fully effective at satisfying customers.
If these AI apps are truly easily integrated with existing apps, it has the power to revolutionize the shopping experience. Making the search for products more efficient and accurate will make retail apps delight customers and curate a loyal fan base. If the AI apps are reasonably priced for small retailers and are easy to integrate, it will help independent retailers compete more effectively against the large chains.
This trend takes the burden of developing skills in-house in AI, and enables businesses to use third-party solutions based on AI capabilities out of the gate. The remaining question will be – as with all other third-party solutions – who does it better?
Not only is the jury still out on this, but the trial is still happening. While AI seems inevitable in every space, consumers will be late adopters. Most businesses have yet to figure out how and where to implement AI so that it is driving key initiatives vs just enhancing efforts. Until organizations fully understand how to leverage AI it won’t matter if they implement 3rd party tools or build on their own tech stacks.
Omnichannel retail was made for AI, or vice versa. The millions of SKUs, reviews, transactional data, etcetera, can now be leveraged instantly—and for endless use cases. The key takeaway is that customers are excited about AI as a co-pilot that helps them get what they want, when they want it, and at the best possible price. As an example, User Generated Content (UGC) coupled with AI is one very powerful use case that is already in place today.
Retailers of the world, start your engines. The race to fold AI into every aspect of your business has already begun. In fact, I believe that every retailer should be leveraging AI technology in some way before the end of 2024. With this technology available to the masses, it will spread quickly while, for shoppers, fade into the background as part of the behind-the-scenes magic.
The principal challenge facing most retail IT shops during the last decade has been what to do with all the data being collected through internal and external sources. AI will certainly help address this issue. Now all the CIOs will need to figure out is how to create queries that actually deliver the actionable insights they need to stay competitive and innovative.
Have we finally discovered THE killer app? Nope. GAI is undeniably a technology with great uses and we will see it change the face of many things somewhere between where they are now and the astronomical levels predicted by the GAI hype. However, even if we assume the face of ecommerce is search, I don’t believe GAI is going to account for massive increases in new ecommerce purchases. And if the technology will be powered by third parties, like the internet, then do we think that maybe we’re starting a new GAI-SO revolution (GAI Search Optimization)? It’s difficult to separate hype from real, problem solving solutions so I think the jury on “default use” of the tech by all businesses is still out. I think if you have a specific problem or challenge that can be addressed by this technology then it’s worth pursuing but if you’re a smaller company, be careful not to invest in “a solution looking for a problem” (aka FOMO Investment).
Generative AI tools will increase productivity across a lot of different areas. Because they can replace humans? No. But because they can accelerate their output. Use cases like search and imagery for merchandising are a great place to start. Because the vendors that provide those solutions have a lot of data they can tap to train and test their models. And because the vendors can provide the solutions based on the data they capture, yes, I believe smaller companies will be able to take advantage of these AI-powered tools.
However, for more ‘back end’ use cases, ones that require a retailer’s own data, a different approach will be required, starting with more traditional predictive AI models, then layering generative AI on top of them. It’s not easy.
The first step is to get a good clean data set that can be used to train and test a model – which is very difficult when your demand, order, shipment, and inventory availability data often sits in different systems, in different formats. Then you need a way to indicate or extract the data signals that are most important for the problem you want to solve – whether it’s demand planning, short term demand sensing, sourcing optimization, etc. Then you can use that data to train and test a model. You also need a way to incorporate the model’s output into your workflows.
Only once you have good predictive AI working will you be able to layer generative AI (a large language model) on top so you can query the data with natural language queries. Some retailers have already started on this journey, others are working on getting their data ready. It’s early days.
The ability to turn a picture into a search query is going to be loved and embraced by shoppers everywhere. Especially if the search can then serve up good/better/best options for the shopper to peruse. AI is going to provide better results and save time in the process.
These apps, as described, continue to reduce the shopping decision to price at the expense of everything else. As such, it will be harder and harder to differentiate on the shopping experience, customer service, etc. Difficult but not impossible.
The art of product discovery is one of the more intrinsic components of the retail shopping experience. Personalization at scale is one of the most challenging endeavors retailers and brands will encounter in the digital world of commerce. Consumers will freely share their personal information with retailers and brands in exchange for compelling offers and great experiences.
We have heard about the AI and GenAI hype cycle and how it will transform our lives. As retailers seek the capabilities and solutions to provide consumers personalized experiences, we should expect plenty of third-party solution providers to step up and help enhance product discovery, improve loyalty programs, and even assist with dynamic pricing strategies.
This is all unproven territory. We should expect companies to take a crawl, walk, run approach to integrating AI and GenAI into their operating models.
I do see this as a key component of the future of retail. I like thinking about this as “access to intelligence.”
I’m skeptical of the new shiny bauble. Many apps use the idea of AI simply to polish their shiny-ness and without serious application.
Mostly, though, while we are all impressed at what predicting the next word can do (the true LLM AI trick), we need to take care in applying these apps. There is an inherent skew built into them because of the data on which they rely.
No. They will not be a revolutionary advantage. Small players should compete as they always have – find something significant to offer customers and do it well.
The key is to get more customers to use a brand’s app. I emphasize the word more because there are currently customers who love apps and those who stay away from them. Early in app development, apps were clunky and hard to use – basically, not user-friendly. Today, they are much better, but brands have to overcome the reputation from these past experiences. AI-powered apps that create better experiences can do that. Brands just need to get their customers to try them out.
Search and speed are the frost steps in the path to purchase and are often overlooked and under-prioritized functions. AI apps, if cost-effective and seamlessly integrated, will help brands engage.
Back-office functions are also often under-prioritized (compared to customer-facing functions) and can be significantly optimized by AI as well.
The digital era of retail/eCommerce has been one where “curation” has in many cases become a thing of the past. AI tools that help consumers zero in on what they want in a world of endless choice and thousands of websites and at the same time promote “discovery”” will be a big part of the future commerce landscape.
These GenAI search and discovery solutions will definitely add some convenience and a few minutes saved in making purchases -but- I am doubtful that these innovations will be revolutionary. It would be interesting to see really how much more time consumers save and how these tools increase conversion rate, reduce returns, and reduce basket abandonment.
Many consumers are moving to more experiences and less consumption. If AI search can search retail based on experiential attributes within the site or the availability of a cafe or special service alongside their product search, I think that would be more revolutionary than product discovery on its own because it solves the total customer experience question.
GenAI and predictive AI as well will become an embedded technology, and eventually becoming invisible and mainstream to the average consumer. Many won’t even realize they’re using AI enhanced responses to their shopping needs. Ecommerce in particular will have dozens of not hundreds of use cases, but most will become indistinguishable from the pages they see today, only that it will seem easier and faster and less encumbering.
3rd party apps will also have their day, but it will be short-lived as the AI revolution has kick-started without them. AI will be embedded everywhere and will become part of the coding and delivery structure for all apps, big and small. There will be embedding into web pages as well as apps, but the most important places will be in back office capabilities where the apps will be accessing information and data- hence the apps will not really be the change agent.
It is refreshing to see technology increasingly bringing new ways to shop in retail and expqnding to more mid-sized and smaller players. I’m a big champion of all boats rise. A consistently progressive experience across all retailers and channels can only be good for the industry as a whole.
Indeed, AI has the potential to transform the world of shopping as we know it but that kind of transformation would require accessible technology at scale. This seems like a step in that direction.
AI-first technology is already revolutionizing the eCommerce experience, especially product discovery. Though specific applications for generative AI are less concrete and still in more experimental phases, we’d be remiss to think AI has not already had an impact on improving both the customer experience and human output capabilities for the teams behind these digital experiences.
This does, however, demonstrate the importance of connected commerce and how third-party solutions make AI technology more accessible for all retailers, removing the need for them to build a system from the ground up. If brands want to keep up with the ever-evolving world of retail and eCommerce, they’ll need to look toward third-party integrations like this to stay in the game.
3rd party apps are well positioned to take advantage of the initial application of AI at retail, and potentially set the standard, providing significant potential for tech firms in AI. The driver being the integration into existing apps, tools, sites etc. Having domain knowledge within the AI tech firm will also provide a good advantage because it relieves the retailer of the burden of understanding how they can utilize what IA can bring forth.
AI ticks the box on categories that continue to generate demand across the retail ecosystem: convenience and personalization. Consumers have grown accustomed to shopping online and receiving personalized recommendations, and now they want online experiences to carry over to physical stores (and vice versa – a seamless and personalized experience). Things like AI, QR codes, and alternative payment options are all changing the ways in which consumers shop, so yes, AI certainly has the power to continue to transform e-commerce – and retail in general. For small businesses, the accessibility of AI through third-party search apps could level the playing field with larger retailers. They can use these tools to improve search functionalities, enhance customer interactions, and gain valuable insights into consumer behavior. While the integration of AI may pose initial challenges related to cost, adoption, and scale, the long-term benefits such as increased competitiveness will encourage smaller companies to embrace AI.
In my opinion, the rise of generative AI-powered tools in e-commerce is truly innovative. Smaller companies now have the chance to level the playing field, offering personalized and convenient experiences previously reserved for tech giants. With solutions like Klarna’s AI shopping assistant, even businesses with limited resources can enhance customer interactions, making it easier to find products and get tailored recommendations. This shift democratizes advanced technology, empowering smaller retailers to meet evolving customer expectations. As these tools become more accessible through third-party apps, I anticipate a widespread adoption among smaller companies eager to provide smarter, more efficient services to their customers.