Microsoft is reviewing different ways to implement conversational AI in Microsoft’s advertising platform. Kya Sainsbury-CarterMicrosoft’s corporate vice president of advertising told Greg Finn at SMX Advanced yesterday.
That’s something we learned yesterday from the conference, which explored how conversational AI could radically change search marketing.
From higher CTRs to increased ROI to conversational AI appearing in ads for the first time, Bing Chat has been specifically designed to address the universal problems experienced with traditional search, according to Sainsbury-Carter.
Here’s a recap of SMX Advanced’s extensive keynote interview.
Address traditional search problems
The new Bing, introduced in February, was designed to succeed where traditional search results can sometimes fail.
“The universal problem with traditional search is that billions of searches are missing the mark. Sometimes it can be difficult to find the information people are looking for, which is why we launched Bing Chat to provide better search results and answers to complex questions,” Sainsbury-Carter said.
Microsoft has worked to solve historical search problems by combining the power of large language models such as open AIs with its own search index.
“This is to ensure that we are delivering results that are current, that are cited and, as I said, that are conversational,” Sainsbury-Carter said.
The new Bing for advertisers: ‘so far, so good’
While it may be early days, Microsoft has confirmed that there have already been more than half a million Bing Chat conversations since its launch, with more than 70% approval for the responses it provides.
In addition, people have created over 200 million images with Bing Image Creator.
Early data also indicates good news for advertisers, as Microsoft’s conversational AI is driving a higher click-through rate than traditional search, which is subsequently leading to increased conversions.
“It’s because the user journey is shortened, so there’s a faster path to conversions. So far, so good, would be the summary here,” said Sainsbury-Carter.
Sainsbury-Carter did not disclose any specific numbers. But it revealed that the increase in CTR applies to both search results and ads, which is promising news for search marketers.
More good news for advertisers: Microsoft is working on more opportunities for advertisers to further increase user engagement, which could once again increase conversions.
Conversational AI is coming to Microsoft Ads
Sainsbury-Carter highlighted the company’s commitment to investing in the integration of chat AI, noting that Bing Chat is “stronger than ever”. She added:
“We’re working hard to understand how we can use chat to drive internal efficiency to make sure our people are working on the most valuable and complex work. So you know there’s the platform aspect and the ‘internal aspect, which we believe are very significant.’
Get the daily search newsletter marketers trust.
From keywords to topics
Microsoft has suggested that advertisers can expect traditional keyword targeting to shift to a topic-centric approach with conversational AI. Explaining the change, Sainsbury-Carter said:
“If I’m trying to find a dress for a wedding in Chicago in July and I type in a traditional search query, I’m likely to get results for dresses for weddings, and for Chicago. But if I put the same query in Chat, Chat can put together something that says, “OK, it’s hot and humid in July in this part of the country, so show me a flowy sleeveless dress in a lightweight fabric.” “What’s happening is that we’re expanding our understanding of intention. ..instead of matching answers to keywords and thus definitely changes the idea of keyword bidding. We believe this provides a great opportunity for ads to better match the conversation and result in better ROI.”
Rethink traditional ads
Sainsbury-Carter explained that Microsoft is taking a new approach to ads, user experience and ROI scenarios:
“We’re thinking more about monetization within the experience. So, for example, in chat mode, when you see sources cited, those citations will be a mix of algorithmic results, as you would see in traditional search, as well as some locations of ads”.
Visual ads
According to Display Ads, Display Ads are another important newer feature that should be of interest to marketers.
“The ads we serve in the new Bing Chat experience include both text ads that appear on hover, as well as more visual ads that serve in line with the conversation, such as a vertical display product ad.”
Microsoft is making display ads a huge priority and has already been testing this feature for hotel ads. It then plans to roll out this expertise to the travel and real estate sectors.
While the focus is currently on conversational AI, Sainsbury-Carter said this is not the end of traditional search:
“Traditional search isn’t going away, and there are good reasons to continue using it. That’s why we’re also focused on delivering more visual and immersive experiences, as well as a new vertical advertising experience you won’t find on any other platform. .”
Quality of ads over quantity
Bing Chat launched with ads, and the plan is to make sure they remain an integrated part of the chat experience, just as they are with traditional search. However, Microsoft is focusing on improving the advertising experience by ensuring that the ads served are relevant and meet user intent.
“It’s less about how many people see your ads and more about the questions people are asking in the new Bing, and how we answer those queries. That’s how we think about it, not so much in terms of numbers but in terms of matching queries,” added Sainsbury-Carter.
Great opportunities for small businesses
Conversational AI makes people query three times longer than traditional search averages, meaning they provide more detail. This additional information is helping bots to better understand user intent and therefore provide more detailed responses.
“We see a huge opportunity here for small businesses. They need that extra detail because they’re not looking to cover the whole world … so that opportunity to be more granular is key.”
Local small businesses looking to improve their SERP rankings have also been advised to ensure they have Bing Maps on their website. “It’s not needed for ads,” Sainsbury-Carter said. “But it does help!”
Drive traffic to news publishers
Microsoft has introduced a chat API, which will be made available to news publishers who want to implement chatbot technology on their websites.
Whether publishers source their AI chat experiences from Microsoft or not, Microsoft Advertising can help them monetize their services by serving relevant ads designed specifically for the chat experience.
“This solution gives publishers and app companies flexibility and control over the user experience, because no one knows their audience better than they do,” added Sainsbury-Carter. “They can choose the ad formats they think are best for their audience and integrate them in the way that feels most natural to their native experience. We’re very excited to bring this to market more broadly.”
Dig deeper: Microsoft’s vision for a thriving content ecosystem in a world of generative AI
Bing Chat isn’t perfect, but Microsoft is learning ‘very quickly’
Sainsbury-Carter acknowledged that security and privacy concerns have been a big topic of conversation for search marketers. He emphasized that addressing these issues has been a priority for Microsoft.
He noted that Bing Chat has been designed in collaboration with AI experts and is in line with Microsoft’s Principles of responsible AI.
“We made sure to leverage Microsoft’s experience to make sure we weren’t reinventing the wheel with this. And we’re continuing to learn lessons from real-world experiences and real-world testing,” Sainsbury-Carter said.
Microsoft has taken many steps to prevent abuse of its technology, including limiting the number of sequential chats per session and limiting the total number of chats per day.
However, while Bing Chat has been developed to Microsoft’s highest standards, Carter admitted that teething problems are expected.
“We didn’t put the product out there and say it’s perfect or you should use it in every instance of every interaction and it’s going to be perfect. We’re learning, but we’re learning very quickly so we can continue to evolve our approach.” .
[ad_2]
Source link
