- Amazon’s new Rufus app combines editorial, content and commerce with AI.
- Our new report on how brands are using AI for content generation
- The era of ‘hustle’ is here
Amazon’s ‘Rufus’ app combines content and commerce with AI
Amazon launched Rufus, an “expert shopping assistant” that has been trained on the e-commerce juggernaut’s product catalog and information across the Internet. Rufus, which launched in beta, will help shoppers by answering questions, delivering comparisons, and offering recommendations. For example, a shopper could start by .asking “what to consider when buying shoes,” and Rufus can deliver recommendations and suggestions to help narrow down their options. Rufus also helps shoppers who have questions about specific products.
Rufus can also help conduct general research or help shop by occasion or purpose, but “it’s still early days for generative AI, and the technology won’t always get it exactly right,” the company wrote in a post announcing the new project.
Why this matters: Editorial and content sits at the heart of Rufus and the future of conversational shopping. The project’s leadership includes Natalie Zmuda, who formerly ran content at Google and now operates as “editor-in-chief” of Rufus and the company’s forays into what it’s calling “conversational shopping.” Conversational commerce has generally been used to refer to the growing practice of businesses speaking to customers in real time – sometimes through humans, and now, through generative AI bots. But how these communications take place and how closely they adhere to a company’s overall brand is a content issue. “I took the job because I was energized by the vision the team had — a vision to revolutionize shopping.” Zmuda said in a post. Walmart’s announcement two weeks ago of a search function powered by generative AI was a similar move: The experience a customer has on a website is now no longer simply transactional.
How brands are using AI for content generation
The rise of generative artificial intelligence presents a dichotomy for brand publishers. On one hand, AI threatens the very foundations of modern online publishing as it trains itself on publicly available content, and AI-driven experiences are baked into search results and other major technology platforms.
On the other hand, it promises to add significant efficiency to the content production process – helping with everything from ideation to editing and content distribution. In this Snapshot, we’ll look at how marketers are currently using AI to their advantage, unpack some of the top challenges and obstacles, and explore where brands hope AI-driven content is headed next.
A new era
We are standing on the precipice of a seismic shift in how content and communications will operate, senior media and content professionals say.
Content will increasingly sit at the heart of companies’ operations, they argue, but it will also become extremely difficult to make content stand out.
Several factors are driving this feeling: The end of organic distribution, thanks to continued link throttling by major platforms, as well as the rise of generative AI and its potentially cataclysmic effect on search traffic are major stressors. Audiences have become smart and tired: They’ve been inundated with content for so long, that they’re hungry for something new. Media brands are shutting down, and the ensuing bloodbath of journalism layoffs foretells difficult times ahead. For many, it feels like the most difficult time ever to be in media.
At the same time, content is becoming the primary way companies communicate to consumers, employees and stakeholders. Consumers are tired of being advertised to, and for all the talk of performance marketing, narratives and stories now appear poised to be a key way brands can exist in the public sphere. The continued chaos that marks the current era means that crisis communications through content is now a C-suite priority. The era of “going direct” may also get more complicated. As Sean Blanda of Crossbeam writes: “In a world where institutions are distrusted and B2B buyers are savvy enough to see a “go direct” playbook being run from a mile away, it will be harder than ever to win the trust of an audience.” In response, more companies are considering carefully how they can improve the quality and originality of what they produce – and that doesn’t come easy. As the Economist writes, who did the posting will soon matter more than what was posted.
What comes next: For anyone in the content business, there’s a feeling that it’s time to show results. How those results actually are shown may vary: Some brands are trying to figure out how to monetize content directly. Others are putting in place more sophisticated tracking techniques so they can get appropriate credit for their work. And many are figuring out how they can use content to gain valuable first-party data on their customers as the demise of the cookie gets underway. As one tech brand editor said: “Time to hustle.”
Humans prefer AI content to that generated by humans – with one major caveat
A new study pitted human-created content vs. AI-generated content in six “battles.” In each, AI won. SEMRush conducted the study, which compared content including social media ads, blog posts, social media posts, product descriptions, and article copy. The company said that for article copy, the AI had to be prompted multiple times to produce a refined result.
“These results show that AI-written content can be effective and resonate with your customers. If you prompt your AI tools well enough, you can create engaging and high-quality marketing copy. However, our experiment does not suggest that AI is enough for content writing,” the study found.
The findings align with an MIT study from a few months ago, which found that for certain types of content, AI was able to generate more engaging copy than humans, indicating that AI could be valuable for rote-type content-related tasks such as descriptions, captions, or summaries without loss in quality.
Job news
Hubspot’s media team is now looking for a marketer to grow audiences for Hubspot Original Podcasts. That set of podcasts includes “My First Million,” The Hustle Daily, and Marketing against the Grain. It’s the latest example of brand publishers increasingly focusing on hiring audience development and growth talent as the era of organic distribution wanes.
Localization works
Creating content in other languages is a relatively rare occurrence for brand publishers. But in a case study on LinkedIn, Campfire Labs CEO Cassandra Naji says that it can help catapult audience growth. Naji compared Asana and its competitor, Smartsheet: Both brands had the same organic traffic three years ago. Today, Asana has 4 million monthly visitors more. Adding content in languages including Spanish, German and French helped the company grow visitors: “The content Asana wrote in these 5 languages accounted for about 70% of the company’s traffic growth over the last three years,” said Naji. For brands operating outside the U.S. – and particularly those experiencing growth slow down – local languages may be the key, said Naji.