Gen Z and Millennials, the younger generations that dominate social media usage, have been found to trust recommendations from friends and family more than they trust influencers. This is despite the fact that these younger individuals heavily rely on social media for researching brands and products. According to data from 2019, the number of Gen Z using social media to research brands and products increased by nearly 40 percent between 2015 and 2019. However, when asked what inspired them to make a purchase in the last month, 48 percent cited a discount on a product, followed by recommendations from friends and family at 39 percent. This shows that despite the influence of social media, personal connections still hold a strong sway over consumer decisions.

Gen Z and Millennials, the younger generations that dominate social media usage, have been found to trust recommendations from friends and family more than they trust influencers. This is despite the fact that these younger individuals heavily rely on social media for researching brands and products. According to data from 2019, the number of Gen Z using social media to research brands and products increased by nearly 40 percent between 2015 and 2019. However, when asked what inspired them to make a purchase in the last month, 48 percent cited a discount on a product, followed by recommendations from friends and family at 39 percent. This shows that despite the influence of social media, personal connections still hold a strong sway over consumer decisions.

Interestingly, consumers tend to trust smaller influencers more than larger ones. In the US and UK, 56 percent of respondents believe that influencers with up to 50,000 followers are the most credible. This could be due to the perception that smaller influencers are more relatable and authentic compared to those with millions of followers. It seems that as influencers gain more fame and popularity, their recommendations become less trustworthy in the eyes of consumers.

Moving on to Pinterest, a visual discovery engine, we can see how computer vision plays a crucial role in exploring users' tastes and preferences. Pinterest's acquisition of Kosei, a machine learning startup specializing in recommender systems, in January 2015 was a turning point for the platform. With machine learning techniques, Pinterest can recommend interests to new users, generate engaging home-feeds, categorize and curate content, predict user engagement, and prioritize localized content. By analyzing users' pins, Pinterest can offer personalized recommendations and surface popular content from users' local regions.

Pinterest's computer vision capabilities go beyond just analyzing images. It also considers captions from previously pinned content and identifies items that are frequently pinned together, even if they don't visually resemble each other. This allows Pinterest to make connections between different items and provide users with related suggestions. In 2014, Pinterest acquired VisualGraph, an image-recognition startup, and established its computer vision team to focus on visual search. This led to the development of Instant Ideas, a feature that transforms users' home feeds by providing similar ideas with just a tap.

The architecture of Pinterest's visual search tool, Lens, combines image and object recognition with discovery technologies to offer users diverse results. The query understanding layer analyzes the input image and computes visual features such as object detection, color analysis, lighting conditions, and image quality. The blender component then utilizes visual search, object search, and image search technologies to return visually similar results, scenes with similar objects, and semantically relevant text search results, respectively. By blending these different sources, Lens provides users with results that go beyond just visually similar images, bridging the gap between real-world camera images and the Pinterest taste graph.

Pinterest conducted various experiments and A/B testing to improve its visual search capabilities. The ability to detect objects in images and match them with visually similar objects has opened up new possibilities for object-to-object matching and discovery experiences. Pinterest's machine learning algorithms learn from labeled bounding boxes for regions of interest in images, allowing them to understand which objects are of interest to users. Real-time object detection is essential for Pinterest's visual search engine, as it enables users to search for any image, including unseen content from the web and camera, in a fraction of a second.

According to Ben Silbermann, the CEO of Pinterest, computer vision plays a vital role in the company's goals. They aim to understand the aesthetic qualities of products and services to provide better recommendations. Additionally, they want to be able to zoom in on specific parts of an image and identify similar objects. Ultimately, Pinterest plans to create a camera tool that allows users to query the world around them. Computer vision serves as the fundamental technology that powers these objectives, helping people discover and engage with the things they love.

In conclusion, the younger generations, such as Gen Z and Millennials, trust recommendations from friends and family more than they trust influencers. Despite the significant increase in using social media for brand and product research, personal connections and discounts play a more significant role in their purchasing decisions. Pinterest, on the other hand, relies heavily on computer vision to explore users' tastes and preferences. By utilizing machine learning algorithms, Pinterest can offer personalized recommendations, categorize and curate content, and provide visually similar results. Three actionable advice that can be derived from these insights are:

1. Brands should focus on building strong relationships with their customers' friends and family members. Encouraging word-of-mouth recommendations and offering discounts can significantly impact purchasing decisions.

2. Influencers with smaller followings should be considered as potential partners for brand collaborations. Their authenticity and relatability can make their recommendations more trustworthy in the eyes of consumers.

3. Businesses should invest in computer vision and machine learning technologies to enhance their understanding of customers' preferences and provide personalized recommendations. This can lead to improved customer satisfaction and engagement.

By incorporating these pieces of advice into their strategies, brands and businesses can effectively navigate the preferences of younger generations and leverage the power of personal connections and technology to drive success.