Social media used to be the place where brands created demand. Search was where people captured it. AI is making that split harder to defend.
New data cited points to a purchase journey spread across more touchpoints, with social platforms generating discovery and AI tools increasingly helping people research, compare, and decide. The important shift is not that marketers suddenly need to be everywhere. It is that each channel is beginning to do a different job.
Discovery is becoming a shared responsibility
The figures in the 2027 Global Digital Report, as summarized by Social Media Today, show how quickly the behavior is moving. Forty-two percent of online adults use social media to research purchases, while 30.4% discover new brands through social platforms. AI is already part of the same process: 26% use it for brand research and 17% discover brands through large language model platforms such as ChatGPT.
That makes social content more than a stream of impressions. A post, review, creator recommendation, or community discussion can become part of the information an AI system draws on when someone later asks what to buy. The path is no longer social post, click, and checkout. It can be social post, AI question, comparison, search, and purchase, with different platforms influencing different moments.
The report also says online adults now use an average of 3.8 channels when researching brands, up from 3.65 two years earlier. That is a small numerical increase with a large planning consequence. A brand cannot assume that the platform receiving the final click created the demand, or that the platform creating awareness will also close the sale.
WeRSM has previously examined how AI is turning brand visibility into a moving target. This data adds a behavioral layer to that problem: visibility is not only about appearing in an answer. It is also about creating the material that makes an answer more likely to be useful, specific, and favorable.
The media mix now needs a job description
The practical answer is not to spread budgets evenly across every major platform. It is to understand what each environment is good at and where audiences actually spend time.
The reported platform differences are striking. Instagram is the favorite social platform for 17.8% of online adults. TikTok has the highest average daily time per user at one hour and 55 minutes. WhatsApp leads in average daily sessions at 20.7. Local patterns vary even more: X leads daily usage time in Japan, Threads leads monthly usage in Taiwan, Pinterest has the highest advertising reach in the Netherlands, and Reddit leads average daily sessions in Ireland.
Those numbers are not a universal budget formula. They are a reminder that “social media” is too broad a category to guide investment on its own. A platform can be valuable because it holds attention, drives conversation, supplies cultural context, reaches a specific market, or helps answer a product question. Those are different jobs.
The same logic applies to AI visibility. Reddit may matter for product comparisons. TikTok may influence trend and recommendation prompts. LinkedIn may carry more weight when an AI system answers questions about a company’s expertise. The right question is not which platform wins overall. It is which platform contributes evidence at the moment a person is trying to decide.
Measurement has to catch up. The source argues that last-touch attribution misses too much of this movement, while multichannel measurement can show how interactions reinforce one another. That does not mean every post deserves credit. It means teams need to distinguish between a channel that creates attention, one that builds confidence, and one that converts existing intent.
AI is not replacing the media mix. It is exposing how incomplete the old version was. Social platforms increasingly shape what people discover and what information systems can later retrieve. The strategic consequence is simple: brands need to plan for the evidence they create across channels, not just the ads they buy inside them.