Social Rocket
People who combine search, social media and generative AI when shopping plan significantly higher back-to-school spending than people who don't use these digital tools, according to a recent Deloitte survey. For social media marketers, this is less proof that AI automatically generates more revenue and more a sign of how product search, inspiration and price comparison are now spread across several digital touchpoints.
The figures come from a US survey of parents with school-age children. They cannot be transferred one-to-one to Germany, but they show a pattern that is also relevant for brands here: social content, classic search and AI tools are increasingly used in parallel.
737 dollars instead of 381 dollars per child
In May 2026, Deloitte surveyed a total of 1.207 US parents of children in grades K–12. On average, respondents planned 557 US dollars per child for the back-to-school season.
The breakdown by digital usage is particularly striking:
- Buyers without the tools considered: 381 dollars per child
- Users of search engines: 494 dollars
- Users of search and social media: 531 dollars
- Users of search, social media and generative AI: 737 dollars
According to Deloitte, 29 percent of respondents belonged to the group that uses all three tool categories. Another 21 percent combined search and social media, 30 percent mainly used search and 20 percent used none of the digital tools considered.
The classification matters: the data shows a correlation, not causation. Particularly digitally active shoppers may simply be more involved in purchase planning, research more and already plan larger baskets.
AI is mainly used for prices, deals and research
A second recent US analysis by Epsilon reaches a similar picture. Almost half of the surveyed parents said they used AI for back-to-school shopping. Among AI users, ChatGPT was the most frequently named tool, at 64 percent.
The main tasks are very practical: comparing prices, finding discounts and offers, and gathering product information. GenAI is therefore not only sitting at the top of the funnel as an inspiration engine, but increasingly right in the middle of research and purchase decisions.
For brands, this means product information must be correct not only on their own website. Prices, availability, product features, reviews and clear benefit arguments should be consistent and easy to find across several digital surfaces.
Social media becomes part of a connected product search
Social content plays a dual role in this mix. Reels, TikToks, creator posts and pins can make products discoverable, but at the same time they provide signals and content that can later feed into search, recommendations or AI-supported research.
This does not make classic audience work superfluous, but more demanding. Our guide on data-driven discovery of social media audiences explains why platform analytics, behaviour and competitor data should be considered together.
Creator content also gains an additional function as a result. Alongside direct reach, specific, quotable content can appear in search and AI contexts over the long term. Our article on creator marketing for AI search covers how briefings are changing for this.
Value messaging remains central despite AI
The technology does not displace the classic focus on price and benefits. Deloitte describes 2026 households that remain strongly price-conscious; Epsilon likewise sees offers, deals and price comparison as central AI use cases.
A combination is therefore sensible for campaigns: social content creates attention and context, while concrete product information, price arguments and offer communication support the subsequent research. Especially for consumer goods and frequently purchased products, this approach can be well linked to the platform roles in our social media playbook for FMCG brands.
For German marketers, the most important caveat remains: the underlying studies measure a seasonal US audience, not the German population as a whole. The dollar figures are therefore not suitable as a forecast for specific German spending. As a behavioural trend, however, they are interesting because they show how little linear digital product search now works.