The B2B shortlist now forms before a seller hears about the deal. SAP’s AI in Customer Experience Report 2027, which surveyed 4,800 consumers, business buyers, CX practitioners and decision-makers in the US and Germany, found that nearly all of the business buyers it surveyed have used AI for purchasing or vendor selection. The same week, G2 and TrustRadius each rebuilt their pitch to software vendors around the AI answer that comes before the sales call.
SAP published the findings on Nov. 12 in a feature by Jessica Keehn, chief marketing officer for SAP Customer Experience. Her argument is that the buying journey has split in two. Discovery, research, comparison and shortlisting increasingly happen inside AI tools. The purchase, fulfilment and service still fall to the business, and in B2B that second half means negotiated prices, contract terms, approvals and implementation plans.
“An AI can identify the right supplier in minutes. Doing business with that supplier still takes considerably longer,” Keehn wrote. SAP calls the result the hybrid AI buyer, one who moves between AI tools and the vendor’s own channels and expects the vendor to be ready on either side.
Advertisement
Three releases, one week, the same premise
SAP was the third vendor in three days to make the case that AI now sits in front of the vendor’s own channels.
G2 sells evidence for the answer engine
On Nov. 10, G2 announced a unified platform across G2, Capterra, Software Advice and GetApp, which it acquired earlier this year. G2 says the combined properties hold more than 6 million reviews across 2,000+ categories and reach 200 million annual buyers. It also introduced answer engine marketing, which it defines as “the discipline of using paid tactics to influence which brands appear in AI-generated answers.” Alexis Zheng, Chief Product and Technology Officer at G2, put the premise plainly: “AI is the curtain, and all that gets through is the evidence.” We covered the ad and intent side of the release when G2 unified ads and buyer intent signals across four review sites.
TrustRadius says its reviews are written for the model as well as the buyer
The same day, TrustRadius, an HG Insights company, named 276 products from 146 companies to its 2027 Buyer’s Choice Awards. Across 9,693 verified reviews, 92% of reviewers gave the vendor high marks on the relationship, which TrustRadius describes as running “from the sales process through implementation.” The company says it has optimised its product pages for LLM crawlers so that verified customer content reaches buyers wherever they research. “Buyers don’t take a vendor’s word for it anymore, and neither do the AI tools they turn to for advice,” said Scott Gordon, CMO at HG Insights.
Invoca sees the same buyer arrive in consumer markets
Invoca’s Broken Funnel Report, also released Nov. 10, covers B2C rather than B2B, but the pattern repeats. Of the 451 US B2C marketing leaders it surveyed, 86% said they can identify leads who started their research in an AI answer engine. Most could not follow those leads to a sale.
Where the shortlist now gets made
Each of the three vendors sells something that benefits if this premise holds: SAP sells catalog and commerce agents, G2 sells visibility in the answers, and TrustRadius sells verified review content. Read together, though, the releases describe one change in where a vendor gets ruled in or out.
In SAP’s framing, by the time a buyer reaches the vendor “the shortlist has already been made.” The inputs to that shortlist are third-party: review content, product attributes, pricing pages, specification sheets and availability signals. SAP says every attribute, image, specification and proof point now shapes whether a product gets found and shortlisted, by people and by machines. G2 and TrustRadius are betting that their review data is what the answer engines weigh most heavily.
The second half of SAP’s finding matters as much as the first. AI takes the lead in discovery and evaluation, SAP writes, and moves to a support role in the later stages, as business context, process and relationships take over. The work a sales team does after the first call stays a human job in this account, and it starts from a shortlist the seller did not shape.
SAP’s data also shows how unready the systems behind that second half are. Ninety-three percent of US CX decision-makers in the survey said they feel prepared for AI-enabled customers. Six percent said all relevant systems are connected end to end.
What it means for the sales leader
In our reading, the sales leader’s job in this split is to make the second half fast and to feed the first half with evidence the team can stand behind. Four practical moves follow from the three releases.
Treat review content as pipeline input
TrustRadius’s awards turn on three questions: does the product deliver, is it worth the price, and is the vendor a good partner. Those questions are being answered by customers and read by AI tools before discovery calls start. Sales and customer success leaders who own the relationship score should know what their own customers have written about the sales process and the implementation, because that text now travels into buyer research.
Ask new opportunities where the shortlist came from
Discovery calls should record whether the buyer arrived with an AI-generated comparison and what it said. That field tells marketing which evidence is reaching buyers and tells sales which objections are already formed. Our October feature on how buyers now arrive with AI answers set out the evidence sellers need to answer them.
Shorten the second half
SAP’s point that doing business with a supplier still takes considerably longer is an instruction to sales operations. Quote turnaround, contract redlines, approvals and implementation scoping are the stages where the human team decides whether a shortlisted vendor wins. Self-service quoting is one route: Ericsson moved contract quoting to SAP Commerce Cloud self-service in October.
Check the six percent problem before buying more AI
If only six percent of US CX decision-makers say their systems connect end to end, the likely failure point is the handoff between marketing data, the CRM and the systems that price, contract and deliver. A sales leader planning 2027 spend can test each tool on whether it reads and writes to the record the next team uses.
Source: SAP News Center
