I expect many revenue teams to spend part of their 2027 budget on AI that writes quotes faster. I think the first money should go to what happens after the customer signs, because a faster quote sends more deals, built in more ways, into a chain of handoffs that most companies still run by hand.
A vendor’s prediction worth taking seriously
Christopher Shutts, Group Vice President of CRM Sales & CPQ at ServiceNow, put the case plainly in his predictions for 2027, sent to SalesTech Edition by ServiceNow’s product public relations team. His heading reads: “In 2027, the biggest sales bottleneck will come after the deal closes.”
“For too long, B2B organizations have treated quoting, contracting, fulfillment, billing, and renewals as separate processes, creating delays and disconnects between what sales promises and what the business ultimately delivers,” Shutts said. He expects AI agents to connect those stages, and he warns: “Companies that automate the front end of sales without modernizing downstream processes risk creating new bottlenecks, billing errors, and delays that undermine the customer relationship.”
Advertisement
He has a product interest in that view. In an October 20 post on ServiceNow’s site, the company says it was named a Leader and a Customer Favorite in The Forrester Wave for configure, price, quote solutions in the fourth quarter of 2026, and describes specialized AI agents that turn meeting transcripts and voice notes into draft quotes and pass orders on to fulfillment. A vendor that sells the connected chain will argue for the connected chain. I still think he is right, and the reason sits in any company’s last 20 closed deals.
Why the handoffs break first
After signature, a deal usually passes through four or five owners: a contracts team, order entry, provisioning or delivery, billing, and later the account team that renews it. Each one works in its own system and re-keys some of what the quote said. That arrangement copes with a steady flow of deals that look alike, because each team learns the usual shape. Some of those steps are already being automated one at a time, as when contract software moved from a tool to an executor, which speeds up one owner without joining it to the next.
AI quoting changes the flow. It produces more quotes and more variety: an added upsell line, a bundle price, a rebate linked at the moment of quoting. Conga’s October 2026 release shows vendors know where that lands. It links pricing, incentives and agreements, and says the result is “less reliance on manual calculations, disconnected systems, and downstream reconciliation.” Downstream reconciliation is the work I mean.
The far end of the chain is getting agents too. Visa, BILL and Billtrust are piloting an agent that pays approved invoices inside supplier portals, with “Payment controls tied to the customer’s approved payment instruction,” as their announcement puts it. When both ends run at machine speed, the manual middle is where the wait builds up.
The strongest objection
The case against my view is speed. Buyers reward the first accurate proposal, a fast quote wins deals that a slow one loses, and finance teams have always absorbed some billing clean-up after the sale. On that reading, a sales leader should buy the quoting agent now and let operations catch up.
My answer is that errors after signature land on the customer. A wrong invoice, a missed delivery date or a product configured differently from the quote reaches the buyer’s finance team and the user who waited for it, and the account team then opens the renewal conversation with that history. We have written about how often nobody owns the contracts that drive renewals. A faster front end makes that ownership gap more expensive, because it feeds it more deals.
What I would do first
Take your last 20 closed deals and trace each one from signature to first invoice. Count the times someone re-entered data, the days each handoff took, and the fields on the quote that have no matching field in the order or billing system. That map is the case for or against a quoting agent, and it costs a week of one analyst’s time.
When you evaluate CPQ vendors, ask each to show an order and an invoice created from a quote its own AI built, with no re-keying. Any vendor that can only demo the quote is selling you the part of the problem you already handle.
Then give one person in revenue operations ownership of the chain from quote to renewal, with the authority to change fields across systems. The quoting agent can come after that, and it will have somewhere to send its work.
Source: ServiceNow on The Forrester Wave for CPQ, Q4 2026 (ServiceNow)
