Forgepoint says its AI sales rep Frank cut manufacturers' quote times by more than 55%

An estimator opens an RFQ that arrived as an email with several attachments and missing specifications ^[5]. Before a quote goes out, that person may have to contact the customer, find similar jobs, build a cost model and get engineering approval, then enter pricing in the ERP and update the CRM ^[6]. In the announcement as published by The Next Web, the steps that need expert judgment are few, and the handoffs and data entry around them stretch the process from hours into days or weeks ^[7].

Forgepoint pitches Frank as a new hire with an email address at the manufacturer's domain, trained on the shop's procedures, pricing rules and approval requirements ^[8]. The work described is narrower. Frank flags missing requirements and writes to the buyer ^[9], while engineers, estimators and salespeople keep the technical and commercial decisions ^[10]. Forgepoint configures each deployment around how the shop already quotes ^[11], so the 30% and 55% figures are averages across differently configured customers ^[3].

Jack Strauss, Forgepoint's CEO, framed the problem as administrative. He said: "We kept meeting manufacturers with strong demand and capable teams, but their best people were spending hours on administrative work." ^[12]

The company's headline number is seven-figure ARR within six months of launch ^[16]. Seven figures covers anything from $1 million to just under $10 million ^[19]. The reported figures do not include a customer count, a price or a retention rate. The nearest usage measure is the hundreds of people using Frank each week ^[17]. It counts users; customers and quotes would be separate numbers.

The evidence that the fastest reply wins comes from a vendor survey: Forgepoint's own research says 78% of companies buy from the first supplier that responds ^[18]. A response is also not a finished quote. Frank's clarification email to a buyer can be the first reply ^[9] while the price waits on engineering.

In our view the evidence shows a product with weekly users and some revenue. A category needs adoption and renewal across more than one vendor's customers, and this is one vendor's first six months.

For a shop deciding on Monday, two measurements sort the pitch. The first is quote time today. Take the reported 55% cut ^[4] and the 24-hour response mark from the Gardner Intelligence study ^[15]. Treating the quote as the response, a shop that now takes more than about 53 hours stays over 24 after the cut ^[13].

The second is where those hours sit. Data entry and chasing missing specifications are the work Frank takes on ^[7][9]. A wait for engineering sign-off sits with people ^[10]. A shop under 53 hours with its delay in the administrative steps fits the case Forgepoint describes. A shop over 53 hours with its delay in engineering review sits furthest from it. The other two combinations need a pilot on the shop's own jobs to show which number moves.