Commerce Tomorrow

A few genuine data points anchor the episode (GMV growth, deployment speed) but the bulk is nostalgic farewell conversation, surface-level AI commentary, and a rapid-fire prediction review that trades depth for breadth. Insight per minute is low.

“Last year we had been north of 30 billion, um US dollar, um, that our customers or we had enabled for our customers, uh through the Commerce Tools platform. The year isn't fully done yet, but we're getting to the end. But I expect we'll be somewhere near um, 40, 42 billion.”

“Only Shopify and us have reported Black Friday Cyber Monday numbers. Nobody else did anymore.”

Nearly every macro observation leans on recycled industry frameworks: Gartner hype cycle, the 'overestimate short term / underestimate long term' cliché, Paul Graham's Default Alive post, and the self-serving two-horse-race framing. Very little that a commerce industry follower would not already know.

“I don't remember who it was who said we always overestimate the impact of technology in the short run and underestimate it in the long run.”

“there's a famous VC named Paul Graham. He, he founded Y Combinator and he has this seminal blog post called Default Alive or Default Dead.”

Dirk is a genuine operator who built Commerce Tools to scale and shares real numbers, giving him practitioner credibility. Kelly has category-building experience but is primarily a marketer. There are no external guests at all, limiting range and challenge significantly.

“I'm long enough now since 25 years in this market. I know that if you have good numbers you share them and only if you have bad numbers you don't share them.”

“I brought Andrew in uh who started now it's liter now half a year back. It was uh 1st of July that Andrew came on board uh and took over my role um as CEO.”

The episode earns credit for named customers with concrete metrics (John Lewis update frequency, Shields sub-five-month go-live, GMV $30B to ~$42B, 30%+ Black Friday API growth, $28M funding round for a named startup). However many predictions are reviewed in two sentences with no supporting data.

“I know they became a customer last summer, already went live before Black Friday and Cyber Monday and had an amazing week out there. So it's uh, uh, less than four or five months on full, flexible, composable kind of solution.”

“John Andrews who's a friend of the show, um, just got 28 million in funding for an AI commerce search startup.”

This is an entirely internal co-host farewell chat with no external guests, no probing follow-ups, and no pushback on any claim. The 'rapid fire' pivot abandons even the shallow depth the first half offered, and questions are consistently softball prompts.

“Good. So, um, we have spent 20 minutes on six of these and it's great. But we also have 20 minutes left and we have to cover uh, all 23. So let's do rapid fire here towards the end.”

“Any, any customers that come to your mind. Um, when look on 24.”