"We're rebuilding the innovation engine at 3M," CEO Bill Brown told investors in July. Then he showed them the operating data behind the claim.
Those 92 launches came during the second quarter, up 44% from the same period last year. First half total: 176. That puts 3M on pace for more than 350 launches in 2026, nearly triple what it introduced three years ago.
Look at what's coming out. Optical connections for AI data centers. High performance fibers for fuel cells. Reflective films built for satellites. Ask 3M, a digital assistant that helps engineers find and compare materials while they work.
The count covers the whole company, including the consumer side you know from Scotch-Brite, Command, Filtrete, Scotch, and Post-it. 3M didn't split the 92 by segment. The examples it chose to highlight came from its industrial, safety, transportation, and electronics businesses.
3M calls the system behind all this its "R&D factory." A product now takes around 250 days to move from the start of development into the market. Brown says the work underway could cut that clock by about 20% by 2027. AI is being fed into the process from idea generation through production.
Speed only tells you how fast things ship. 3M grades the ambition too.
About 75% of current launches sit in Class 3, which covers incremental products. Classes 4 and 5 push into adjacent markets or build something new for a market that didn't buy from 3M before. About a quarter of the portfolio lives in those higher classes today. Inside its industrial businesses, 3M expects that share to climb past 40%.
The launch count starts to matter once you can see how many of those bets are aimed at new demand.
Expanded Beam Optics shows what one bet can turn into. The technology makes durable fiber optic connections for data centers, and Microsoft has been testing it for years. 3M says that testing cut the time to install circuits and start earning revenue by about 85%.
Microsoft Azure became the first announced hyperscale cloud provider to deploy it, in July. 3M expects the business to bring in $40 million to $50 million this year. Brown thinks it could grow four or five times larger over the next several years.
3M is also tracking what the newer portfolio produces in revenue. Products introduced during the past five years should generate about $4 billion in sales this year.
The company watches this through its new product vitality index, which measures the slice of total sales coming from recent launches. Brown expects it to hit the mid teens this year and 20% next year. Get there, and one dollar in every five comes from something introduced within the last five years.
Launches still need a path into accounts. AI tools now help account teams plan their week and pick which opportunities to chase first. Cross-selling has booked $110 million in opportunities so far, with another $120 million in the pipeline. That pipeline grew more than 40% during the quarter.
3M reported 5.4% adjusted organic sales growth in Q2. CFO Anurag Maheshwari pointed to sharper selling and a bigger push from new launches. The company raised its full year organic growth outlook after the quarter closed.
So 3M has given innovation a throughput number, a clock, an ambition mix, a revenue figure, and a route into customer accounts. The point is simple. Leaders can see growth taking shape early enough to do something about it.
3M · Q2 2026
Sources: 3M Q2 2026 earnings call · 3M Q2 2026 results · 3M and Microsoft announcement · Ask 3M announcement
What to do this week.
Map your own innovation clock before you try to cut it.
List every offering that reached a customer in the last twelve months. For each one, write down how long it took from approved development to first sale. That's your map. The slow spots will show up fast.
Then find where it slows down. Where does research take weeks or months when it could take hours? Where is your team synthesizing customer data or sizing a market by hand? Those are the spots where AI can move the clock. 3M is cutting its 250-day cycle by feeding AI into the process at every stage, from spotting emerging signals early to accelerating what used to be manual research work. You can do the same, even if you start with just one bottleneck.
Pick one slow spot and ask where AI could remove friction there. That's how the clock moves.
From the portfolio
Most teams don't stall because they lack ideas. They stall because validating a concept takes months and costs a fortune before anyone knows if it's worth the bet.
FifthRow compresses that cycle. Teams move from problem space to validated concept in four to five weeks. The same work typically runs six to nine months, or a half-million-dollar consulting engagement.
A $20B consumer goods company brought seven problem spaces. Five weeks later, four concepts were validated in front of real consumers. Another team started with a category that had no ideas in it. Four weeks later, five concepts were in market. Two advanced. Three were stopped on evidence instead of opinion. The value isn't just speed. It's knowing which bets deserve the budget before you've spent it.
If this is an area you're looking to get stronger in, reach out. I'd love to hear the context and problem spaces you're sitting on so we can engage the FifthRow team to run this as a live example.
PS. Board of Innovation's AUTONOMOUS summit is next week, September 9 and 10. The agenda dropped this week, and it's built around a problem you'll recognize: AI makes teams faster without making the company faster. Several sessions go straight at how to close that gap and get AI speed to show up in the P&L. That's the same gap 3M is attacking with its R&D factory. It's virtual, and I'll be there both days. Register at autonomoussummit.ai, then reply and we can compare notes afterward.
If your CEO asked where future growth is taking shape inside the business, how fast could you answer?
