$1B in Six Months: The Cost of Late Signal Detection

Good to Great by Jim Collins has been on my bookshelf since undergrad. Survivorship bias critiques aside, I’ve always admired how great leaders have the capacity to confront uncomfortable truths about the position of their business, with some even going out of their way to deliberately look for early signals of disruption even when things are going well. Young David loved reading those case studies. And, in retrospect, I guess it’s not really a surprise I went on to study foresight and lead so many projects focused on early signal detection.

The book is top of mind for me today because one industry in particular continues to demonstrate what missing or misreading early signals actually costs. And the numbers just keep getting larger — see PepsiCo paying an eye-watering $1.95B for Poppi in March.

For me, one of the clearest examples is in beverage alcohol, where the cost of late detection is reflected in close to a billion dollars of acquisitions by incumbent brands over the last six months:

• Anheuser-Busch InBev paid up to $490 million for an 85% stake in BeatBox in December.
• Molson Coors acquired Atomic Brands, owner of Monaco Cocktails, in March — its first move into spirits-based RTDs.
• Sazerac picked up Dirty Shirley that same month.
• Mark Anthony Group bought The Finnish Long Drink in April.

Close to a billion dollars in six months. All in spirits-based RTDs. None of them built by incumbents.

One can’t help but feel like the RTD format, and the innovation inside it, has gotten away from incumbents. The window for building, partnering, or backing RTD brands at lower cost has long since closed, and now incumbents are stuck paying a premium to buy insurgent brands.

Insights and innovation professionals should be jumping up and down pointing at the premium being paid for brands like these. There’s a measurable price tag here.

This is what detection lag looks like. And it is yet another example of why organizations need foresight and insights programs that go beyond confirmatory, low-depth research.

Why these deals had to happen

Spirits have been the only real growth engine in beverage alcohol for some time now, reaching 42.4% of US revenue in 2025. Coming out of the pandemic, much of the industry bet big on premiumization as the next leg of growth. And for a while, it worked really well.

Then it stopped. Excluding RTDs, US spirits revenue fell 4% in 2025, with luxury brands declining for the third year running. Players who diversified early into RTDs, and into different occasions, are weathering the correction better than those who stayed focused on the premium thesis.

To put it plainly: RTDs are carrying the category. Spirits-based RTDs reached $3.8 billion in 2025, up 16.4% — and it’s the format all four recent acquisitions bought into.

Incumbents need RTD success stories to build shareholder confidence.

Three Kinds of Lag

It’s worth being precise about what kind of lag is being priced into these deals. Three matter, each tied to a different organizational competency:

• Detection lag: the time between a signal forming and the organization seeing it.
• Interpretation lag: the time between seeing it and understanding what it means.
• Action lag: the time between understanding and deploying capital or capability against it.

The deal premiums in alc bev reflect all three. But detection is where I want to focus — because you can’t interpret what you can’t see, and you can’t act on what you haven’t interpreted.

Most organizations large enough to support an insights function are great at confirming and sizing signals that are already well on their way toward the top of the adoption curve. Long after early innovators have made inroads.

But they’re less capable at detecting what’s forming early. Some are also shedding or restructuring their futures/foresight functions in the face of economic uncertainty.

Addressing Lag

I get why detection work is underweighted. Most insights budgets are loaded toward interpretation — brand trackers, segmentations, U&A studies, concept tests, conjoint exercises. I enjoy doing all of these for clients. They’re all important and have real value. Resources are also finite, which is why so many people reach for off-the-shelf trends decks from Mintel, Canvas8 and others.

But every interpretation-based inquiry shares the same thing: a pre-determined frame. The question is already known. The population to ask is already defined. The problem space is assumed to be relatively stable.

Interpretation work is also easier to scope, easier to brief, and easier to defend in a budget conversation. Detection work is harder on every count. So, it’s easy to wind up in situations where nearly all your spend is on confirming the strategy while almost none of it looks for emerging threats and disruptions.

Bringing it back to the drinks business: by 2019, hard seltzers had grown larger by volume than the entire vodka category, and IWSR was forecasting they would triple in size again by 2023. The signals were very visible.

An interpretation frame treats that signal as a sizing question: how big is this market, what share could we capture? Then you launch a hard seltzer SKU that fits within your existing portfolio strategy.

A detection frame asks something different. Not how big is hard seltzer, but what does it mean that consumers chose a 5% canned product over a $30 bottle of vodka, in this volume? What core business assumptions does that call into question?

Interpretation is responsive. Detection is sense-making for the purpose of meaningfully differentiated strategic action. When organizations cut futures functions or load all of their resources into interpretation, they lose the ability to elevate signals that say “the strategy may no longer be working” — until the strategy has visibly stopped working.

If you’re as nerdy as I am, and want to explore this from a systems perspective, read about Stafford Beer and his Viable Systems Model — but that’s a painful aside because it’s…dense.

A useful discipline, particularly now when stable assumptions are becoming rarer, is to ring-fence 10–15% of every insights budget explicitly for detection work.

There are three big questions to ask that can help you make sense of how to design and scope this sort of work:

Do we have a system that challenges the assumptions underpinning our current strategy?

Most organizations are well-equipped to answer “is the strategy performing?” and barely equipped to answer “are the assumptions the strategy depends on still true?” Different questions, different feedback systems. The first is about measurement and feel-good dashboards. The second is more like an early warning system.

What signals would change our strategic direction if we saw them?

This is a question about what would need to be true to change your mind. Knowing in advance what evidence would force a rethink is half the work of robust strategy, and it’s too often overlooked.

Where might the next competitor or category boundary emerge from, and what do we know about how cultural codes or norms are being shaped in those areas?

Disruption rarely starts with new products. It starts with codes being rewritten — what’s coded as desirable, authentic, status, health, indulgence and so on. Winners detect and take advantage of these new codes early on.

---

The four deals at the top of this piece add up to close to a billion dollars in the last six months. That’s the cost of detection lag, paid by organisations that had access to the signals five years before they had to write the cheques. The story is bleaker still when you look at the public markets, the cost-cutting programmes being announced, and the multi-year strategic theses being publicly walked back.

The gulf between what’s been invested in the capability that could have prevented this and the cost of the missteps themselves is hard to ignore.

Detection work isn’t expensive to resource, but it has to be built and maintained before you need it.

If you’re working on what that early detection looks like inside your organization, I’d love to talk it through with you.