Here's a pattern I've seen play out at dozens of companies: a team is working hard, the strategy is thoughtful, the investment is real — and growth still stalls. The instinct is to question the execution. Usually, the problem is the diagnosis.
Most companies aren't failing because they're solving their problems badly. They're failing because they're solving the wrong problems for their stage.
I've watched teams at $15M invest in marketing mix modeling infrastructure they won't have the data volume to use for another three years. I've watched teams at $80M still pulling revenue data into spreadsheets every Monday morning — because no one built the foundation when the company was small enough to do it cleanly. Both teams are working hard. Both are behind.
The mistake isn't bad strategy. It's stage mismatch.
The Three Stages Where Growth Breaks
Each growth stage breaks in a different place. And each requires a completely different fix.
$0 to $30M: You don't have a growth system yet. At this stage, the most common mistake is reaching for complexity before you've built the foundation. You don't need an attribution model. You don't need a data warehouse. You need the basics defined: which metrics actually matter, who owns each one, and whether the numbers are trusted by everyone in the room. Without that foundation, every analysis is contested and every decision is slower than it needs to be.
$30M to $100M: You have data but not decision velocity. The foundation exists. The problem is you've added headcount, process, and organizational complexity faster than your decision-making infrastructure could absorb it. What used to take a two-person conversation now takes a cross-functional alignment meeting. The data is there. The speed isn't. Teams at this stage often diagnose a strategy problem when they actually have a process problem. The fix isn't more analysis — it's eliminating friction between insight and action.
$100M+: The playbook that got you here won't scale. At scale, the system exists and the muscle exists. The problem is that the executives who built the system don't agree on how to evolve it. You're early again — but with a much more expensive team and a lot more to protect. The instinct at this stage is to optimize the existing system. That's usually wrong. The system needs rethinking, not refinement.
Why AI Makes This Worse
AI has made stage mismatch more common, not less. The availability of sophisticated tooling — attribution platforms, predictive LTV models, AI-powered growth analytics — has compressed the perceived distance between stages. Teams at Stage 1 now have access to Stage 3 complexity they genuinely cannot use yet.
The tools are impressive. The data foundation to support them doesn't exist. The result is investment in the wrong layer of the stack — a team that's technically sophisticated but operationally behind. The right use of AI looks completely different at each stage: at $0 to $30M, automate data hygiene; at $30M to $100M, accelerate decision loops; at $100M+, surface cross-functional patterns executives can't see from their individual vantage points.
How to Actually Diagnose Your Stage
Before you decide what to fix, answer these honestly:
- Are your key metrics clearly defined, owned, and trusted by everyone who uses them? If not, you're at Stage 1 — regardless of your revenue.
- Do you have data but find that decisions still take too long and require too many people? That's a Stage 2 problem — even if you have a sophisticated analytics stack.
- Is your C-suite aligned on what you're optimizing for as you scale, or do the same conversations recur without resolution? That's Stage 3.
Figure out which stage you're actually in — not where you want to be, not where your headcount suggests you should be, but where you actually are. Then solve for that. The most expensive growth mistakes don't come from bad strategy. They come from applying a Stage 3 solution to a Stage 1 problem.
David Manela is co-founder of Exactius, a growth and data science company. Follow him on LinkedIn for more frameworks on growth, marketing, and capital allocation.
David Manela
David Manela is the founder of Exactius and creator of the Growth Operating System — a framework for deploying capital-efficient, compounding growth inside scaling companies.
FAQ
Frequently asked
How do I find out where my growth system is breaking before it gets expensive?
Diagnose your stage before you choose a fix. This article sets out three tests. If key metrics are not defined, owned and trusted by everyone who uses them, you have a Stage 1 problem whatever your revenue. If you have data but decisions take too long, it is Stage 2. If the C-suite cannot agree on what to optimize as you scale, it is Stage 3.
How much data do you need for marketing mix modeling?
More than many early-stage companies have. In a worked example for a national-level model, Google's Meridian documentation says two years of weekly data, or 104 data points, is too few and suggests three years of weekly data instead. Google avoids setting a single minimum, because what matters is the number of data points per model parameter, which falls as channels and controls are added.
Why do growth problems change as a company scales?
Management research has described this pattern. In "Evolution and Revolution as Organizations Grow," Larry Greiner wrote in Harvard Business Review that management practices that work well in one phase of growth may bring on a crisis in another. This article applies the same idea to growth systems, with different break points at $0 to $30M, $30M to $100M and above $100M in revenue.
Are growth stalls caused by internal or external factors?
Mostly internal, according to two large studies. Harvard Business Review's 2008 stall-point research by Matthew Olson, Derek van Bever and Seth Verry found that external factors such as regulation explained only 13% of stalls at the very large companies studied. Bain & Company's 2016 Founder's Mentality research, drawn from 8,000 companies, found that 85% of the barriers to profitable growth for the average company were internal.
How common are growth stalls?
Common, even at the largest companies. In "When Growth Stalls," published in Harvard Business Review in March 2008, Matthew Olson, Derek van Bever and Seth Verry reported that 87% of the Fortune 100 size companies they studied over half a century had stalled at least once. The research covered very large companies, so treat it as a reference point for smaller businesses, not a forecast.
What does it cost to work from data no one trusts?
Gartner research estimates that poor data quality costs organizations at least $12.9 million a year on average. That figure describes organizations in general, so use it as a sign of scale, not a benchmark. This article treats untrusted metrics as the defining Stage 1 problem, because when every analysis is contested, every decision takes longer than it needs to.
Related Reading
Keep going
Ready to fix the system?
Your growth system is either compounding or degrading.
Book a diagnostic call. We'll identify where your growth system is breaking and what it's costing you.