The Uncomfortable Questions That Decide Growth

I'm fascinated by pivotal moments. The moments where a company either decides or loses. Where it moves or stagnates.

Most growth writing covers the steady state. Optimize the funnel, test the creative, improve the conversion rate, tighten the CAC. That work matters and I've spent my career doing it. But it isn't where companies are made. Companies are made, or quietly unmade, in a handful of moments where the current path stops working and someone has to say so out loud.

I've been inside four of those moments at four very different companies. A consumer hardware brand bleeding cash. A B2B hardware company drowning in its own product catalog. A monitoring startup that couldn't raise on its own category. A public company that made the move everyone recommends, brought acquisition in-house, and watched it fail anyway. Different industries, different stages, different channels. The moment was the same every time.

Here's what I believe after living through them. Growth is more than funnels, journeys, data, and luck. At the pivotal moment, growth comes down to an intense focus on the customer coupled with uncomfortable internal questions. The questions everyone in the building already half-knows the answer to, and nobody wants to ask, because the answer costs something. The founding story. The product line someone loves. The vendor relationship. The org chart.

The skill isn't asking questions. It's asking the specific question your company is avoiding, then acting on the answer with full force.

How to know you're in one

A pivotal moment doesn't announce itself. It usually looks like a plateau, and plateaus get explained away. Seasonality. The algorithm. The economy. So before the stories, here's the diagnostic. You're probably in a pivotal moment when:

Effort stopped moving the number. You're working harder than last quarter and the output is flat or worse. More spend, more campaigns, more content, same revenue. When input and output disconnect, the problem isn't volume. Something structural is wrong, and structural problems don't respond to effort.

The story and the data disagree. The deck says one customer, the order data says another. The positioning says one market, the sales calls sound like a different one. Every company I've seen at a pivot point was carrying a gap between what it said about itself and what its customers were actually doing. The size of that gap is the size of the opportunity.

Everyone's working on everything. Budget spread across the whole catalog. Ads targeting every persona anyone ever brainstormed. A roadmap with nine priorities. Spreading is what teams do when they don't want to choose, and refusing to choose is refusing to learn. Nothing gets enough force to prove or disprove anything.

If two of those three are true, you don't have an optimization problem. You have a question you're not asking.

Here's what that looked like in four rooms I was in.

Cubii: who is actually buying this?

I joined Cubii as the first hire. It was me and the founders. Compact under-desk elliptical, great Kickstarter, and by the time I arrived we were losing $40K a month. Ad spend on Facebook and Google before you even counted inventory or salaries. The runway math was not a long conversation.

The founding story was young professionals. Movement for people stuck at desks in offices. It's a good story. It raised money. And the uncomfortable question was sitting right there in the purchase data: is that who's actually buying?

We did the unglamorous work. Surveys, focus groups, early-adopter feedback, then pattern analysis on who actually converted and stayed. The buyers were not the story. They were health-conscious and weight-conscious people, older buyers who wanted low-impact movement, women who cared what the thing looked like in their living room. Our founders hadn't dreamed of some of these segments. That's not a knock on them. Every founding story is a hypothesis, and this one had done its job. It just wasn't the business.

Saying that out loud cost something. It meant the deck was wrong, the targeting was wrong, and a chunk of the creative and messaging we'd paid for was pointed at people who weren't coming. Then we acted on it, and this was a rebuild, not a tune-up. Messaging rebuilt per segment. Health benefits for one, ease and low-impact for the older buyer, design for the style-conscious. Funnel rebuilt around a clear path to purchase instead of a Kickstarter-era landing page. Email built into a real engine, not a newsletter.

And we built something that didn't have a name yet: a paid content engine running on earned PR and real customer testimonials. Instead of polished brand ads, we put paid distribution behind proof. Press coverage became ad creative. Customers telling their own stories became the campaign. Today that's a standard playbook with a dozen names attached to it. At the time we were making it up, because when your buyers don't match your founding story, borrowed credibility from press and real customers closes the trust gap faster than anything you can say about yourself.

We went from losing $40K a month to profitable in 60 days. Same product. The company later exited for over $100 million, selling to customers the original pitch deck never mentioned.

The mechanic: your founding story is a hypothesis, not an identity. The customer who shows up gets a vote, and at the pivot moment their vote is the only one that counts.

CamDo: what actually makes money?

CamDo made rugged camera systems, mostly for construction time-lapse. Good SEO, decent traffic, an education-first content strategy a previous agency-trained marketer had built. Guides, calculators, downloads. Tens of thousands of visits a month and almost no transactions. Activity everywhere, revenue nowhere.

The uncomfortable question was about the catalog and the strategy at the same time: what here actually makes money? Not what could, not what should. What does.

The data answer was brutal and clarifying. A handful of products drove the overwhelming majority of revenue. Everything else was noise that cost money to advertise and diluted the message. And the "nurturing" funnel was attracting researchers, not buyers. Thousands of guide downloads, a couple dozen customers.

So we stripped it down. I've called this stopping the peanut butter. Budget off the whole catalog and onto the three to five products that earned it. Lead magnets replaced with direct offers. Product pages rewritten as sales pages with one obvious action. Remarketing rebuilt around the core products with enough budget to actually learn, instead of a dozen starved campaigns serving wallpaper.

Fewer visitors. More buyers. Revenue tripled inside a year.

The mechanic: focus is a decision about what dies. Every product you keep advertising and every persona you keep targeting is budget and attention taken from the thing that's actually working. The instinct at a plateau is to add. The correct move is almost always to cut, and to give what survives full force.

Therma: what market are we actually in?

Therma, now GlacierGrid, monitored refrigeration. Sensors that told restaurants and facilities when a cooler was failing. Useful product, real customers, and a problem no funnel could fix: we needed to raise a Series A, and niche refrigeration monitoring is a crowded, commoditized category with small budgets and no urgency. You can run flawless campaigns in that category and still be capped by the category itself.

The uncomfortable question wasn't about marketing at all: what market are we actually in? Because the answer we'd been giving, monitoring, was the least valuable true answer available. The same technology, positioned differently, was an energy platform. Cooling as a controllable load on the grid. Different buyers, different budgets, different urgency, different competitive set.

People treat positioning like a messaging exercise. It isn't. Positioning is choosing which competitive set you belong to, and the set determines the price you can charge, the buyer you sell to, and how urgently they feel the problem. Changing the frame meant rebuilding the go-to-market underneath it. New ideal customer profiles, new qualification criteria, new lead scoring, and a real service-level agreement between marketing and sales with disposition feedback both directions so the targeting improved every week. That last unglamorous piece did more than any campaign.

Marketing-driven qualified leads grew 600%, acquisition costs came down, and the company raised a $19M Series A. Same sensors. I went deeper on this kind of category diagnosis in Hard to Market Is a Diagnosis, Not a Condition.

The mechanic: sometimes the growth ceiling is the category you chose, and no amount of execution inside the wrong frame breaks through it. The frame is a decision. Most companies inherited theirs and never re-decided.

The RealReal: the fix everyone recommends had already failed

The pivotal moment doesn't only live in startups. When I consulted for The RealReal, I didn't walk into the story you'd expect. They had already done the thing every growth blog recommends. Customer acquisition had been transitioned in-house. And it had failed. By the time I arrived, the entire acquisition staff had turned over. A public company with paid acquisition at nine-figure-revenue scale, and the department that ran it was gone. No playbook, no institutional knowledge, no team. The org chart said in-house. The capability had walked out the door.

That's the uncomfortable question this room had to face, and it's a harder one than firing an agency: we made the structural change everyone said to make, so why didn't it work? Because in-house is a structure, not a capability. Moving the work inside the building doesn't move the skill inside the building. The company had insourced the accountability without yet having the playbook, and when the people left, there was nothing underneath.

So the work wasn't a transition. It was a rebuild from bare walls, in partnership with the Director of Customer Acquisition and the VP of Growth, and it was hands-on. A playbook of tactics built from scratch across the paid social channels. First-party data infrastructure so targeting ran on what customers actually did. Attribution moved off last-click, which had been quietly overcrediting retargeting. Creative testing on a real cadence, with feedback in days instead of monthly reports. Senior people doing the work themselves until the playbook existed to hand to anyone.

Paid spend scaled from $300K to over $2M a month at 7x return, into a channel driving over $100 million a year in attributed revenue, with acquisition costs down 40%.

The mechanic: the consensus fix relocates the problem when the capability doesn't come with it. In-house versus agency is a structure question. Growth is a capability question. Answer the second one first, because a structure change without the capability underneath just moves the failure closer to your own desk.

The pattern in the four rooms

Look at what these four moments have in common, because it's not the industry, the stage, or the channel.

None of them were channel problems. Cubii's ads weren't the issue, the target was. CamDo's traffic wasn't the issue, the architecture and the catalog were. Therma's campaigns weren't the issue, the category was. The RealReal's spend wasn't the issue, the missing capability under the new structure was. In every case, the team could have kept optimizing the visible thing for another year and stayed exactly where they were.

Every answer was already in the building. The order data knew who was buying Cubii. The revenue report knew which CamDo products mattered. Everyone at Therma knew the monitoring category was small. Everyone at The RealReal knew the department hadn't survived its own transition. Pivotal moments are rarely about discovering new information. They're about someone deciding to act on information the company has been politely ignoring.

Every question had a cost attached. That's why it went unasked. Admitting the founding story was wrong. Killing products someone championed. Rebuilding a go-to-market mid-raise. Rebuilding a department after the fix already failed once. An uncomfortable question is uncomfortable precisely because the answer obligates you. Ask it and you either act or you've publicly chosen not to.

And every fix was boring. This is the part I find funny after all these years. Not one of these turnarounds involved a growth hack. Segmentation research. Cutting a catalog. Qualification criteria and an SLA. A playbook rebuilt by hand. After the hard question gets asked, the work that follows is repetition and discipline. The drama is all in the asking.

Running the question on your own company

If you suspect you're at a pivot moment, here's the protocol I'd run. It's simple. It isn't easy.

Write down the question you'd least like a board member to ask. Not the hardest question in general. The one that makes your stomach drop because you suspect the answer. In my experience the company's real question is usually within one step of that one. If you can't think of it, ask your team what question the company avoids, and ask them separately. The overlap in their answers is your question.

Answer it with data the company already has. Order data, cohort data, win-loss calls, revenue by product, pipeline by source. Not a new research project. The point of using existing data is that it removes the escape hatch of we need more information. You almost always have the information. What's missing is the willingness to look at it as a verdict instead of a dashboard.

Then act with full force or explicitly decide not to. The failure mode isn't asking the question and getting it wrong. It's asking, seeing the answer, and hedging. Keeping most of the catalog. Sort of repositioning. Bringing half the spend in-house. A pivot executed at half force gives you the costs of the change and the results of the status quo. Cubii didn't add the new segments to the old targeting, we rebuilt around them. CamDo didn't trim the catalog, we cut it to the bone and put everything behind the survivors.

One honest caution: not every plateau is a pivotal moment, and treating every rough quarter as an identity crisis is its own disease. Metrics have noise. Sometimes the answer to the uncomfortable question is that the strategy is right and the execution needs another two quarters of repetition. That's a legitimate answer too. The discipline is in genuinely asking, not in always pivoting.

The companies I've watched win weren't the ones with the best steady-state marketing, though the discipline compounds. They were the ones that could tell the truth to themselves faster than their competitors could, at the exact moment the truth got expensive.

So, plainly: what's the question your company is avoiding right now? You almost certainly just thought of it. That reaction is the diagnostic.

The rest is whether you ask it out loud. And if you want a second set of eyes on what the answer means for your growth, that's a good use of a 20-minute conversation.