When the Product Can't Sell Itself

I kept running into the same disconnect between what I read about growth and what I see in my daily work. So I dug in.

---

Every other week, something about product-led growth shows up in my feed. A thread about how Linear grew to a billion-dollar valuation with $35,000 in marketing spend. A case study about Figma’s bottom-up adoption. A take on why sales teams are a relic if your product is good enough.

I read these and I think: that is a completely different universe from the one I work in.

My day job is in field service and route optimization software. The kind where implementations take months, where the product integrates with ERP systems, and where the person who decides to buy is a VP of Operations or an IT Director. The person who uses the software every day is a dispatcher or a field technician. A free trial would be meaningless here. There is nothing to try without months of integration work first.

This is not an abstract interest for me. I sit in rooms where product strategy and go-to-market get discussed. When PLG comes up, and it does come up, the conversation tends to stall quickly. Either someone dismisses it because “our product is too complex for that,” or someone gets excited about free trials without thinking through what a trial of a route optimizer with no data in it would actually demonstrate. Neither reaction is useful.

I also build Morningbrew on the side, a specialty coffee journal and roaster map. And there, the PLG playbook works exactly as advertised. A user lands on the site, logs a coffee, browses roasters. Two minutes later they know if this is for them. No call. No demo. No committee.

That contrast kept nagging me. I wanted to get past the surface-level takes and understand what, if anything, the PLG model actually offers when the product cannot sell itself. Where is the line between “this does not apply to us” and “we are leaving opportunity on the table”?

So I spent some time with the research, the case studies, and the numbers.

The model, as advertised

The PLG loop is clean. A user discovers the product. They try it. They experience value. They invite others or expand usage. Revenue follows adoption.

It works when three conditions hold. The user can reach value quickly. The user has influence over the buying decision. And the product has a natural way to spread inside an organization.

For Morningbrew, all three are true. I see it in the data. Someone signs up, logs their first coffee, and either comes back or does not. There is no gap between the person using the product and the person deciding whether to keep using it. They are the same person.

For a route optimization engine, none of these hold. The product needs months of integration, real operational data, and vehicle constraints before it can show meaningful results. The dispatcher who would use it daily has no say over whether the VP approves a six-figure contract. And there is no viral loop. Nobody invites a colleague to a field service management system.

The buyer never touches the product. The user never touches the budget. I kept thinking about that gap.

According to an OpenView survey, only 25 percent of SaaS providers with more than 1,000 employees consider PLG a significant part of their growth strategy. That number surprised me. It means three quarters of the enterprise software market has already concluded that the model, as advertised, does not fit.

But here is where it got interesting.

What I found when I looked at the success stories

The companies everyone points to as PLG winners are not actually doing pure PLG. That was the first thing I did not expect.

Linear grew to a $1.25 billion valuation on product quality and word of mouth. $35,000 in total marketing spend. Engineers loved it, invited their teams, usage expanded. Pure PLG story, right? Except that as larger companies joined, Linear added sales-assisted motions for enterprise accounts. The product opened doors. But someone still had to walk through them.

Worth noting: Linear is an issue tracker. A single engineer can sign up, create a workspace, and feel the speed advantage in thirty seconds. That is a fundamentally different product from one that needs three months of ERP integration before it can show you a route. The PLG part works because time-to-value is near zero. That constraint shapes everything.

Figma is the case I found most revealing. According to their S-1 filing, 70 percent of enterprise deals started with a user on a free plan. That is real bottom-up adoption. But Figma also built a serious enterprise sales team. And here is the detail that stuck with me: their early sales hires were required to demonstrate deep product empathy during interviews. The sales team was not there to replace the product’s role. It was there to extend it into rooms the product could not reach on its own.

Datadog is the one that matters most for my context. Cloud monitoring is not a simple tool. It requires technical expertise to implement. Yet Bain’s research found that Datadog designed for time-to-value so aggressively that most users see results within one hour. And the revenue split is telling: 27,000 customers, but the 12 percent spending above $100,000 per year represent 86 percent of annual recurring revenue. The product fills the top of the funnel. A large sales team closes the deals that actually pay the bills.

McKinsey calls this pattern product-led sales. Their survey of 625 SaaS buyers found that 65 percent prefer both product-led and sales-led experiences in the same purchasing decision. Not one or the other. Both.

That reframed the question for me. It is not PLG versus sales. It is: how much of the work can the product do before sales gets involved?

What this might look like on the right side of the spectrum

This is where it gets less clean and more speculative.

A route optimization engine cannot demonstrate value in a five-minute trial. That part is a given. The product needs real data, real constraints, real integration. But that does not mean the product has to stay completely silent until a salesperson explains it from scratch.

What if the sequence flipped? Instead of “sales qualifies, then product delivers,” what if the product built some conviction first, and sales converted it?

An interactive demo loaded with sample data that lets a prospect see optimization results without any integration. Self-service content and ROI calculators that let technical evaluators build their business case on their own time. Published pricing, so a prospect does not need a discovery call just to understand whether the budget fits.

Each of these sounds straightforward on paper. In practice, they all have costs. An interactive demo needs engineering time that competes with the product roadmap. An ROI calculator only works if you have real benchmark data behind it, not hypothetical percentages. Published pricing is the hardest, because it collides with how sales teams are incentivized and how competitors monitor each other. These are not product decisions. They are business model decisions that happen to express themselves through the product.

None of this replaces the sales team. But it changes what the sales conversation is about. Instead of explaining the product from scratch, the salesperson walks into a room where the prospect already has context. Instead of building conviction, they navigate the buying committee.

Working on Morningbrew in the evenings is a useful reset. A user lands, logs a coffee, browses the map. Two minutes later they know whether this is for them. No demo request. No account executive. No implementation timeline. That simplicity is what PLG was designed for. And it is also why the model does not transfer whole into enterprise software. The loop breaks when value takes months to appear. But the underlying instinct, that the product should do as much convincing as possible before a human steps in, that part transfers fine.

Where I landed, for now

I do not think PLG is the answer for the kind of software I work on. The conditions are too different. The implementation is too complex. The buyer and the user are too far apart.

But I also do not think the right response is to ignore it and let sales carry all the weight.

The Datadog example stays with me. Their products are complex. Implementation is not trivial. But they invested in making the first experience fast, even though full deployment takes much longer. Bain’s research says most users see value within one hour. That is not because monitoring tools are simple. It is because the team decided that the product should build conviction early, regardless of what comes after.

The product does not need to do everything. It needs to do enough that when a salesperson enters the conversation, the prospect is already past “what does this do” and into “how do we make this work for us.”

The product and the sales team are not competing strategies. They are two gears in the same engine. I think that is the actual insight buried under all the PLG discourse.

Interactive demos feel like the most obvious starting point for complex B2B. Self-service educational content is straightforward. Published pricing is harder, because it touches sales incentive structures and competitive positioning.

But the angle I find most interesting is the one I see discussed least: equipping the end user to advocate upward.

In field service software, the dispatcher who uses the tool every day has opinions about what works and what does not. The planner who manually builds routes for 40 vehicles knows exactly how much time a better optimizer would save. These people have conviction. What they lack is a way to translate that conviction into a business case their VP will take seriously. They cannot walk into a budget meeting with “I liked the interface.” They need something closer to “here is what this would save us per route, per day, across our fleet.”

That is bottom-up adoption adapted for a world where the user cannot swipe a credit card. The product does not sell itself directly. But it arms the person closest to the problem with the evidence to sell it internally. Figma understood this. Most enterprise software does not even try.

None of this is fully resolved. But the spectrum is real, and most of the industry is still treating it as a binary choice.

---

Sources: McKinsey (Product-Led Sales, 2023), Bain (PLG Development, 2024), Elena Verna / Amplitude (2025), OpenView PLG Benchmarks, Figma S-1 Filing, Linear case studies (Eleken, Thoughtlytics, Lenny’s Newsletter), Datadog / The General Partnership (2025).