The hardest stretch in B2B SaaS: $10M to $75M ARR

There is a phase in every B2B SaaS company where the founder-led motion that got the company to twenty million in ARR stops working, and the repeatable scaling motion that will get the company to seventy-five million has not yet been built.

This phase is not a stage. It is a transition. Companies are usually in it for somewhere between eighteen and thirty-six months. Some companies traverse it cleanly. Most do not. The ones that fail to traverse it are not failures in the conventional sense. They become twenty-five million dollar companies with thirty-five million dollar burn rates, and the next funding round is about survival rather than scale.

I have spent most of my career inside companies trying to cross this gap. I want to write down what I have learned about why it is hard and what actually works.

Why the founder motion stops working

The founder-led motion at five to twenty million ARR has a specific shape. The founder is the principal seller. The product is sold through deep relationship, technical credibility, and the founder’s ability to absorb edge cases inside individual deals. Pricing is negotiated case by case. The handful of reps in the org are essentially closing leads that the founder warmed.

This works until it stops. It stops for three reasons.

First, the founder hits a personal bandwidth ceiling. Sixty deals a year at three hundred thousand ACV is twenty million in new ARR, and almost no founder can sustain that pace for more than two years.

Second, the deals that the founder absorbs through sheer effort are not replicable by anyone else on the team. The “process,” to the extent there is one, lives in the founder’s head. New hires fail to close because they do not have the founder’s context, the founder’s relationships, or the founder’s ability to bend the product to fit the deal.

Third, the customer base starts to demand things that founder-led motion cannot deliver: predictable renewals, real customer success, structured expansion conversations, a coherent product roadmap influenced by aggregated voice-of-customer rather than by the last three loud accounts.

The CRO hired to traverse this transition is hired to build the system that replaces the founder. Not metaphorically. Literally. The CRO’s job is to take what the founder does intuitively and turn it into a process that ten or twenty or fifty people can execute.

Why this is harder than it sounds

The standard playbook says: hire a VP of Sales, hire a VP of Marketing, hire a VP of Customer Success, define an ICP, build a sales process, install a CRM. Most CROs can list these activities. Most CROs even do them. And most CROs still fail to traverse the gap.

The reason: the playbook describes the outputs. It does not describe the sequence, and the sequence is what matters.

Hiring a VP of Sales in month three, before the ICP is defined and the qualification framework is locked, produces a VP of Sales who builds a team around the wrong segment. Building a customer success function before the renewal motion is mapped produces a CS team that fights fires for two years. Installing a CRM before the sales process is defined produces a CRM that nobody trusts because it does not reflect how deals actually move.

The right sequence, in my experience, is roughly this:

Lock the ICP and the disqualification criteria first. Most companies have a stated ICP. Most companies are still selling to accounts that do not fit it. The first thirty days are about getting honest about which accounts are converting and which are not, and writing down the disqualification rules.
Map the deal stages against actual won deals. Not against a theoretical sales process. Pull the last twenty wins. Identify the actual movements that distinguished them. Build the stages around what happened, not around what should have happened.
Install the qualification framework and train the existing team on it before hiring anyone new. Adding headcount before the qualification framework is locked just multiplies the noise.
Build the forecast discipline before scaling marketing spend. If the team cannot forecast accurately at current pipeline volume, doubling the pipeline does not solve the problem. It makes the problem larger.
Then, and only then, hire the next layer. VP of Sales to scale the qualified-pipeline motion. VP of Marketing to feed the qualified-pipeline definition. VP of Customer Success to operationalize the renewal and expansion motion that is now mappable.

This sequence is unglamorous. It does not look like progress for the first ninety days. Boards get nervous. The CROs who succeed in the transition are the ones who hold the sequence even when the board is pushing for “more pipeline, more reps, more revenue.”

What changes in the AI era

I want to add one more layer that has only become visible in the last eighteen months.

The traditional ten-to-seventy-five-million scaling motion was a headcount-led motion. Each million in new ARR required somewhere between five and ten incremental commercial headcount. Sales, SDRs, CSMs, Sales Ops, Marketing. The economics worked when ARR per commercial headcount was around two hundred and fifty thousand. They stopped working when companies tried to scale before that ratio was locked.

In the AI era, the ratio is changing. At Calliope I am running a motion at three commercial headcount that traditionally requires twelve or more. The agents are doing the work the SDRs used to do. The CSMs are using AI-augmented tooling for renewal pattern recognition. The forecast model runs against transcript data, not against rep self-reporting.

This is not a cost-savings story. The headcount we are not hiring at the SDR layer is headcount we are spending on senior commercial talent who can run the agent layer, design the pilots, and develop the champions. The total dollars are similar. The shape is different.

For Series B and C companies in 2026, this means the scaling motion that worked from 2018 to 2022 needs to be redesigned. Not augmented. Redesigned. The CROs who can architect the new motion will produce companies that get to seventy-five million ARR at a meaningfully lower headcount and a meaningfully lower burn rate. The CROs who run the old playbook will produce companies that look fine at thirty-five million and then stall.

What I am still figuring out

I do not want this to read as if I have the playbook fully written down. I do not. I have done this transition successfully at Tehama, at PKWARE in the post-acquisition variant, and at Equus in the expansion variant. I am doing it now at Calliope. Each cycle teaches me something new.

What I am most uncertain about right now: the right shape of the customer success function when the renewal motion is partly agent-driven. The right comp model for a senior AE running an agent-augmented territory that produces three times the pipeline of a traditional AE. The right way to structure marketing when the SDR-fed inbound motion is largely automated and the leverage is in champion development.

I write about this here partly because writing forces precision, and partly because the operators I respect most are the ones who share what they are figuring out in real time. The CROs who pretend they have a finished playbook in 2026 are either lying or are not paying attention.