Frameworks
Patterns that travel from one operating seat to the next.
I focus a lot on pattern recognition. See the same problem enough times, at different companies, at different stages, and something clicks. I find myself adjusting my approach so it fails less and works better. Part of that probably comes from my project management and Six Sigma background, but I just can't see something not working repeatedly without wanting to design and implement a scalable fix.
That said, we don't always have the luxury of that level of meticulous due diligence in a SaaS startup. That quandary is just another problem to solve. By taking a hybrid, adaptable approach, I've been able to build frameworks that fit the actual size of the problem in front of me, not bigger, not smaller, with room to expand them later if the situation calls for it. I'm putting them here because they explain how I think about the priorities around Customer Success and the adjacent functions in a SaaS company. Collectively, this is what I mean when I talk about bringing an operating system into a new organization: not a rigid playbook, but a set of frameworks I reach for, adapt to what's actually in front of me, and build out further as the problem demands it.
Flagship Framework
The Five Drivers of Retention
NDR and GDR are the scoreboard, not the reason for the score. Every time I've worked on a retention problem, it's come down to one or more of these five things actually changing:
- 01
Product Value
Does the product actually do what it was sold to do? At one particular company early in my tenure, platform stability problems were driving churn on their own, separate from anything Customer Success could fix with better process. At another, there was misalignment of value expectations between the pre- and post-sales cycles. When that happens, Customer Success ends up doing the work of reconciling what was promised with what's actually true, and finding a resolution both sides can live with.
- 02
Product Adoption
Adoption is a critical driver, and it's often inconsistently tracked, in part because measuring it well usually takes a custom approach rather than an off-the-shelf metric. Companies also need to avoid taking it for granted: customers can be paying and still not be using enough of the product to feel the value. At Hyperscience, getting time-to-value down by 50% had to happen before the retention work could really take hold. You can't retain someone on value they haven't experienced yet.
- 03
Customer Success Execution
Customer Success, Professional Services, Support, Technical Services, Training: this is usually where I've spent most of my time. Does the org actually have the structure and ownership to act on a risk signal or leading indicator before it becomes a lost renewal? Moving renewal and expansion ownership into a dedicated CSM team at Sharpen was probably the single biggest lever in that NDR turnaround, more than any individual tactic.
- 04
Commercial Alignment
Pricing and renewal terms have to be defensible, and a lot of teams are quietly afraid to ask for what they're worth. We pushed renewal increases averaging 5.2% against a 2% target at Sharpen, which only worked because the value case was real. And at CyberShift, the channel partnership model wasn't going to work financially until partner enablement came before partner volume — that's a commercial alignment problem as much as an execution one.
- 05
Market Forces
Some of this is just outside anyone's control: competitive pressure, timing, brand awareness. But how you respond to it, proactively or reactively, still matters. When there are early signals of an acquisition or cost cuts coming, that's usually the moment to get back in front of stakeholders and make the value case again before anyone asks you to. CyberShift's shift toward channel partners created a new kind of dependency, and a new margin problem, that needed its own approach rather than just doing the old thing harder.
None of these five tells the whole story by itself, and in my experience the real work is figuring out which one is actually driving the number before you spend a quarter (or a year) fixing the wrong thing. That's basically the same diagnostic instinct behind CARRavan.
Operating Frameworks
A few more patterns I lean on.
Each one answers a specific operating question, not a retention metric directly.
Capacity-to-Retention
This is a question I've gotten from boards more than once, in different words: how much Customer Success investment is actually enough?
Under-resourcing shows up slowly. Not as immediate churn, but as response times creeping up and account health visibility eroding, and it usually doesn't hit the NDR number until two or three quarters later, which makes it easy to miss in real time. Over-resourcing hides a different problem: retention looks fine, but it's being propped up by a few people working really hard rather than by a repeatable system, and the cost shows up eventually in burnout and turnover instead of the P&L.
The right level of CS capacity is bespoke to each company's product lifecycle stage and the market it serves. Early companies without much brand awareness, solving new problems with a nascent product, need a lot more CS support. At the other end of the spectrum, a mainstream, mature product with name-brand recognition and an educated customer base won't need much at all, or can charge a premium for it (think Zendesk, Salesforce, Jira, Gainsight, Asana). A modest time-and-motion analysis of what "good" customer support actually looks like will usually surface the most important drivers, including one I think about a lot:
Product Tax
The operational burden your product places on Customer Success. Immature or overly complex products often require significant CSM involvement just to compensate for technical issues, poor usability, or incomplete documentation. That's work that has nothing to do with driving adoption or value realization.
The higher the product tax, the more CS resources get diverted from proactive retention toward reactive firefighting. Addressing the factors behind product tax over time reduces the engagement burden on the team, and changes what "right-sized" capacity even means as the product matures.
At Sharpen we ran this with a lean team of 4 holding 300+ accounts, using AI-assisted workflows instead of adding headcount, partly by necessity, given where the budget was, but it ended up proving the point: it's not really about a ratio. It's about whether the team has enough room to catch a risk signal before it turns into a renewal conversation, not the day after.
Managing Imperfect Deals
Don't optimize for perfect deals. Optimize for an organization that can successfully absorb reasonable complexity.
Every SaaS organization takes on deals with a little "hair" on them. Whether driven by competitive pressure, evolving product-market fit, ambitious customer expectations, or the realities of an emerging market, not every customer arrives in Customer Success under ideal circumstances. That's a normal part of building and scaling a SaaS business.
I've seen this play out most clearly around integrations. At CyberShift, Sharpen, and Cascade, competitive pressure and customer expectations pushed sales to commit to integrations the product didn't actually have yet. Rather than walk away from that revenue, sales looped Customer Success into the sales cycle early to scope out what each integration would actually require. That let us size the risk honestly and start reserving engineering capacity ahead of time, instead of scrambling after the contract was signed. It also gave product and engineering a chance to learn from real requirements and eventually build a standard, scalable way to handle that category of request instead of custom-building it every time. We kept the revenue, stayed competitive, and the product came out stronger for the next customer who needed the same thing.
I've never seen Customer Success benefit from treating these deals as sales failures. They're a chance to show adaptability and operational discipline instead. The goal isn't zero imperfect deals; it's making sure I have the process and cross-functional alignment in place to manage reasonable commercial risk well.
When Sales, Customer Success, Product, Services, and Support work together with shared accountability, these customers often become some of the most valuable learning opportunities in the business. They expose gaps in the product, onboarding, pricing, messaging, or operational process, and closing those gaps makes things better for the customers who come after.
Liberate the Talent
The most consistent pattern across every one of my roles isn't a metric. It's a belief that organizations routinely underestimate the people already inside them. When I join a company, finding the person whose potential exceeds their current role is one of the first things I look for. They're usually not hard to find. They're just overlooked, untrusted with enough scope, or waiting for someone to ask more of them.
I run that process through four stages:
Teach — establish clear expectations and the reasoning behind them, not just the task.
Do — lead by example, so people are learning from observed behavior, not just instruction.
Mentor — stay engaged enough to let someone practice and occasionally fail safely.
Liberate — step back once they've proven they can operate independently. The goal is autonomy, not dependence.
Two examples from a lot of years of this:
At Hyperscience, a software developer on my Technical Services team had no management experience when I needed someone to run the function, so I put him in charge. He needed almost no coaching to get there. He built the team's operating cadence from scratch, went on to help create the platform's first SDK, and improved Level 1 support solve rates from 53% to 78% while absorbing a 113% year-over-year increase in support volume without breaking SLAs.
At Sharpen CX, a sales rep with no management experience was still finding her footing when the CEO moved expansion, upsell, and renewal ownership from Sales to Customer Success. She'd never run a team, but she had the instincts and the customer relationships, so I put her in charge of the new CSM function. Within nine months, through team attrition that left her down to two effective CSMs, she helped take GDR from 60% to 96% and NDR from 70% to 100%, including a 5.2% average renewal rate increase against a 2% target. She was promoted to VP after I left.
Neither of them needed me forever. That's the point.
Want to dig in further?
These frameworks come to life inside operating engagements — applied to your retention numbers, your team, your commercial reality.