Every company I've worked with — a five-person startup, a four-hundred-person scale-up, a multinational with more business units than I have fingers — has asked me some version of the same question: what tools do we actually need?

Wrong question. The right one is: what stage are we in, and what does the stack need to do at this stage?

A company in its wild-west years is optimizing for speed and survival — break things, ship fast, prove the core offer works before the runway runs out. A company scaling toward enterprise is doing something almost the opposite: locking in efficiency, tightening the operation, substantiating the core well enough to justify the next round or the IPO. And a true enterprise is playing a third game entirely — protecting a mature core while constantly turning over the soil, weeding out what's dying, cultivating whatever's about to become the next high-growth line, usually with an M&A team involved somewhere in the mix.

Same five jobs need doing at every one of those stages. What changes is how many tools live in each job, how specialized they get, and how much governance sits around them.

Below are three stack blueprints — small, mid-market, and enterprise — each with an interactive graphic you can build up layer by layer using the slider underneath it. Drag it from the bare skeleton, to the tech categories, to the single tool I'd actually point you toward, to the specialized tools that optimize it. At full extension, every name is a live link and the whole graphic is downloadable to keep on hand.

One caveat before we start: this is my recommendation, not gospel. This field reshuffles every quarter. Treat this as a snapshot from where I'm sitting today, not a permanent answer.

The same five jobs, no matter the size

Every stack below is built on the same five logical layers, because the jobs don't change — only the answer to "who does this job right now" does.

Buy — and evaluate — by which of these five jobs a tool is actually doing, not by category hype or how good the demo looked. A tool that can't tell you which job it's doing usually can't tell you its ROI either.

Now here's how that skeleton gets dressed at each stage. Each graphic below has four states: the bare layers, the tech categories, my single recommended tool per category, and up to three specialized tools that optimize or extend it. Drag the slider, or hit the download button to jump straight to the full build.

Small: breaking ground

If you're under fifty people and still finding product-market fit, your stack has exactly one job: don't get in your own way. This is the wild-west phase — fail quick, break things, make it happen, get the core offer in front of real customers before the money runs out. Nobody on this team has "operations" in their title yet, which means every tool has to be usable by whoever's closest to the problem, not by a specialist who was hired to run it.

The temptation at this stage is to buy tools like a bigger company would, because that's what "professional" looks like on a pitch deck. Resist it. It's not a hunch — premature scaling shows up again and again in research on why high-growth startups fail, not as a footnote but as a leading cause.

The minimum viable stack below is built around that constraint: one tool per job, consolidated wherever a single platform can honestly cover two jobs at once, and specialized add-ons kept to the handful that solve a real, current pain — not a hypothetical future one. Notice the same platform shows up in more than one layer. That's not an error. At this size, the fewer logins your team needs, the faster you actually move.

Mid-market: building the root system

Somewhere between fifty and a few thousand people, the job changes. You're not proving the core offer works anymore — you've proven it, you've raised on it, and now a board, an investor, or just plain math wants to see the efficiency that turns "promising" into "durable." This is the substantiate-the-core phase: tightening operations, focusing priorities, and proving ROI hard enough that the next round — or the acquisition offer — actually happens.

This is also where a real RevOps or marketing ops function usually shows up for the first time, and not coincidentally, it's where tool counts grow fastest of any company-size segment.

That growth is fine, even healthy, as long as it's deliberate. This is the stage where the specialized-tools overlay in the graphic below actually starts to matter: not because more tools automatically means more revenue, but because a dedicated ops function can now support complexity a general-purpose platform can't handle alone — deeper enrichment, real intent signals, conversation intelligence, actual pipeline forecasting instead of a gut check in a spreadsheet. The risk at this stage isn't under-tooling anymore. It's hiring specialized tools before you've hired the person who'll actually own them.

Enterprise: turning over the soil

By the time a company is a true enterprise, the question stops being "what do we need" and becomes "what do we need to stop doing." The core business is mature — worth protecting, not reinventing — but staying relevant means constantly cultivating whatever's about to become the next high-growth line, usually alongside an M&A team that just bought, or is about to sell, an entire business unit with its own, completely separate stack.

This is tech rotation as a permanent condition, not a project with an end date. Weed out what's old and dying, cultivate what's new, protect the core — and, this is the part enterprises get wrong most often, resist the urge to force-merge every acquired business unit onto one platform just because procurement wants a single vendor. Sometimes the synergy is real. Sometimes the fastest way to sabotage a high-growth acquisition is to strip it of the scrappy, independent tooling that made it move fast enough to get acquired in the first place.

And that turnover is warranted. Estimates of the average enterprise stack range from around 91 tools to well over 250 depending on whose audit you trust and how they define "tool" — but every methodology I've seen agrees on the uncomfortable part: enterprises actively use well under half of what they're paying for. One widely cited figure puts stack utilization at 33%, down from 58% just a few years earlier.

The enterprise build below leans toward platforms engineered to survive that kind of churn — built for governance, multi-business-unit use, and M&A integration from the ground up — with a specialized layer that's meant to be swapped, not permanent.

Match the stack to the stage, not the ambition

None of these three builds is "better" than the others — they're solving different problems. A fifty-person company running the enterprise stack isn't sophisticated, it's drowning in tools nobody has time to configure. A five-thousand-person company running the small-company stack isn't lean, it's under-governed and one bad quarter away from a data mess nobody can untangle.

The move, at any size, is the same: know which of the five jobs above your current tools are actually doing, be honest about which ones are dead weight, and change the stack when the business changes — not before, and not two years after. Wild-west companies should feel a little uncomfortable outgrowing this list within a year. Enterprises should expect to have this exact conversation again next year, because if you're not turning over some soil annually, something's probably dying underneath you without your noticing.

I'll be updating all three of these blueprints as the tools — and my opinions — change. This is a snapshot from where I'm sitting today, not a permanent verdict. If you think I've got a pick wrong, or you're running something in one of these five jobs that deserves a look, tell me. That's genuinely how this list gets better.