What PSA Software Actually Costs for a 30–50 Person Firm

I ran an MSP in the 30–50 person range, and I made both migrations: off one of the two big industry PSA suites, onto Zendesk, and then onto the other big industry suite because that's what the industry used. One of those moves I regret. Pricing pages will give you a per-seat number; none of them will tell you what a PSA costs in productivity, which in my experience is where the real money moved. This article has both — the current numbers, and what actually happened to my firm. We sell a competing option at PurpleOwl, so I have an interest; our pricing is in the table below, run through the same formula as everyone else's.

What does PSA software cost per user?

The current reported figures: ConnectWise runs $60–$85 per user per month for mid-market firms [reported]¹, Autotask $85–$130 per technician when bundled [reported]², HaloPSA roughly $85–$109 per agent depending on team size [reported]³, and the unified RMM+PSA tools (Syncro, Atera, SuperOps) $79–$179 per technician [reported/published]⁴. A general-purpose ticketing tool like Zendesk publishes $19–$55 per agent for its mainstream plans [published]⁵. ConnectWise, Autotask, and Kaseya BMS don't publish prices — you get a quote after a sales conversation.⁶

Here's what surprised me putting this together. From my own MSP years, I remember the ratios rather than the invoices: Zendesk cost us about a third of our first industry suite, and our second industry suite about double Zendesk. Check those ratios against today's published and reported numbers and they still hold, a decade-plus later. The general-purpose ticketing tool still costs a fraction of the industry PSA, and the industry PSA still charges roughly double for being the industry PSA. The structure of this market has persisted even as every individual price changed.

What do the pricing pages leave out?

Implementation and onboarding. Reported figures run $8,000–$12,000 for a partner-led HaloPSA implementation and $10,000–$50,000 for ConnectWise onboarding depending on user count and complexity [reported]⁷, with migrations budgeted at 8–14 weeks [reported]⁸. The seat price is not the price; the price is assembled — seat plus implementation plus the add-ons that arrive later. Never compare seat prices. Compare assembled prices.

And in the MSP world specifically, there's the resale stack. Once you run a PSA, every vendor with an expensive security tool or SIEM figures out that you talk to a lot of folks and comes selling. Those line items ended up mattering more to our costs than the PSA seats did.

What did switching PSAs actually cost me?

Our first PSA was one of the two suites the industry runs on, and at the time it reminded me of SAP: there's nothing you can't do, and four thousand clicks. Figuring out how the thing worked felt like joining a cult. The deeper problem was fit — its heart and soul were a break-fix business, tracking every billable minute in a reactionary way, and we were a maintenance business managing relationships across time. Wrong shape, plus fees we didn't love. We left.

Zendesk was the opposite philosophy. We adopted it early, back when a ticket could only have three priorities — and that constraint was the point. Their bet was that the flow of tickets shouldn't be complicated, and the tool should hold guardrails for you. Getting our data out of the first suite was not a great experience (it helped that I write software for a living); getting it into Zendesk was flawless — once we had the mapping right, everything fell where we expected. The result was measurable: people got an hour or two of their day back, because as an IT firm we couldn't skip documentation, only pay for it in extra steps or in pain later. And it was a clean experiment — same people, same IT issues, and not one client knew or cared that we'd switched systems. The ticket flow continued and the gain was unmistakable.

The other industry suite was the move I regret. We chose it to do things the way the rest of the industry does — same reasoning half the industry uses, and it sounds like marketing spiel because that's what it turned out to be. It was the quintessential demo. Underneath, things weren't as configurable as they looked, and the asymmetries showed up fast: the user interface let me put HTML into a note, but the API accepted only plain text — which crippled the integration tools we lived by. Its configuration items, the established way you're supposed to track things, felt like wearing somebody else's wardrobe and trying to make outfits work. The first thing we hoped to replace — moving one process off spreadsheets — took years to get right. The punchline: roughly half of the benefit we gained moving from the first suite to Zendesk, we gave back moving to the second.

I'll be fair about it. I'm not sure we got screwed, and some of those limits are probably obsolete by now. And I never found — in my own environment or anyone else's — that merely using the industry's software made a company a different company. We were already good at what we did; the tool choice just decided how much of that ability reached the customer. But switching a PSA is not trivial, and done wrong with a fixed crowd, it puts you in the spiral: not enough time, getting beat up, maybe losing customers while you dig out.

Does per-seat pricing matter at 30–50 people?

Less than the blogs claim, honestly. At this size, the difference between PSA seat prices is small next to salaries. What actually moved money in my firm was fit — the productivity swing between Zendesk and the suite that replaced it was worth more than every license fee combined — and inertia. By the time you get a PSA working, switching to a competitor that's broken in different ways is like doing multidimensional differential equations by hand. So firms punt, for years. I understand why; I also watched what punting cost.

There is a ceiling on how long shouting compensates. Between roughly 35 and 50 people you can still resolve the exceptions by yelling across the company. Past that, the misfits your PSA can't hold stop being annoyances and start being process failures.

What would I buy today?

The table uses one formula for every row: unified cost = license + (one-time costs ÷ seats ÷ months). Incumbent figures are [estimate] blends of the published and reported sources above; ours are first-party, marked as such, same formula, no favors. PurpleOwl rows are flat — they don't change with seat count, which is the quiet variable in this table. What's in those rows is on Custom PSA Software for MSPs.

Option	30 seats, 1-yr	30 seats, 3-yr	50 seats, 1-yr	50 seats, 3-yr
Ticketing tool (Zendesk-style) [estimate]	$7K–$25K	$20K–$65K	$11K–$38K	$35K–$105K
Dedicated PSA, published or quoted seat pricing [estimate]	$32K–$79K	$75K–$160K	$50K–$110K	$125K–$230K
Industry PSA, bundled ecosystem [estimate]	$60K–$105K	$120K–$200K	$90K–$155K	$195K–$310K
PurpleOwl custom build (first-party)	$110K–$440K	$130K–$520K	same	same
PurpleOwl guided lease, $1,800/mo (first-party)	$21.6K	$64.8K	same	same
PurpleOwl self-serve lease, $300/mo (first-party)	$3.6K	$10.8K	same	same

Incumbent pricing checked August 2026; vendor list prices change, so re-check before relying on a figure.

Honest annotations. The ticketing row is cheap because it's not a PSA — no agreements engine, no RMM, no industry integrations; it worked for us because our processes carried what the tool didn't. The custom build's first year overlaps and can exceed the bundled-suite band; fit and ownership are the argument there, not price. Lease tokens (roughly $1–$2 per built item, a couple thousand dollars during setup) are excluded the same way incumbent add-ons are; internal labor is excluded from every row, ours included. And the "same" cells are the point: per-seat columns grow with your headcount, flat columns don't, so the 3-year winner at 50 seats is decided by marginal seat cost more than by any list price.

When would I tell a firm to buy the incumbent suite anyway? When your processes are commodity and you want them to stay that way, when nobody internal can own requirements, or when you need the certified-admin hiring pool and the marketplace of pre-built integrations more than you need fit. That's a real set of firms, and the industry PSAs serve them fine.

For everyone else, the question underneath this market is who compromises. A suite starts from the vendor's architecture and works backward toward you; in the first week you discover the things your business does that the platform doesn't support, and you file a feature request. Software built from your process works forward instead — the thing gets built. My years on the second suite were the cost of choosing the first direction while being the second kind of firm. Run the formula on your own quotes, add the assembled costs, and then ask which direction you're buying. If it's the second direction, start at Custom PSA Software for MSPs.

---

Competitor figures are blends of published and reported sources and are marked that way in the text; ours are exact. Any of them is easy to check with a few AI prompts, which is how we found them. The MSP story is my own firm; the two industry suites in it are unnamed by choice.

Sources

Competitor pricing checked August 2026.

Footnotes

1. checkthat.ai, ConnectWise pricing — https://checkthat.ai/brands/connectwise/pricing ↩

2. Flamingo, Autotask alternatives and pricing — https://www.flamingo.run/blog/autotask-alternative ↩

3. Scopable, HaloPSA vs ConnectWise migration costs — https://scopable.io/blog/halopsa-vs-connectwise-migration-costs ; Flamingo, HaloPSA review — https://www.flamingo.run/blog/halopsa-review ↩

4. Scopable, Atera vs Syncro — https://scopable.io/blog/atera-vs-syncro-for-msps ; Atera, Atera vs Syncro — https://www.atera.com/blog/atera-vs-syncro/ ; Flamingo, PSA software for MSPs — https://www.flamingo.run/blog/msp-psa-software ↩

5. Zendesk pricing — https://www.zendesk.com/pricing/ ↩

6. ConnectWise PSA, request pricing — https://www.connectwise.com/platform/psa/request-pricing ; Gartner Peer Insights, Autotask PSA — https://www.gartner.com/reviews/product/autotask-psa ↩

7. Flamingo, HaloPSA review — https://www.flamingo.run/blog/halopsa-review ; Vendr, ConnectWise buyer guide — https://www.vendr.com/buyer-guides/connectwise ↩

8. Scopable, HaloPSA vs ConnectWise migration costs — https://scopable.io/blog/halopsa-vs-connectwise-migration-costs ↩