Bullhorn or custom CRM? The honest answer for boutique recruitment firms

Every comparison post you'll find on this topic is written by someone selling one of the two. Bullhorn's blog explains why Bullhorn beats custom. Custom dev shops explain why custom beats Bullhorn. Both are partly right. Neither tells the boutique recruitment owner what they actually need to know.

I build custom CRMs on Microsoft 365. I have skin in this game. But I've also turned down clients who would have been better served by Bullhorn or Vincere, and I'm going to tell you exactly when that is below.

When Bullhorn (or Vincere, or JobAdder) is the right answer

Off-the-shelf SaaS recruitment platforms exist because they solve a real problem well. A 10-desk agency on Bullhorn typically pays $15,000 to $25,000 per year and gets a system that genuinely works (Treegarden Bullhorn analysis). For many firms, that's the right trade.

Specifically, Bullhorn / Vincere / JobAdder is the right answer when:

• Your workflow is genuinely standard. Candidate intake, pipeline stages, client submissions, placements. If your team's workflow looks like 80% of recruitment agencies, off-the-shelf is shaped for you.
• You don't have time for a build. Bullhorn implementation: 4-8 weeks to live, mostly configuration. Custom: 6-12 weeks of design + build. If you need a system running yesterday, SaaS wins on speed.
• You're comfortable with the per-seat economics. 10 users at $165/user/month = $19.8k/year, scaling to 25 users = $49.5k/year. If the placement volume justifies the seat fees, it works.
• You want product roadmap as a service. Bullhorn ships features quarterly. Custom build = you decide the roadmap (which means you also pay for it).
• You're never going to read the FAQ on data ownership. If you don't care where your data lives or who controls the schema, the lock-in concern is theoretical for you. Pragmatic, and pragmatically fine.

When custom on Microsoft 365 is the right answer

The flip side. Custom-built recruitment systems on your own Microsoft 365 tenant make sense when one or more of these is true:

• Your workflow is genuinely non-standard. Multi-tenant client portals, unusual pipeline stages, specific compliance workflows (healthcare, finance, regulated industries), industry-specific document generation. If you keep bending Bullhorn around your process, the bend is the cost.
• You're in a regulated industry. Healthcare (FADP, HIN, Art. 321 SCC), financial services, defence, government-adjacent. Where data sovereignty matters and SaaS-cloud answers don't satisfy your procurement, custom on your own tenant is often the only acceptable answer.
• You already pay for Microsoft 365. Most boutique firms have M365 sitting underused: calendar, email, Teams, basic file storage. Building on top of what you already license costs nothing in additional per-seat fees and uses infrastructure you've already paid for.
• You're tired of the per-seat trap. 10 users on Bullhorn at $165/seat is fine. 25 users at $165/seat starts to bite. 50 users and your CRM is a six-figure annual line item. Custom is a build (€12k-28k of scope depending on tier, these days usually paid as one fixed monthly number that also covers hosting, support and training), and the economics flip at a smaller desk than most people expect (the math is below).
• Data ownership matters to you (or your clients). You want to know your candidate data, client relationships and historical pipeline live on infrastructure you control, with the schema and documentation in your own repository.

The five-year cost math, shown both ways

For a 10-desk firm. Standard placement work. Real numbers.

On hard costs, custom comes in at roughly €31,000 vs ~€114,000 over five years for a 10-desk firm. (Bullhorn is priced in USD; figures converted at ~1.08, and Bullhorn does not publish list pricing, so its number is an industry estimate.)

That gap is wide enough that it deserves an explanation rather than a victory lap, because the two columns stop being the same kind of thing in year three. The build is paid off inside the term. From month 25 the €850 stops, and what continues is care: support, hosting and a small development budget, which you can also stop. The Bullhorn column has no such moment. It is rent, and rent has no end date. So the honest reading is not "custom is 73% cheaper", it is "one of these finishes paying for something and the other does not."

Those numbers still come with honest asterisks:

• The Bullhorn number reflects mid-market pricing; volume discounts at scale can lower it 20-30%.
• The custom number assumes a foundation-tier build that fits standard scope. Regulated builds (€18k-28k) shift the math up but stay competitive. The table below spreads it over two years because that is what this example picks; you choose the term, and a longer one means a smaller monthly and a bigger total.
• The €31,000 assumes you keep paying for care for the whole three years after the term. Drop it and the five-year figure falls to about €22,000, but then nobody is watching the system.
• The Bullhorn number includes vendor product development (you get new features without paying for them); the custom number includes only what's in your contract (new features = new scope).
• Bullhorn doesn't publish pricing: the ~$20k/yr is an industry estimate for 10 mid-market seats, not a quoted rate; a small or hard-negotiated deal can run lower.
• "€0 incremental" assumes you're already on Microsoft 365 and the build uses standard connectors (a SharePoint-based system, how I build). A design that needs premium/Dataverse connectors or Power BI reporting would add per-user licences.

Per-seat economics scale down and a build's cost does not, so the smaller the desk, the better Bullhorn looks. On these numbers the two meet somewhere around three or four desks, and below that renting is simply cheaper. That figure is Bullhorn's, not a universal one, and it moves with whoever you are actually comparing me to. Against a platform at roughly $85 a seat the crossover sits meaningfully higher, and against one at $169 it sits close to Bullhorn's own. If you want the number for your case, tell me which product is on your desk and I will run it with their price rather than a convenient one. I would not push a firm that size toward a build on cost alone, because at three desks you are usually buying something other than savings: a workflow nobody else sells, or data you cannot keep in someone else's system. For 25 desks the cost argument stops being close.

The lock-in question every owner should ask before signing

This is the part Bullhorn's sales team doesn't volunteer.

When you sign with an off-the-shelf ATS, four mechanics quietly lock you in over time:

1. Your data lives on their servers. If you cancel, you get a CSV export, which captures maybe 40% of what your team actually built (custom fields, workflow state, integration data, message history, attachments). The rest you lose.
2. The export format is proprietary. Even the data you do get out doesn't map cleanly into another system. Migrating to a different ATS = manual re-entry or a five-figure data services bill.
3. Your team's workflow is shaped to their UI. Six months in, your consultants think and work in Bullhorn-shape. Moving them off is a months-long re-training exercise that hits placement velocity.
4. Per-seat fees compound for years. Vendors raise prices 5-10% annually. They know you're not leaving. Leaving is more painful than the price hike.

This is why almost no firm ever switches ATS once they've been on one for two years. Review roundups report agencies quoted 5,000 to 10,000 US dollars just to get their own candidate records back out, before a single hour of the new system is paid for. Add the build and the retraining on the other side and most firms decide it is not worth the risk.

The honest three-question decision tree

Skip the comparison spreadsheets. These three questions decide it for most firms:

1. Is your workflow standard or specific? Standard → Bullhorn fits. Specific or regulated → custom fits.
2. Are you growing past 15-20 desks in the next 3 years? Yes → per-seat economics will hurt; custom math wins. No → SaaS economics are fine.
3. Does data ownership and exit cost matter to you? Yes → custom is the only answer that delivers that. No → SaaS is the pragmatic choice.

If you answered "standard / no / no", buy Bullhorn or Vincere this quarter and stop researching.

If you answered "specific / yes / yes", custom is your answer and the math will back you up.

If you're mixed, that's the conversation worth having on a discovery call.

The honest bottom line

The recruitment software market does not reward indecision. Firms that have been "evaluating options" for 18 months are usually firms still running on spreadsheets, losing placements they didn't know they had (see 5 signs you've outgrown spreadsheets).

Bullhorn is a serious tool for a serious price. Custom on M365 is a serious tool for a different shape of buyer. The wrong answer is "let's wait six months and see." The right answer is to honestly assess where you fit, and pick.