I could build you a calculator right now that proves we are always the cheaper choice...
Here is how I would do it. I would go find the highest DBA salary survey I could get my hands on, one of the ones that samples a couple dozen people in expensive metros and reports it as a national average, or take the 99th percentile salaries over on Brent’s salary survey. I would take that number and apply a “true cost of an employee” multiplier of 1.8 or 2.0. Then I would add health insurance and payroll taxes and retirement on top of that multiplier, even though the multiplier already had all of them in it. I would assume you replace the person every couple or few years and carry the full replacement cost as an annual line with recruiting costs thrown in. I would sprinkle in a downtime statistic from a survey of Fortune 500 companies and apply it to your environment universally. And I would land somewhere definitely north of $250,000 a year for one DBA, put our price next to it, and let the calculator do the rest.
That math exists. You can find it in about four minutes. Some of it is on hiring vs using a service calculators and webpages out there right now.
I’m not going to play that game here, though… Our first value is Trust360 and everything else we do sits on top of it. If I fudge the one number on this page that you can actually go check, why would you believe anything else I tell you about backups or corruption or whether your availability group is configured correctly?
So here is the honest version, and I will tell you how it ends right now, up front, before you scroll.
On a per-hour basis, hiring a DBA beats us. Every tier we sell. For most salary ranges. It’s not really close on that hourly basis, and it’s even further on our smallest plan, the Micro plan.
Which leaves the actually interesting question, which is why anybody hires us at all. Stick around for that part, because the answer is real and it is not “but you get 24/7 coverage!” This is part of our series of posts about Buying DBA Services. Trying to give a bit of a transparent look behind the scenes at where it makes sense to partner with a remote DBA shop like ours and where it may not.
What assumptions are we using?
I am putting these first, before the numbers, because every single one of them is a dial. Turn any of these and you get a different answer, and anybody who buries their assumptions at the bottom (or leaves them out entirely) is hoping you will not notice which dial they turned.
| Assumption | What we used | Where it comes from |
|---|---|---|
| Mid-level DBA base salary | $110,000 | BLS puts the median for database administrators at $104,620. Robert Half’s 2026 guide projects $95,500 to $137,500. ZipRecruiter’s 2026 average is $119,317 with the middle half between $108,500 and $131,500. The Brent Ozar salary survey has some really great details also. And it may make this number a bit higher depending on what you mean by mid-level. |
| Senior DBA base salary | $160,000 | A 20-year SQL Server specialist is not a $110,000 hire, and if you think they are, you probably haven’t tried to hire one lately. $160k gets us to around the 70th percentile of the 15+ year DBAs in Brent Ozar’s 2026 salary survey. And I think $160k is doing a disservice to the top-notch, most experienced folks, but that’s fine. I don’t want to inflate things here. |
| Employer share of health insurance | $20,143 | KFF 2025 Employer Health Benefits Survey, family coverage. Use about $7,885 if you are modeling single coverage |
| FICA (Social Security and Medicare – The employer part) | 7.65% | Applies to the full salary at both levels |
| 401k match | 3% | We’re at 3.5% – some are up to 4-5% went conservative here. |
| Workers comp, FUTA, SUTA | About 2% of wages | Varies a lot by state and it actually has a cap so the 2% may be high – but every state (as I’ve learned) is dramatically different |
| Equipment, software, training | $3,000 to $5,000 | Add in conferences and it can go high, and after hiring Richie Rump and looking at what AI has done to the price of memory, this is definitely a bit low… |
| Recruiting | Two versions shown | Either in-house cost per hire, or an agency at 22% of first year salary, amortized over a 4 year tenure (I do think DBAs tend to jump around – but in a bumpy economy they jump a bit less..) |
| Tenure | 4 years | Some folks build in 2-3 years, I went with 4. I see DBAs with all sorts of tenure out there. Seems like the great ones might jump around a bit more in a good economy, though. I did – but maybe I was always a consultant.. |
| Non-DBA time | 25% to 35% | How much “on the clock” time isn’t actually spent DBAing.. |
We did not apply a total compensation multiplier and then add benefits again on top of it. That multiplier already contains health insurance, payroll taxes, retirement, and paid leave. Doing both counts the same dollars twice and it is worth about $30,000 – $40,000 of fake cost.
And we did not put a downtime statistic anywhere in this article. You can find all sorts of surveys saying an hour of downtime costs high six-figures, and those surveys sample big enterprises. Using them to scare a mid-market company running four SQL Servers is a good way to make things seem more expensive.
And for real – check out the Salary Survey Brent Ozar does every year. The data is rich – and there are some outliers perhaps and its anonymous human submitted and he doesn’t edit or massage it so the numbers are raw – I think we could have made this be $125k for the mid level non Jr. DBA and maybe even $165/170k for the Senior DBA – more for superstars. But the numbers work fine either way.
What does a full-time SQL Server DBA actually cost?
Somewhere between about $149,000 and $214,000 a year fully loaded, depending on whether you are hiring a solid mid-level person or a genuine senior specialist. That works out to a multiplier of roughly 1.3 to 1.4 times base salary, not the 1.8 or 2.0 you will see quoted.
A mid-level DBA at $110,000 base
| Component | Annual |
|---|---|
| Base salary | $110,000 |
| FICA at 7.65% | $8,415 |
| Health insurance, employer share, family | $20,143 |
| 401k match at 3% | $3,300 |
| Workers comp, FUTA, SUTA | $2,200 |
| Equipment, software, training | $4,000 |
| Recruiting, amortized over 4 years | $1,200 to $6,100 |
| Fully loaded | About $149,000 to $154,000 |
A senior DBA at $160,000 base
| Component | Annual |
|---|---|
| Base salary | $160,000 |
| FICA at 7.65% | $12,240 |
| Health insurance, employer share, family | $20,143 |
| 401k match at 3% | $4,800 |
| Workers comp, FUTA, SUTA | $3,200 |
| Equipment, software, training | $5,000 |
| Recruiting, agency at 22%, amortized over 4 years | $8,850 |
| Fully loaded | About $214,233 |
Salary is the big number and it stays the big number. The benefits and taxes and recruiting are real and they matter, but they do not double your cost.
How many hours of actual DBA work do you get?
This is where the interesting part is, and it is the number nobody wants to talk about.
| Hours | |
|---|---|
| 52 weeks at 40 hours | 2,080 |
| Less 15 vacation days | -120 |
| Less 4 sick days | -32 |
| Less 11 holidays | -88 |
| Hours they are actually at work | 1,840 |
Now take out the part of the week that is not database work. Standups, status meetings, the security questionnaire somebody needs filled out, the ticket that turned out to be a network problem, helping with the thing that is not their job but nobody else can do it either, the interview loop for the other open req, annual training, and the twenty minutes every morning that go somewhere nobody can account for.
I’ve been the full-time DBA in the office. And I lived it.. The time goes fast – hallway meetings, getting sucked into a debate, idle chat. I’m also not adding the “extra hours” that many of us DBA types work. The math is rough here – but I’m going somewhere rather generic here.
At 25% non-DBA time you get about 1,380 hours of actual database work in a year. At 35% you get about 1,200. If you think those numbers are too generous or too harsh for your shop, move them, that is fine, the point is that you are not buying 2,080.
So what does an hour of real DBA work cost you?
| At 1,380 hours | At 1,200 hours | |
|---|---|---|
| Mid-level DBA, $149,000 loaded | $108/hr | $124/hr |
| Senior DBA, $214,233 loaded | $155/hr | $178/hr |
What does DBA as a Service cost per hour?
Our pricing is on the website and has been since we started. Transparency saves everyone time.
| Our plan | Annual at list | Quarterly Hours * 4 | Effective cost per hour |
|---|---|---|---|
| Micro | $24,000 | 40 | $600 |
| Silver | $40,000 | 100 | $400 |
| Gold | $56,000 | 180 | $311 |
| Diamond | $76,000 | 300 | $253 |
| 20 hour bucket | $7,000 | 20 | $350 |
Is it cheaper to hire a DBA or use a managed service?
Per hour, hiring wins, and it is not really that close of a call.
| Option | Cost per hour of DBA work |
|---|---|
| Mid-level employee | $108 to $124 |
| Senior employee | $155 to $178 |
| Us on Diamond | $253 |
| Us on Gold | $311 |
| Us on Silver | $400 |
| Us on Micro | $600 |
There it is. We are two to four times more expensive per hour for our most expensive hourly (and lowest priced per quarter) plan, and I do not have a clever way to make that table say something else, so I am not going to try.
If you want to stop reading here and go post a job req, I understand, and honestly for some of you that is the right move and we’ll talk more about that in a bit and I already posted a couple weeks ago about clients where we aren’t the best fit.
Why does Micro cost more per hour than Diamond?
I’ve been posting a lot about the “business of business” and I posted last week a bit about how our levels work so I won’t go to deep here, but I’m sure you may be like “uhh why is the Micro, lowest priced quarterly plan the most expensive?!?” Because there is a floor under every engagement we take, and the floor costs us about the same whether you buy 10 hours a quarter or 75.
Here is everything that runs on your environment regardless of which plan you are on, and whether it eats your hours:
| What runs no matter which plan you buy | Does it consume your hours? |
|---|---|
| Skopos, our monitoring platform, deployed and hosted and maintained | No (though building in filters and tweaking alerts for requests does consume hours) |
| One SQL Sentry license included | No |
| A daily health check produced and reviewed by a human on our support team, every business day | No – we don’t bill for the review of that report – the creation of the tickets from it and the priority escalation. Our billing starts when we start actually connecting and working with you. |
| Alert triage and review, Monday through Friday, 8 to 5 Eastern | No – reviewing alerts, jumping to the P1s and creating tickets – like the health checks daily – isn’t billed. |
| The on-call rotation, including what we pay the DBA who takes the week and the answering service and all of that. | No (though we do charge a fee per access, waived for Gold and Diamond clients for 1 or 2 calls per year – and we really work our tails off to be proactive and not get calls – but I still have to pay one of our DBAs to be on-call just in case – and we have backups and a whole process for that.) |
| The plumbing behind a 15 minute P1 response, the answering service, the escalation chain, the backup escalation | No |
| Your named lead DBA and internal accountability and internal management and internal leadership review on accounts. | No (though the DBA is billing when they are working or talking with you, naturally) |
| Actual work on your servers | Yes, billed in 15 minute increments |
Spread that base cost – especially the tooling and monitoring and reviews across 40 hours a year and it is brutal. Spread it across 300 and it nearly disappears. Micro’s $600 an hour is not us charging a premium for expertise at the low end, it is a small number of hours carrying a fixed cost that does not shrink. We don’t sell monitoring only – If we did, we’d some of that fixed price would move out of the plans and the effective hourly number would look quite a bit lower.
Which leads to a recommendation that costs us money: if you are evaluating Micro on an hourly basis, you are evaluating it wrong. Micro exists to buy you monitoring, daily eyes, and a phone number that a full senior heavy team of US based FTE DBAs answers, with a small allotment of hours attached. If what you actually want is hours, and not monitoring, buy a bucket instead. It is $350 an hour and it comes with none of that other stuff, and we will tell you that on the call, because we would rather sell you the right thing.
So why on earth would you hire us?
Because the per-hour number is the wrong question for most of the companies reading this.
Per hour, we lose. Per year, when a company doesn’t have sustained heavy DBA needs, we start to really win.
| Annual cost | |
|---|---|
| Mid-level DBA, fully loaded | $149,000 |
| Senior DBA, fully loaded | $214,000 |
| Us, Gold (180 hours, plus monitoring and on-call) | $56,000 |
| Us, Silver (100 hours, plus monitoring and on-call) | $40,000 |
The employee is more efficient per hour. They are also a fixed cost, and you buy all 1,180 hours whether you need them or not. In fact, when I left the managed services company I was working for trying to build a SQL Server practice with them – I realized we just weren’t going to do it. Most IT MSPs have a harder time justifying a full-time SQL Server DBA services practice. The rates are higher, the size of the team to really support is a bit high and the sales teams aren’t used to selling it. I saw the writing on the wall that we just weren’t growing so I talked to the partners who take a chance on bringing me in. I did the math. I said you don’t use me for a lot of billable hours. So I’m teaching, blogging, marketing, hanging out, and just kind of being around a lot of the time – which means when I do work – my hourly cost is way higher than the math here shows – and my salary was on the lower end. I was ready to be foolish enough and go off on my own and I made a plan where I would leave – they’d make a REALLY good deal on me per hour – and I had one really low rate if I was training or just doing biz dev with them – one higher (but still absurdly low) rate if I was billing. They’d be my anchor client – and get first dibs on my time and for that we’d have a 3, 6 and 9 month guarantee of my availability and their billing. They needed DBA as a Service – I just didn’t know it would be a thing then – and I gave it to them (as a solo person – which means coverage and support is tough when I got busier and had to juggle clients – it’s good to have a team now.)
Where is the actual crossover?
This is the number I would want if I were you, so here it is.
Start with Gold or Diamond and add overage as you go. Depending on the level of resource and the salary range and your details – somewhere between 400-800 hours a year is where things start to cross over to “we’re definitely more expensive”
Translate that to a week and if you consistently need more than about 10 to 15 hours a week of real, hands-on database work, hiring somebody starts to outright win on cost if that’s the only metric you care about.
Note the word “real.” Not 15 hours of somebody being available. Fifteen hours of actual database work. Many of the companies I talk to who are convinced they need a full-time DBA are looking at a backlog, not a regular workload, and those are different things. Clear the backlog in year one or a half a year and then ask yourself what year two looks like.
The four other reasons, which are not about cost
You are buying one person’s skill set when you hire. Nobody is simultaneously an availability groups expert, a replication expert, a query plan person, an Azure architect, and a security specialist. They are strong at two or three of those and learning the rest on your time.
| The thing you need twice a year | One person | A team |
|---|---|---|
| An AG design reviewed before you commit | Whoever you hired, probably learning unless that was their strength | Somebody who has done dozens – and we’ve seen them all fail in really unique ways. (Cue the farmers insurance jingle – we know a thing or two… because we’ve seen a thing or two.) |
| Replication upgraded without reinitializing eight subscribers | Blog posts at 11pm and some AI suggesting DMVs that may or may not exist | Someone who supports all kinds and contracts with a guru whose specialty it is |
| A query plan that makes no sense | Depends entirely who you hired | The person who used to teach the tuning course to MCM candidates or who is well known among our team as the queen of query performance – plus a strong set of others who are just as good as Andy and Tara at much of this. |
| An Azure architecture and cost review | Probably not their background | A former Microsoft data platform guy who was on the team to help launch Azure in our CDO and Azure Architect, Buck. (And we cross train – all learning from the firehose of knowledge that he is.) |
They go on vacation, and then they leave. One DBA is capacity, not coverage. We wrote a whole thing about the perils of being a solo DBA and I will not repeat it here, but the short version is that one is none, two is one.
When things go badly wrong, we flex, and we try not to profit from your bad week. Here is our actual policy – a client can leave us with 30 days’ notice (no cell-phone contracts – why would I want to keep a bad fit client hostage to a contract) and each quarter we evaluate and pick the level that’s right for the hours needed for a client. If someone needs more hours in a quarter once it’s picked they use overage.
In reality, though we like to be flexible and lean into our Generosity value. When it’s just a bit of overage, it makes sense to just make everyone’s life easier and not make the AP folks at our clients hate us and confuse them with “mid-quarter upgrades” – but we also care way more about long-term relationships. Our clients simply stay with us – for a really really long time – like our team – we will often work with a client and maybe upgrade next quarter to the higher level and let them steal hours ahead or, if the need arises soon enough, look the other way and let someone upgrade the current quarter. Saving overage for the smaller one offs.
| Situation | What we do |
|---|---|
| You need more hours in a quarter than your plan includes | We sell overage at your plan’s rate |
| It is a real emergency and the overage is going to be ugly | We will often do a courtesy mid-quarter upgrade or upgrade the next quarter and let someone “steal from the future” instead, because that costs you less than the overage would |
| Why we do that | Because we do not want an incentive that pays us more the worse your month goes. |
You will pay more in a quarter where your environment tried to fall over. You are paying for the hours that saved it. And in most years, added up, our clients running anywhere from one instance to scores of them have spent less than a hire would have cost, without giving up a phone number at 2am. So our base is a lot less than the full-time if the hours each week aren’t high – and in a “oh no” flex situation – you may have a more expensive month, but you still work out better by the year.
Sometimes the answer is both, and this is the option most articles skip. Keep your DBA. Put a team behind them.
| Need | Your DBA alone | Your DBA plus us |
|---|---|---|
| Vacation coverage | None/Training backups and documentation | Covered |
| Escalation when they are stuck | Google, or a forum post, Ask your nearest LLM | Named senior DBAs who have probably seen it – and our monitoring has probably already identified the P1 and we’re probably already starting up the P1 meeting. |
| Specialty depth | Whatever they happen to know | The team’s knowledge |
| What it does for them | Isolation, and eventually burnout if its busy enough | Peer relationships with MVPs, authors, community tools folks, mentors and 25 year specialists, and somebody to check their work and teach them |
| If they resign | Your environment walks out with them | Continuity |
| Cost | Their fully loaded salary | Their salary, plus the right plan for your needs |
That last row is worth sitting with. A capable mid level or Junior in-house DBA plus a Silver plan runs about $150,000 – $180,000 all in. Hiring a second DBA for coverage runs whatever the loaded cost for the level you need is. And your existing person gets a peer group in their corner.
When should you hire a full-time DBA instead of outsourcing?
Hire when most of these are true:
- You genuinely have 15 or more hours a week of hands-on database work, sustained, not a one-time backlog
- A real chunk of that work is development, ETL, or data engineering, which we are frankly too expensive to do 40 hours a week with a managed DBA services company
- You want somebody in your daily standups, in every change control meeting and IT team meeting, and/or in the building
- Deep knowledge of your specific application matters more to you than breadth across database specialties
- You can hire well, and you have a plan to keep that person interested after the backlog is gone
Be careful about hiring when:
- The pile of work you are looking at is remediation, not steady state
- You are hiring for coverage instead of capacity, because one person is not coverage
- You cannot answer who mentors them, who reviews their work, or who covers their vacation
And if you read all of that and land on “we should hire,” then hire. Tell us and we can even sell you a bucket of hours good for the year and help you write the req, sit in on the technical interview, and if you want, review their work quarterly so they have somebody senior to argue with. We would rather have that relationship than sell you a plan you should not have bought.
The short version
Per hour, we cost two to four times what an employee costs. That is true and it is on this post in a table above.
Per year, if you don’t have enough to keep a DBA sufficiently busy, we can cost a fraction of a hire and you get monitoring, a team, and coverage when somebody is out that can still flex up for the emergencies.
Above about 10 to 15 hours a week of real,sustained, DBA work, go hire someone, but we can still help provide that coverage and expertise – maybe you can hire a skillset down and we can level them up with you and they stay a bit longer.