Your Payroll, Not the IRS Limits, Picks Your Retirement Plan

The most common retirement planning question small-business owners ask me is some version of: "What is the maximum I can put away?" It is a fair question, and I understand why it comes first - the IRS publishes those ceilings in big clear numbers every November, and comparison charts are everywhere. But the ceiling is not what gets most owners into trouble. The floor is.

The floor is what your payroll can actually sustain when you hire employee number three, or when Q4 comes in twenty percent light, or when you realize the plan you chose two years ago now legally requires you to fund it for four people instead of one. That math is what comparison charts skip - and where most plan design mistakes actually happen.

I have worked with business owners who started with the right plan and ended up with the wrong one - not because the rules changed, but because the headcount did. The plan stayed exactly the same. The payroll obligation did not. What follows is a look at how each major plan type behaves as a business grows - not just what the IRS allows, but what your payroll will actually demand once employees are in the picture.

Why the Contribution Limit Is the Wrong Starting Point

If you search "best small business retirement plan," you will find the same table in different clothes: how much can the plan holder put in? The SEP-IRA and Solo 401(k) both cap out near $69,000 for 2025. The SIMPLE IRA allows $16,500 in employee deferrals with a modest employer match. The Safe-Harbor 401(k) lands in the same neighborhood as the SEP. Put them side by side and the SEP and Solo 401(k) look like obvious winners - higher ceilings, simple setup, no administration overhead.

The table is not wrong. It is just incomplete. It shows you the ceiling when what you need to understand is the floor. The ceiling tells you the most the IRS will let you shelter. The floor - which almost no comparison chart mentions - tells you the minimum your payroll must fund for other people once you have employees who are also eligible for the plan. Those are two very different numbers, and as a business grows, the gap between them matters far more than the IRS limit itself.

That distinction becomes significant once your business grows past one person. A SEP-IRA at 15 percent of your own compensation is a personal savings choice. A SEP-IRA at 15 percent of compensation for yourself plus four employees earning $60,000 each requires $36,000 in additional employer contributions before you have changed anything about your own account. Same plan name. Completely different financial commitment. Same is true for every plan type - the IRS limit is static, but the employer obligation grows with every hire.

The sustainable plan is the one you can fund in a slow year, not just in your best year. That is the standard I use when working with clients on this decision. A business owner who picks a plan based on the highest limit she can contribute in a strong year - without modeling the employer cost at five or eight employees - is setting up a problem she will feel at exactly the wrong time: a lean year, with a payroll obligation she cannot easily skip.

What follows is an honest look at how each major plan type actually behaves as a business grows, and what to do when the plan you started with stops making sense.

What a SEP-IRA Really Costs When You Add Employees

The SEP-IRA is the most popular first retirement plan for self-employed owners, and the reasons are easy to understand. No annual administration cost.

No non-discrimination testing. No third-party administrator to pay. You set a contribution percentage each year - or you contribute nothing at all. If revenue is strong, you put in more. If it is a hard year, you skip entirely. That flexibility is genuinely valuable when your cash flow varies, which is the default condition for most solo operators and small consulting businesses.

Here is what the simple version of that pitch leaves out: the same percentage you contribute for yourself must apply to every eligible employee. Eligibility under a SEP-IRA requires being at least 21 years old, having worked for you in at least three of the last five years, and earning at least $750 in the current year. Once an employee meets those criteria, your contribution rate applies to them at the same level it applies to you. There is no way to selectively exclude eligible employees or use a lower rate for non-owners.

The math follows directly from that rule. At a 15 percent contribution rate with four employees averaging $60,000 in compensation, your employer cost for those employees is $36,000 - before you have contributed a single additional dollar to your own account. Run the number at 20 percent and it gets larger. As one self-identified CPA noted in a widely read discussion among small-business owners on Reddit: choosing a Solo 401(k) over a SEP "indicates that you absolutely have no plans for employees in the future." The implication runs the other direction too - once employees are part of your plan, the SEP-IRA's economics look very different than they did when you were the only participant.

The SEP-IRA's genuine advantage - full flexibility to skip contributions in a lean year - does survive as the team grows. If you contribute nothing, you contribute nothing for everyone. That remains a meaningful safeguard. But the decision to skip a year also means every eligible employee receives nothing that year, which has retention implications that pure tax math does not capture. A plan that protects your cash flow at the cost of your team's retirement savings is a tradeoff worth naming clearly before you commit to it.

My read: a SEP-IRA is excellent for a solo owner or a business with one or two employees where the owner is comfortable with the cost scaling as the team grows. Once you have three or more employees and want to contribute meaningfully in most years, the economics of a different plan often become more attractive - even if the limit looks lower on paper.

The SIMPLE IRA: A Stable Team's Best Friend (With Limits)

The SIMPLE IRA was designed for businesses with 100 or fewer employees, and it solves one of the SEP-IRA's core problems: the employer cost is predictable and bounded before the year starts. You choose between two structures. Option one: match employee contributions dollar-for-dollar up to 3 percent of compensation. Option two: contribute 2 percent of compensation for every eligible employee, whether they participate or not. Either way, you know exactly what you are committing to in January rather than discovering it at year-end.

That predictability is the SIMPLE IRA's strongest argument. At a 3 percent match, an employee earning $60,000 who participates costs you no more than $1,800 per year. The cost scales directly with headcount and pay levels - no eligibility-window surprises, no percentage-of-everyone calculation that compounds as the team grows. One small-business owner who switched to a SIMPLE IRA after running a 401(k) for years described the appeal simply: it was just "simple and easy" with lower costs to manage than the plan it replaced.

The tradeoff the SIMPLE IRA asks for is a lower employee deferral ceiling. In 2025, the limit is $16,500, compared to $23,500 under a traditional 401(k). For most employees - and most owners earlier in their careers - this is largely theoretical, since the majority of people are not maxing either limit. But for an owner who wants to shelter the maximum possible as the business matures, the SIMPLE IRA's lower ceiling becomes a genuine constraint.

There is also a notice requirement that can create friction in volatile years. If you want to reduce the employer match, the IRS requires at least 60 days' advance written notice to eligible employees, and you can only make this change once per year. A business cash crunch rarely waits for a 60-day notice window. If your revenue moves faster than your plan design allows, the SIMPLE IRA's predictability cuts both ways - it is a commitment that does not bend easily when you need it to.

The SIMPLE IRA also imposes a two-year lockup on early withdrawals. Employees who take money out in the first two years of plan participation face a 25 percent penalty rather than the standard 10 percent. Clear communication at enrollment prevents most problems here - but it is worth stating plainly, because employees who face unexpected penalties rarely attribute that outcome to plan design.

Where the SIMPLE IRA earns its place: a business with a stable team of two to ten people, predictable enough revenue to commit to a defined match, and an owner who is not yet trying to maximize personal deferrals. In that scenario, it is often the most cost-efficient plan available.

What Will Matter Most in the Next 12 to 24 Months

Two things are converging that will put retirement plan design back on a lot of business owners' agendas over the next couple of years, and neither of them is primarily about contribution limits.

The first is SECURE 2.0, which was signed into law in late 2022 and has been rolling out in phases ever since. Starting in 2025, businesses with fewer than 50 employees that start a new 401(k) can receive a tax credit of up to $5,000 per year for the first three years to offset startup costs - plus an additional credit for employer contributions made during those years. That changes the economics meaningfully for small businesses that previously ruled out a 401(k) on cost grounds. The math that made a SIMPLE IRA look cheaper when you run the numbers may now favor the 401(k), once the credits are factored in. If you adopted a SIMPLE IRA in the last few years partly because a 401(k) felt too expensive to set up, it is worth running the comparison again with current numbers.

The second is labor market pressure. According to a 2023 ADP survey, 60 percent of employees rank a retirement savings plan in their top three workplace benefits. That number has been climbing, particularly in the professional services, healthcare, and technology sectors where small businesses compete hardest for skilled talent. A business that offers only a SEP-IRA - where the employee makes no deferral and all contributions are at the owner's discretion - is now at a noticeable disadvantage compared to one that offers a matching 401(k). The employer math still matters, but so does the recruiting math. A plan your employees can contribute to and see matched is a different tool than one that exists mainly for the owner's benefit.

Neither of these forces requires an immediate change. But both are reasons why the plan that made sense when you last thought about it may not be the plan that makes sense for the business you are running now - or the one you will be running in two years.

Solo 401(k) vs. Safe-Harbor 401(k): Two Very Different Problems

The Solo 401(k) offers the highest contribution ceiling relative to income of any plan available to a self-employed person with no employees. As both employer and employee, you can contribute up to $23,500 as the employee deferral and up to 25 percent of compensation as the employer side, up to a combined $69,000 for 2025, or $76,500 if you are over 50. For a self-employed professional earning $100,000, the Solo 401(k) often shelters more than a SEP-IRA at the same income level, because the employee deferral applies before the 25 percent employer calculation - a math advantage I covered in more detail in a separate comparison of SEP-IRA vs. Solo 401(k).

The catch is structural. The Solo 401(k) requires that you have no full-time W-2 employees other than your spouse. The moment you hire one, the plan is disqualified. There is no grace period, no transition window, no grandfathering. You terminate the Solo 401(k) and move to a plan that accommodates employees. For someone who will remain truly solo indefinitely, it is an excellent vehicle. For anyone who expects to hire within the next year or two, it is a plan you are choosing in anticipation of replacing it - and that replacement needs to be ready before the hire, not after.

The Safe-Harbor 401(k) exists to solve a completely different problem: non-discrimination testing. Under a standard 401(k), the IRS runs annual tests to confirm that highly compensated employees (generally those earning over $155,000) are not benefiting disproportionately compared to lower-paid employees. If the plan fails, the owner has to receive some contributions back - sometimes well after year-end. That is an accounting problem nobody enjoys, and it creates uncertainty about how much the owner can actually contribute in a given year.

A Safe-Harbor 401(k) bypasses the testing entirely by requiring a mandatory employer contribution for all eligible employees - either a 3 percent non-elective contribution to everyone's account, or a match of 100 percent on the first 3 percent of deferrals plus 50 percent on the next 2 percent. Meet that threshold and the owner can contribute the full personal deferral without worrying about what participation looks like across the rest of the team. The administrative cost typically runs $1,500 to $5,000 per year depending on the provider - a real fixed expense that does not scale down when revenue does. For a business where overhead is stable even when revenue fluctuates, that cost is manageable. For a business with highly seasonal or project-dependent cash flow, it is a commitment to model carefully before signing up.

How to Read Your Payroll as a Plan Decision

The question I find most useful is not "which plan has the best limit?" It is: "What is the most expensive year this plan can have, and can the business sustain that?" Not the most expensive theoretically possible year under the IRS ceiling. The most expensive realistic year given your current and expected headcount, your compensation levels, and the contribution rate you are planning to set.

For a SEP-IRA, that means calculating your employer cost at your target rate across every eligible employee you expect to have three to five years from now - not just who is on payroll today. SEP-IRA eligibility is based on historical service, not current status. An employee who worked part-time two years ago and is now full-time may already be eligible. If your team has been with you for several years, most of them likely are. Run the number at your future headcount before you commit to a contribution percentage that seemed manageable when you were the only one in the plan.

For a SIMPLE IRA, the math is more direct. Your employer cost at a 3 percent match is simply 3 percent of total eligible payroll - a number you can model immediately and compare against your expected revenue range. The question is not whether you can afford the match when business is good. It is whether you can sustain it during a year when revenue drops and you are looking for places to reduce expenses. The 60-day notice requirement means the match is not something you can reduce on short notice - which is another way of saying it is a commitment that the payroll needs to be able to carry through uncertainty, not just through growth.

For a Safe-Harbor 401(k), the question is whether the mandatory employer contribution and administrative cost can be treated as fixed overhead - not a discretionary line item that gets cut in a hard year. The annual administration fees do not scale down with revenue. They are as fixed as your payroll processing cost or your office lease. If your business has overhead that holds up even when top-line revenue fluctuates, that is a plan you can commit to confidently. If your revenue is volatile, the Safe-Harbor 401(k) deserves careful modeling before you adopt it.

One more thing I always tell clients, because it tends to stop people from making good decisions: the plan you start with is not the plan you are stuck with. Moving from a SEP-IRA to a SIMPLE IRA, or from a SIMPLE IRA to a Safe-Harbor 401(k), involves timing requirements and administrative steps - a SIMPLE IRA must be formally terminated before November 1 to allow a new plan to take effect the following January - but none of these transitions are as difficult as running the wrong plan for five years because switching felt too complicated to think about. The earlier you build in a review trigger, the more options you have.

Retirement plan decisions feel like a one-time setup task, and a lot of business owners treat them that way. Pick a plan, open the account, contribute when cash flow allows, revisit it never. That works fine until the business changes - and businesses change. Headcount grows. Revenue gets lumpy. You hire someone who has been part-time for three years and suddenly find out they are SEP-eligible. Or you try to cut the SIMPLE IRA match during a slow quarter and learn you needed to give 60 days' notice to do it.

The plan that fits your business today should be revisited any time your headcount, revenue structure, or compensation profile changes meaningfully. That does not mean changing plans every year - transitions have their own costs and timing constraints. It means having a working model of what your employer cost looks like at your current and projected payroll, and checking that model against reality periodically. The contribution limit tells you the best possible year. The payroll tells you what the typical and worst-case years actually look like. Both numbers belong in the decision.

← All articles