It usually arrives by email in early April. The subject line is polite. The number inside is not. A founder who spent the year hitting revenue targets opens the return, sees a tax bill twenty or thirty thousand dollars larger than expected, and does the only thing that feels available in the moment: pays it, mutters something about doing better next year, and gets back to work.

Next year looks the same. That is the part worth sitting with. A surprise bill in April is rarely a tax problem at all. It is the visible symptom of a decision, or a non-decision, made eight months earlier, when there was still time to do something about it. By the time the return is filed, the money is gone and most of the levers that could have moved the number have already locked.

Three patterns account for the bulk of these surprises. None of them are exotic. All of them are fixable, and the fix has nothing to do with finding a cleverer deduction in March.

Mistake one: your CPA files the return, but does anyone plan it?

This is the most common failure mode, and also the easiest to miss, because on the surface nothing looks broken. Returns get filed. Deadlines get met. The accountant is responsive and pleasant. The relationship feels healthy right up until the bill says otherwise.

The tell is timing. If the only time you hear from your accountant is when documents are due, you have hired a preparer, not an advisor. A preparer records what already happened. An advisor changes what is about to happen, while the year is still open. There is a simple diagnostic for which one you are paying for: run through the list of strategies that should come up in any real planning conversation, things like entity structure, retirement plan contributions, the timing of large equipment purchases, and the Section 199A deduction. If none of those have been raised with you proactively, you are paying for preparation and receiving it under the label of planning.

Picture a Mesa contractor who doubles revenue in a strong build season. Their longtime accountant files the return on time, the way they always have, and the April bill lands tens of thousands of dollars higher than the year before, because nobody adjusted the quarterly payments or revisited the entity once the numbers changed. The work was fine. The timing was the problem. After the second surprise in a row, the owner does what they should have done a year earlier and starts looking for CPAs in Mesa, Arizona who lead with planning instead of filing, screening for one thing above price: does the firm talk about next year before this year’s return is even closed? The fee gap between a filer and a planner is real, and for most businesses past the startup stage it pays for itself several times over in a single cycle.

And the cost of getting this wrong compounds. Every year spent with a reactive accountant is a year of windows that opened and closed without anyone reaching through them. The deduction you could have documented, the retirement contribution you could have timed, the election you could have filed: each one had a deadline, and each one passed in silence.

Mistake two: treating estimated taxes as a rough guess

The IRS does not wait until April to get paid. It expects business owners to pay as they earn, through quarterly estimated payments. Anyone who expects to owe at least $1,000 for the year, after withholding, is generally on the hook for them.

Miss those payments, or lowball them, and the penalty machinery starts. The underpayment charge is pegged to the federal short-term rate plus three percentage points, and the IRS resets it every quarter. For early 2026 it has run at 7 percent, easing to 6 percent in the second quarter, compounded daily. That is not a parking ticket. On a meaningful underpayment, carried across several quarters, it quietly becomes one of the more expensive forms of borrowing a business can stumble into.

Two details trip people up. First, the penalty is calculated per quarter, so paying extra in September does not erase a shortfall from April. The clock already ran. Second, there is a safe harbor that makes the whole thing avoidable: pay at least 90 percent of this year’s tax, or 100 percent of last year’s (110 percent if your prior-year income was above $150,000), and the penalty disappears regardless of how the final number lands.

The businesses that get burned here are almost always the ones growing fast. A great year means last year’s safe-harbor figure no longer covers the real liability, and nobody updated the quarterly checks to match. Software like QuickBooks will tell you what you earned. But it will not tell you what to set aside, or warn you that your safe harbor has moved, unless someone configures it to and actually watches the trend. That someone is rarely the owner, who is busy earning the income in the first place.

Mistake three: choosing an entity once and never looking again

Most businesses pick a structure on day one, when revenue is theoretical and the priority is just getting the thing registered. Sole proprietorship or a single-member LLC, usually. Clean, cheap, fine for the moment.

Then the business grows, and the structure that fit at $80,000 of profit keeps running on autopilot at $250,000, where it can quietly overcharge the owner. The classic example is self-employment tax. Past a certain profit level, electing S-corporation status and paying yourself a reasonable salary can shave a real amount off that bill, because distributions above the salary are not hit with self-employment tax the way sole-proprietor profit is. How much it saves depends entirely on the numbers, and the salary has to be defensible, which is exactly the kind of judgment call a planner exists to make.

The broader point is that entity choice is not a one-time event. It is a setting that should be reviewed as the business changes shape. Plenty of owners never revisit it, because the DIY filing tools they started with, TurboTax and the like, are built to process the structure you already have, not to question whether it still serves you.

What the three have in common

Look closely and the same shape shows up in all three: a decision made once, then left to run while the business outgrows it. The accountant relationship set on autopilot. The estimated payment based on a stale number. The entity chosen for a company that no longer exists.

None of these get fixed in April. They get fixed in conversations during the year, when there is still room to act. The cheapest tax strategy almost always turns out to be the meeting you took in October, not the check you wrote five months later. For a business that intends to keep growing, that meeting is not an expense. It is the cheapest insurance on the books.