Fusion CPAs

Whether you're taxed at 37% or 24% changes what a deduction is actually worth. A $50,000 write-off can save you up to $18,500 in your highest-rate year. That's the whole logic behind planning a spike year on purpose, not just filing it in April. I get into the moves that actually do this here: hubs.li/Q04wzk2Y0

We've been talking about RSUs a lot lately, and the planning conversation looks completely different depending on the company. If you're at a a high tech, privately held startup founded within the last 5 years, it's not a given that the company is going to be around in 20 years. The conversation there is about 22% vs 37% withholdings, whether to sell shares as they vest to cover the tax, or to hold onto as much stock as possible while the company's still privately held. Whereas, if you're at a company that's been around for 20 years with even more robust executive incentive packages such as deferred compensation plans where you reduce your salary now and get it years down the line when your tax rate might be lower. The RSU conversation may be leaning more toward diversifcation. More toward 37% and sell. Two very different RSU conversations at two different ends of the planning curve.

One of our clients had used a Kaiser Permanente plan for years. The company covered the full employee-only premium, and employees could add family coverage at their own cost. That worked well while the whole team worked out of the same state. Kaiser is a common choice for employers in its home markets, but it’s a regional plan. As of 2026, it operates in eight states: California, Colorado, Georgia, Hawaii, Maryland, Oregon, Virginia, and Washington. Members generally need to live or work inside a Kaiser service area to enroll. Outside those areas, coverage is generally limited to emergency and urgent care. By the time this client’s team had grown across several states, only about half fell within Kaiser’s service areas. The employees living outside that footprint were paying into a plan that didn’t give them practical access to routine care where they actually lived.  The owner’s question made sense: if the plan didn’t cover those employees properly anyway, why not let them buy their own policy and reimburse the premium? It sounds like a simple fix, but it's actually much more involved. 🔗 We dig in here: hubs.li/Q04tGz3T0

In 2026, the line item keeping CFOs up at night is not payroll or rent, it is the AI bill. LLM token spend doesn't budget like the software you're used to. Seats scale with headcount. Tokens don't. Route the same task to a different model and the cost can swing a lot, and a pilot that runs a few hundred dollars a month can turn into a five-figure line item the moment it ships to the whole team. Most companies adopted these tools faster than they built the controls to watch them. The fix isn't complicated: code every software and AI dollar to a vendor and department, set a budget at that level, and review it monthly. In our latest piece, we break down the four-step playbook, and why at the small and mid-sized business level, it's usually the controller who runs it, not a CFO. 🔗 Read it here: hubs.li/Q04tjW8P0