AI in Sales
AI Intake Agent or a Seasonal Admin? What Accounting Firms Should Price Out Before Next Tax Season
An honest, run-the-numbers comparison for accounting firms facing the tax-season lead surge: what a seasonal hire really costs, what a missed lead really costs, and when the human is still the right call.
by Jerrod Anthraper
Every December, the same conversation happens in accounting firms: "We need someone on the phones before tax season." So the firm posts a listing for a seasonal admin, interviews between year-end deadlines, and trains the new hire during the exact weeks the office is busiest. Meanwhile, a business owner who just fired her bookkeeper fills out the firm's contact form on a Saturday in February. By the time anyone sees it Monday afternoon, she's already had a call with the firm that answered Sunday morning. The engagement was worth $8,000 a year. Nobody at the firm ever knew she existed.
The leak: tax-season leads arrive precisely when nobody can answer them
Accounting is the most seasonal professional service there is, and inbound demand behaves accordingly. Inquiries spike from January through April — exactly when partners and staff are buried in returns and have the least capacity to respond. The cost of slow response is not abstract: 78% of sales go to whoever responds first, and the average B2B response time sits around 42 hours. During your busiest quarter, 42 hours is optimistic.
Worse, a large share of those inquiries arrive when the office is closed. Across service businesses, 40-60% of inbound leads come in after hours — evenings and weekends, which is when business owners finally have time to deal with their own books.
Why it happens: intake is everyone's side job
At most small and mid-sized firms, nobody owns lead response. The office admin answers when she's not processing e-file authorizations. A partner returns web inquiries when a return wraps early, which in March is never. The firm's capacity is sized for August; its demand arrives in February.
The traditional fix — a seasonal hire — has a built-in contradiction: you recruit in December, train in January, and by the time the person is competent, the season is half over. Then they leave, and next year you start again. It can still be the right call, but most firms make it without ever running the numbers.
How to actually run the comparison
Before next season, spend an afternoon on these five steps. They'll tell you whether you need a seasonal admin, an AI intake agent, or both.
1. Pull last season's lead log and timestamp everything
Export every inquiry from January through April: form fills, phone messages, emails to the info@ address. For each, note when it arrived and when someone first responded. Two numbers matter — your median response time, and the percentage of leads that arrived outside business hours. Firms that do this are usually surprised by both.
Don't forget the invisible column: hang-ups and abandoned calls. Your phone system logs calls that rang out or hit voicemail without leaving a message. Most of those callers didn't change their minds about needing an accountant — they just dialed the next firm on the results page. Count them as leads, because your competitors did.
2. Price the seasonal hire honestly
The wage is the smallest part. Add recruiting time, the hours a partner or manager spends training and supervising in January, payroll overhead, and the ramp period before the hire can competently qualify a prospect. Then note the coverage ceiling: a full-time hire covers 40 of the week's 168 hours. Your Saturday-morning business owner still goes to voicemail.
Factor in fragility, too. A seasonal hire who quits in mid-March — and some do, once they experience mid-March at an accounting firm — leaves you re-running this entire process during your worst possible month. Consistency has a value that never shows up on the payroll line.
3. Price the missed lead
Take your average first-year fee for a new client and your historical close rate on inquiries you actually responded to quickly. Multiply against the after-hours and slow-response leads from step one. This is the number the seasonal-hire conversation always skips: the cost of the leads you never really competed for. For context, leads contacted within five minutes are 21x more likely to qualify, and responding within 60 seconds can lift conversion by up to 400%.
Run the arithmetic on one realistic scenario before you dismiss it. A firm averaging $3,000 in first-year fees that missed twenty after-hours leads last season, closing even one in four with a fast response, left roughly $15,000 in first-year revenue on the table — recurring, if those clients stay. Swap in your own numbers; the point is to put a dollar figure next to the voicemails.
4. Be honest about what a human does better
A good admin handles things no intake automation should touch: the longtime client with a complex situation, the walk-in, the prospect whose "simple question" is actually a landmine. And if your new business is overwhelmingly referrals from existing relationships — people who will happily wait two days because their attorney vouched for you — a seasonal hire, or nobody at all, may genuinely be the right answer. The comparison only tips toward automation when a meaningful share of your leads are strangers who found you online and will not wait.
5. Run a split test for six weeks
If the numbers from steps one through four are ambiguous, don't decide on instinct. Put an AI agent on first-response duty for after-hours and weekend inquiries only, keep your human process for business hours, and compare consult-booked rates for six weeks. You'll get a clean read on what speed alone is worth to your firm before you commit to anything.
Track three metrics on both sides of the split: time to first response, percentage of inquiries that turn into a scheduled consultation, and the show rate for those consultations. The last one matters more than firms expect — a lead booked within minutes of reaching out shows up at a very different rate than one scheduled three days later, after the urgency has drained out of the problem. Six weeks of clean data will settle the argument better than any vendor pitch, including ours.
The proof
Tykon's framework has been through six-plus weeks of rigorous testing before we put it in front of clients, and the benchmark is public: above a 32% win rate, achieved with only 27% of the work compared to a human sales team — near parity with human performance on outcomes, but faster, more consistent, and scalable across every hour your office is dark. That last property is the one that matters for a seasonal practice: the agent's capacity in February is identical to its capacity in July.
Where Tykon fits
If your lead log says the leak is real, this is what we built James for. It answers every inquiry — form fill, missed call, web chat — in under 60 to 90 seconds, around the clock, qualifies the prospect against your criteria, and books the consultation directly onto the right person's calendar. No December recruiting, no January training, no April goodbye. We back it with a 90-day money-back guarantee, and we won't take the engagement unless we see a credible path to 3x ROI within 90 days — if your inquiry volume doesn't support that math, we'll tell you. A reasonable next step: run step one above on last February's inbox, and if the after-hours share is anywhere near half, bring the log to a 20-minute call and we'll run the rest of the math with you.