A firm that has never had a sales function does not have a hiring problem. It has an ordering problem. Almost every firm we watch attempt this does the steps in the wrong sequence, hires first, and then spends a year discovering that the missing piece was never the person.
The Sequence That Works
Build in this order. Write the offer and its price down. Decide what the first hire is actually for. Hire for the conversation you sell, not for a resume. Pay in a way that survives your sales cycle. Then manage against two numbers, not a dashboard. Everything else, including the tooling, waits.
We Are Not Going to Re-Argue Build Versus Outsource
This post assumes you have already decided to build.
If you have not, that is a genuinely different question and we answered it separately in should your firm build an internal sales team or outsource sales. Go and read that first if the decision is still open, because the arguments do not compress well and half-deciding is how firms end up with an expensive hire and no system.
We do this for a living as an outsourced sales team, so you would be entitled to expect this post to conclude that you should not build. It does not. Some firms should build, and the ones that succeed at it do the five things below in the order listed. A reader who follows this and never speaks to us has been served correctly.
Step One, Write the Offer Down Before Anything Else
If the founder is the only person who can explain what the firm sells and why it costs what it costs, there is nothing for a hire to sell.
This is the step that gets skipped because it feels like marketing rather than sales, and it is the reason most first hires fail. Write down what you sell, in what packages, at what price, and what the buyer gets. Then have somebody outside the firm read it back. If they cannot repeat it correctly, no salesperson will either.
Firms that sell hours have the hardest time here, which is worth naming rather than dodging. We wrote about the underlying shift in selling tax planning without competing on price, and the offer work is where a sales function either gets a foundation or does not.
Step One and a Half, Check That Demand Exists
This one sits between steps because firms skip it and then blame the hire.
A salesperson converts demand. They do not usually create it, and in professional services the person who can create it from nothing is rare, expensive, and not the person answering an ad for a first sales role at a firm. Before you hire, count how many genuine inbound conversations the firm had in the last three months. Referrals count. Website inquiries count. A partner’s golf conversation counts. Cold outreach that nobody has run yet does not.
If the number is under about two a month, a closer will sit idle, lose confidence, and leave, and the firm will conclude that sales hires do not work here. The honest answer at that point is that the constraint is demand and the money is better spent generating it.
If the number is decent and conversion is poor, you have found the case for hiring, and it is a strong one. Conversations arriving and not converting is the most fixable revenue problem a firm has, because the expensive half has already been paid for.
Step Two, Decide What the First Hire Is Actually For
There are three different jobs and firms routinely advertise all three in one posting.
There is the person who books meetings. There is the person who runs the conversation and closes. And there is the person who manages the other two. They are different skills, different pay, and different personalities, and asking one person to do all three produces somebody who does the easiest of the three all day.
For a firm with existing inbound demand, the closer is the first hire. For a firm with no demand at all, a closer will be idle and demoralised inside six weeks and the honest answer is that you have a marketing problem before you have a sales one. Decide which firm you are before you write the job ad.
Step Three, Hire for the Conversation You Actually Sell
Your buyer is a business owner making a considered, trust-heavy decision, often about money they are already worried about. That is not a transactional sale and a high-volume transactional seller will hate it.
The hiring context is tighter than most owners expect, too. The Bureau of Labor Statistics projects about 124,200 openings for accountants and auditors each year over the 2024 to 2034 decade, with a median annual wage of $81,680 in May 2024. You are recruiting in a market where the technical people you might have promoted internally are themselves in demand and well paid. Promoting your best manager into sales because they are good with clients is a common and expensive move.
Interview for the conversation. Have candidates run a real discovery call with you playing a real prospect, and score what they asked rather than what they said. The failure mode you are watching for is the candidate who advises instead of asking, because that is the person who will diagnose brilliantly for 40 minutes and close nothing.
Step Four, Pay in a Way That Survives Your Cycle
Firm sales cycles are long. A compensation plan copied from a company with a 14 day cycle will starve your hire through the exact months when they are learning.
We wrote the detail in how to pay the first salesperson you hire, and the summary is that firm owners consistently set base too low for the cycle they actually run, then interpret the resulting churn as a hiring problem. Work out your real average time from first conversation to signed engagement, and make sure the plan keeps somebody solvent and calm across at least two of those cycles.
Step Five, Manage With Two Numbers
Most firms build a dashboard nobody reads. Start with two numbers instead.
The first is conversations held per week. It is the only leading indicator that is entirely inside your hire’s control, and it tells you within two weeks whether the problem is activity or ability.
The second is the conversion rate from conversation to signed engagement. That one tells you which. High activity and low conversion is a skills and offer problem. Low activity and decent conversion is a demand or discipline problem. You cannot fix either until you know which one you have.
Add anything else later. When we run this as a service, reporting and measuring are two of the six components we deliver, and they exist because a sales function you cannot see is a sales function you cannot correct.
What Working Looks Like at 90 and 180 Days
Firms give up on this too early or far too late, usually because nobody agreed in advance what progress would look like.
At 90 days, working means the hire can run a full conversation unaccompanied, can tell you why a specific deal stalled without being asked, and is holding a stable number of conversations a week. Revenue at 90 days is close to meaningless in a market with long cycles, and judging on it produces the wrong decision in both directions. It fires people who were about to succeed and keeps people who got lucky once.
At 180 days, working means the conversion rate has moved and you can see why. By this point the offer has been tested on enough real buyers that the objections have stabilized, and either the rate is climbing or you have a specific, nameable reason it is not. A hire who reaches 180 days and cannot tell you which part of the conversation loses deals has not been supervised, and that is on the firm rather than on them.
Somewhere in the second quarter you should also see the founder’s calendar change. If it has not, the handover has not happened, regardless of what the pipeline says. That is the measure most firms forget to set, and it was usually the actual reason for building the function in the first place.
What Breaks Most Often
Four things, and none of them is the hire.
The founder does not actually let go. The hire is given the small deals and the founder keeps the good ones, which means the hire never learns the conversation that matters and the founder never gets their time back. We wrote about the handoff itself in founder-led sales and when to let go.
Nobody writes anything down. Everything the hire learns lives in their head, so when they leave, the firm restarts from zero. The firm sales playbook is the antidote and it is a document, not a project.
Delivery and sales never speak, so sales sells what delivery will not do, and the first three engagements are painful enough that the partners quietly withdraw support.
And the number is set from what the founder used to do. A founder closing warm referrals at a rate a new hire working cold will never see is not a benchmark. Set the first target from the pipeline that exists, then raise it.
A Note on Tooling
Firms building a sales function for the first time reliably start with software, because software is a purchase and the rest of this is work.
Resist it for the first quarter. A shared spreadsheet with one row per conversation, the date of the next step, and the reason it stalled will tell you everything the two numbers above need, and it will teach you what your process actually is before you pay to encode a process you have not designed yet. Firms that buy the system first end up with a system shaped like the software’s assumptions rather than their own.
Buy the tooling once you can answer three questions from memory. What are the stages a deal actually moves through here, what has to be true to move between them, and who updates the record. If you cannot answer those, no software will fix it, and most of them will quietly make the reporting worse by giving you a dashboard full of numbers nobody trusts.
If You Would Rather Not Build It
There is no shame in the other answer, and plenty of firms reach it after step one.
Standing up a sales function is a management job that lands on the partner group at the exact time they are least able to absorb it. If the appetite is not there, our results and the TRM CPA case study show what the outsourced version looks like from the firm’s side, including where the founder’s time went afterwards.
Whichever route you take, do step one either way. Nobody, internal or external, can sell an offer that has never been written down.
