How to Decide Who You Sell To, What You Say and How You Reach Them

Most sales plans are too big to use. They run to forty slides of segments, personas, channels and funnel stages, and when a deal stalls nobody can say which part was wrong. Sales has three levers and only three. Who you target, what you say to them, and the channel you use to say it. Every other sales problem is one of those three wearing a disguise.

Founders usually fix the wrong one. They change the message when the list was wrong, or buy a new channel tool when the message never landed. You can only tell which lever failed if you pull them one at a time and watch what happens, and the fastest way to watch is to be on the phone yourself. A founder who has had fifty real conversations with strangers in the target profile knows more about the market than any deck, and every sales hire after that inherits something true.

The goal: Produce a one-page Sales Triad, written in the prospects' own words and backed by your own call log, that a new salesperson can read in ten minutes and act on the same day.

Background

The triad is three short answers.

1. Who. The company and the person. The company is a profile you could hand to a researcher: the vertical, the size band, the systems they already run, and the event that makes them ready to buy. The person is a title and a seniority level, and the reason that person feels the problem personally. If you cannot name the title, you do not yet have a target.
2. What. The one problem, the one proof and the one ask, in a few sentences you can say out loud in under a minute. The problem in the prospect's words, the evidence that you solve it, and a specific small next step, usually a meeting.
3. How. The channel that reaches that title: phone, email, LinkedIn, events and webinars, referrals, or a partner. Name one primary channel. A second is allowed only once the first has results.

This play is the compact version of work the Customer Segmentation and Buyer Persona plays do at length. If you already have those, extract the triad from them and then test it here. If you do not, write your best hypothesis now. You will correct it against real calls within a month, which is faster than the persona research would have been.

One practitioner benchmark for a phone sprint, offered as a rule of thumb and not a promise: with a parallel dialer, roughly 5–8% of numbers connect, roughly half of those turn into a conversation of more than a minute, and 15–25% of conversations become a meeting. Your own numbers will move with your market. Record them in step 5 and use yours from then on.

Steps

1. Write the first draft of the triad on one page, in three boxes. Do not research first. The draft is a set of hypotheses with a name on each, and the sprint exists to test them.
   • Who: the company profile, the title and the trigger event.
   • What: the problem, the proof and the ask, each in one sentence.
   • How: the one channel, and why you think it reaches that title.
2. Build the list. Aim for about 2,000 contacts that fit the company profile and hold the target title, with a direct dial or mobile number for each. Buy or build it from a data provider and clean it before you dial. At the rule-of-thumb rates, a list that size gives you something like a hundred or more live connections, dozens of real conversations and a meaningful number of meetings, which is enough to learn from and not so much that you waste a market on a bad script. Check your numbers against the national Do Not Call registry and any state lists before you load them, and ask counsel how business-to-business calling rules apply to your market.
3. Set the sprint up as research. Load the list into a parallel dialer, a tool that calls several numbers at once and connects you to whoever answers first. Block 60 to 90 minutes a day for two to three weeks. Decide the stopping rule in advance: either about forty real conversations, or the end of the third week, whichever comes first.
4. Make the calls yourself. Open with the one-sentence problem, ask whether it is real for them and listen. After each conversation, before the next dial, write three lines in a shared log:
   • Who answered: title, company traits, anything that surprised you about the fit.
   • What happened: their exact words when they leaned in, and when they pulled back.
   • The outcome: meeting, follow-up or no.
5. Change one lever at a time. After about ten conversations, read the log. If the wrong people keep answering, change Who. If the right people answer and go flat, change What and use their phrase for the problem, not yours. If you cannot reach anyone, you have a channel or list-quality problem, so check the list before you blame the phone. Change one lever, run another ten conversations, and compare. Track the connect rate, the conversation rate and the meeting rate as you go.
6. Rewrite the Sales Triad when the sprint ends. Tighten Who to the profile and title that actually engaged. Replace the message with the words prospects used. Name the channel that worked, with the rates you measured. Keep a short list of the objections you heard, in their wording, and what you said that moved them.
7. Turn it into the working assets: the Sales Scripts play for the conversation, the Buyer Persona play for the people you now know, and the Sales Funnel Creation play for the stage conversions. Then hand the triad to your first sales hire and have them make their first fifty calls from it while you listen. Do not hand the sprint off until the meeting rate holds at or above your measured floor for two consecutive weeks of someone else's calls.
8. Give the triad an owner and review it every quarter. The sales lead owns the page and the log. Keep the log open. Once a quarter, make a hundred or so calls yourself against the current page, compare the numbers to the last quarter and change a lever if the numbers moved. Rerun the full sprint whenever you add a segment, change the product's core promise or change the price.

Notes

The sprint is about learning, not about meetings, though you will get some. Founders who treat it as lead generation stop when they have enough pipeline and never fix the message. Founders who treat it as research keep the log and carry the findings into everything else, including the website, the deck and the onboarding.

An automated email or AI outreach tool can send two thousand messages in a day, and it will tell you almost nothing about why nobody replied. A live call tells you in thirty seconds. Test automated channels against the triad afterward, with the Outbound Channel Test play, once you know what a good conversation sounds like.

Troubleshooting

My time is worth more than cold calls. Your time is worth most when you are the only person who can tell the market what you sell. Forty conversations is a fortnight of mornings. A year of a salesperson working from a bad message costs more than that in salary alone.

Nobody picks up. Check the list before the script. Wrong numbers, switchboards and untargeted titles produce silence. Try calling at different times of day, and test twenty numbers from a hand-built list against twenty from the purchased one.

I get conversations but no meetings. Then Who is probably right and What is wrong, or the ask is too large. Ask for fifteen minutes to compare notes on the problem, not a demo.

I already know my customers. Then the sprint will confirm it in a fortnight. Keep the log either way. The most common result is finding that your best customers share a title you had not written down.

My buyers are not reachable by phone. Some markets are not, and the numbers will show it by the second week. Then run the same sprint in the channel your buyers do answer, with the same log, and keep the one-lever-at-a-time rule.