Make owner-led sales a system with AI: 4 steps and the first to hand over
Split owner-led sales into 4 steps and hand AI the step that stalls in busy weeks first, so selling keeps going.
Key points
- Split sales into 4 steps, review last
- AI researches; the owner writes the why
- Start AI where you stall in busy weeks
- Feed reply reasons back into targeting
Contents
Look at owner-led sales as 4 steps
A&A perspective
Seen as four steps (targeting, research, the proposal and reviewing replies), the owner's sales work shows where it stalls. Systematizing sales, as used here, means running those four steps at a fixed time each week and handing the research and drafting within them to AI.
A&A perspective
In targeting, make sure you can say in one line why you are contacting this company now. That line is your reason for picking it. Research means checking the company's business and recent moves in public information. The proposal is the first message or proposal document, written from that reason and the facts you found.
A&A perspective
Reviewing replies means looking at who answered, who did not and who declined, and why, then changing whom you pick next. Without this fourth step, you miss in the same way every week.
A&A perspective
In a company where the owner still carries most of the selling, all four steps get squeezed into the owner's spare hours. In a week when the core business gets busy, one of them stops, and the steps after it stop too. The company then slides back to waiting for referrals or for an outsourced sales agency.
A&A perspective
Automate & Augment (A&A), the company writing this article, thinks AI adoption should not start with sending more messages. Having AI write far more messages does not give you more reasons to pick a company.
Hypothetical example
On a weekly calendar, the four steps might look like this: 30 minutes on Monday morning to choose this week's companies, research on Tuesday, proposals on Wednesday, and 15 minutes on Friday to review replies. The days and times are only a hypothetical example.
AI does the research; the owner writes the reason
A&A perspective
Of the four steps, hand research and drafting to AI, and keep the reason for picking a company, and the decision to send, with the owner. Two overseas examples use AI in sales in a way close to this split.
From the sources
Anthropic, a US company that builds AI, has published a customer story about Lindy, a North American company. Lindy offers a service that lets each business build its own AI to carry out work on its behalf.
From the sources
According to the story, a software startup hands two jobs to Lindy's AI, finding prospective customers and tailoring the first outreach message to each one, and books more than 40 meetings a month. Its sales team can spend the freed-up time on high-value conversations and on important accounts, the story says.
From the sources
The story also says that teams using it can hand routine work to AI while keeping control of strategic decisions themselves.
From the sources
The other example is Lovable, a Swedish company whose service builds apps from plain-language instructions. Clay, a service that uses AI to research the companies and people you sell to, published it as one of its own customer stories.
From the sources
At Lovable, when a prospect submits the enterprise inquiry form, AI looks up the industry, the person's role, whether a relationship already exists, and what the company is likely focused on now. For every meeting, the salesperson receives a brief covering the company's profile and the problems it is likely trying to solve.
From the sources
Lovable also researches job candidates the same way with AI when hiring. Even then, recruiters read what the AI found and make every decision on whether to reach out themselves, the story says.
A&A perspective
A&A thinks the owner's sales should be split along the same line. Hand research and drafting to AI; the owner writes in one line why to contact this company now and decides whether to send. If you cannot write that line for a company, do not send to it this week.
A&A perspective
Both examples are customer stories published by the vendor, not figures checked by a third party. They come from companies with several salespeople, so adapt them to a week that the owner runs alone.
Pick the first step by what stalls in busy weeks
A&A perspective
Pick the first thing to hand to AI from the step that stalled in a busy week. Where it stalls is where the owner's time runs shortest.
From the sources
The Lovable employee in charge of building AI-driven work processes says that when he spots a bottleneck, he experiments a bit and sees whether it works. The story describes the approach as testing small and adding volume once something works.
A&A perspective
A&A thinks this approach carries straight over to an owner's sales. Recall a busy week last month and name the one step, of the four, that stopped first.
A&A perspective
Every step has a research or drafting part you can hand to AI. If targeting stalled, for example, have AI build the list of candidate companies while the owner still writes the one-line reason. If research stalled for lack of time, hand over the research; if the message never got written, hand over the proposal draft.
A&A perspective
If you choose research, how to hand prospect research to AI and test it is covered in a separate article, "Give AI the prospect research first, not the email: test it on 10 companies".
A&A perspective
After handing a step over, compare the time it takes for about two weeks against the time before. Also count how often the AI's draft was wrong. A step with many errors can always go back to the owner doing it by hand.
Hypothetical example
Here is a hypothetical way to see whether it pays off. Suppose prospect research took the owner 3 hours a week and, now that the owner only checks AI drafts, it takes 1 hour. That frees 2 hours a week, about 8 hours a month.
Hypothetical example
Price the owner's time by the revenue it would bring in if spent on the core business. Multiply those 8 hours a month by that price, then subtract the AI subscription and the time spent on the initial setup. Set the result next to what you now pay a sales agency or spend on hiring, and you get a rough sense of how many months it takes to pay off.
Feed reply and decline reasons back into targeting
A&A perspective
Once a week, gather the reasons behind replies and declines, and fix one condition for choosing next week's companies. Without this, the four steps each run once and end there, and your reasons for picking companies never get better.
A&A perspective
Neither customer story describes this review directly. It is a step A&A added to the owner's sales week.
Hypothetical example
For example, in 15 minutes on Friday, give AI the list of companies you contacted this week and have it sort them into replied, no answer and declined, summarizing the reasons. Have it keep decline reasons in the other side's own words.
Hypothetical example
The owner reads that summary and changes just one condition for next week's targeting. If it shows, say, that companies whose contact person changed recently tend to reply, choose next week's companies by that condition.
A&A perspective
For companies you decided not to contact, leave a one-line reason too, such as "not enough information" or "wrong timing for now." You will spend less time hesitating over the same company next week.
A&A perspective
Once the weekly review is running, the reasons for picking companies live outside the owner's head. Unlike meetings bought from a sales agency, those reasons build up inside the company, and they become material when you hand sales to someone on the team.
Summary
A&A perspective
Split the owner's sales into 4 steps (targeting, research, proposal and reply review) and place them in the week.
A&A perspective
Hand research and drafting to AI; the owner writes the one-line reason for picking each company.
A&A perspective
Choose the first thing to hand to AI by the step that stalls in busy weeks.
A&A perspective
Once a week, gather the reasons behind replies and declines, and feed them back into targeting.
Split into four steps (targeting, research, proposal and reply review), the owner's sales shows where it stalls. Hand research and drafting to AI, and have the owner write the one-line reason for picking each company. Choose the first step to hand to AI by what stalls in busy weeks, and once a week feed the reasons behind replies and declines back into targeting.
FAQ
Should we drop our sales agency and switch to AI?
There is no need to switch right away. First hand one step the owner does personally to AI, compare the time over two weeks, then set it next to the agency budget and decide.
Can we start with just ChatGPT?
For the research and drafting steps, you can try a conversational AI such as ChatGPT. The owner opens the source pages to check what the AI found.
Can we send a proposal that AI wrote as it is?
Before sending, the owner reads it and checks that it matches the one-line reason for picking that company. In Lovable's hiring example too, recruiters decide for themselves whether to reach out.
Sources & editorial note
Primary pages read for this article. Publication dates below belong to the sources; access dates record our research.
- Lindy empowers teams to scale with AI Agents powered by Claude
Anthropic · Publication date not stated
Accessed 2026-10-22 - How Lovable uses Clay to scale one of the fastest-growing startups in history
Clay · Publication date not stated
Accessed 2026-10-22