A founder got on a call with me last week to ask about hiring my AI development agency to automate their sponsor outreach. I told them to go send 20 to 50 emails a day by hand instead, before hiring us.

They run a newsletter and events business. They were sending about five cold pitches a week, and roughly 70% of their sponsors came to them inbound. Everything that wasn't inbound came out of those five emails.

This issue is what I said on that call, the math behind it, and why I think the order matters for almost any team, mine included. If this breakdown is useful, reply and tell me. If it isn't, reply and tell me that too, I read all of them.

If you're new here: I was beehiiv's first sales hire and built the GTM team that helped grow the company to ~$35M ARR. I’m now the CEO of an AI Agency called Creme Digital. I've built a lot of cool things, and now I help other people build cool things too.

They wanted an agent that could pitch sponsors and explain what their business does, because they were wearing every hat and didn't have the capacity to pitch more. Totally reasonable, and something I’ve built successfully for a number of businesses. They'd already paid someone a monthly retainer plus a cut of whatever got closed. Classic early stage sales comp model, nothing wrong with it, but hard to motivate someone with that model unless they’re a killer.

I have built and sold exactly what they were asking for. The easy move was to say yes and send a proposal.

But what we'd have been building was automation for a process that barely exists. Five emails a week is not a sales motion yet. And our retainer would have eaten most of what their business brings in each month. I told them that on the call, which is not a normal thing to say to a lead.

I walked them through the math.

Writing one pitch by hand takes two to three minutes once you've done a handful. Call it three. That's 20 an hour. One hour a day, 20 new accounts.

Five days a week, that's 100 new accounts. They were only doing five. If the same founder did one extra hour on the calendar with 20 times the volume, their business would be in a completely different state.

Then the follow-ups start to kick in and compound.

Every account you contacted last week gets a second touch while you're sending 20 more new ones. Ten weeks in, you've touched about 1,000 accounts and there's more follow-up than one person can do. That's the day you hire someone or build a tool, and by then you know exactly what to tell it.

if you automate this before you know what the pitch should be, or what the actual product is that you're selling, or what people you should be reaching out to, then you end up A/B testing everything for a lot longer than it would take if you already know what works and what doesn't.

I'd pick the number you can hit every single day over the number that sounds impressive. The first couple of weeks will probably feel bad. Replies are slow, the first template is wrong, and you're mostly learning which sentence gets answered. I told them I thought they could double or triple their revenue just by running that route for a month or two.

I like to know what good looks like before I optimize it. I closed $4M of deals at beehiiv, and I can tell you what a good first email sounds like because I wrote a lot of bad ones first.

Now agents do a lot of that work for me. Nobody writes those instructions from a blank page. Knowing what good looks like before automating a process is the difference maker in having agents who work and agents who produce a bunch of useless slop.

There's an old line from a pediatrician named John Gall, who wrote a book on how systems fail in 1975:

An outbound agent is a complex system. If you build it before a simple version works, you're debugging the offer and the agent at the same time and you can't tell which one is broken.

Someone on my team said it even more plainly on the call. Figure out what works, then build the bot.

I gave them one pitch tip, and it applies to anyone selling sponsorships, newsletter ads, or really anything with an audience attached.

Lead with the fact that your readers can't be reached anywhere else. A brand can buy ads on Meta or in other newsletters. It can't buy the trust you've built with your people. That's what you're selling, so say it in the first line, and don't discount it to get the meeting.

Near the end they told me they don't have a lot of sales experience, and weren't sure they could ask big companies for big money.

I told them I get that feeling too. Some days I look at my calendar, see what has to happen to grow my company, and I just don't want to sell. Then I remember that every founder has to sell, because otherwise there's no company. Ironically, the more senior I get in a company and further away I get from being a salesperson, the more I become a salesperson.

They said they needed to put a lot more time into it, because it's the critical activity.

I strongly agreed. You can't grow a company without sales. It's as simple as that.

If you already have a process that works and you're out of hours, that's when an agent pays for itself. You know what a good email looks like, you know what a reply is worth, you know what the LTV is of your customers and how much you need to make for an engagement with someone like me to be worth it, and the thing between you and the next level is just time and expertise. We look for exactly that in an audit: what's already working, and what's eating your team's day to day.

If that's you, you can book an AI audit here. If you're not there yet, you don't need us yet, and I'll say so on the call, same as I did last week.

And if you send fewer than 20 a day right now, send 50 tomorrow and tell me what happens.

In either case, let’s stay in touch if I can help support what you’re working on in any way.

Daniel Berk
CEO, bald