THEAGENCY.STRATEGY BUILDS WHAT’S NEXTStart a conversation →

Scale with AI / Episode 14 / October 17, 2026

Scale with AI — October 17, 2026

AI marketing is moving from writing content to making decisions — the right message, to the right person, at the right moment. The first measured results are in, Pennsylvania's tough new outreach law starts tomorrow, and the biggest asset you're sitting on is your dormant customer list.

The Agency LLC · 7 min · Updated October 17, 2026

Scale with AI podcast cover art

Listen. Then put it to work.

Read the full transcript ↓

Quick Take

  • AI marketing is shifting from a copy machine to a decision loop — one message per person, chosen by what they just did and tested against a holdback group.
  • The first measured decisioning results are big: 137% more opens and 207% more clicks on a reactivation campaign, from better decisions, not more content.
  • Pennsylvania Act 47 takes effect Sunday, October 18 — 9 AM–7 PM Monday–Saturday, no Sundays, written consent for AI-voice marketing, and QUIT or CANCEL count as opt-outs.
  • Three of four AI-referred sales now involve specialist products outside the top 100 categories — small businesses get found when their information is clear enough for an assistant to recommend.
  • Your dormant customer list is usually worth more than your next ad campaign — reactivate it with consent on file, correct hours, and an honest holdback test.

The Shift

For two years, most businesses have used AI in marketing as a copy machine — it writes a subject line, drafts a post, makes a picture. That saves time, but this week the first published customer results arrived for a different kind of system: one that makes decisions in real time. A traditional campaign is a schedule — built Monday, sent Thursday, the same thing to everyone. A decisioning system is a loop. A customer opens a message, visits a pricing page, or goes quiet for two weeks; each is a signal. The system scores which next step is most likely to move that specific person, sends only that step, and learns from what happens. No two people walk the same path. And the timing matters because most of a business's revenue is not hiding in strangers — it is hiding in people who already know you: past customers who went quiet and quotes that never got an answer. That list is an asset sitting idle, and a decision loop is built exactly for it.

The Fine Print

Pennsylvania's Act 47 takes effect Sunday, October 18, 2026, and it is one of the toughest telemarketing laws in the country. Marketing calls and texts are allowed only between 9 AM and 7 PM, Monday through Saturday — nothing on Sundays or legal holidays. A text counts exactly like a call, and so does a voicemail, including a ringless voicemail that lands without the phone ever ringing. Automated and AI-voice marketing requires the person's prior express written consent — written consent you can produce, not a handshake or a past purchase by itself. Opting out is broader than STOP: QUIT, END, REVOKE, OPT OUT, CANCEL, and UNSUBSCRIBE all count and must be honored. Penalties run $1,000 per violation, $3,000 when the person reached is 60 or older. The bigger point: states are now writing their own rulebooks with different hours and consent language, so any follow-up system that crosses state lines has to know which rules apply before it reaches out.

Who’s Doing It

The decisioning results are unusually well measured, because each ran against a control group. A reactivation campaign aimed at customers who had gone quiet produced 137% more opens, 207% more clicks, and nearly 14% more purchases per thousand people targeted — with the audience and offers unchanged; only the AI's choices differed. An onboarding journey produced more than 1,200 extra first-time purchases and lifted newcomer conversion by 42%, with converters spending about three times the average. A churn-prevention campaign reached a 10% purchase rate against just over 1% in the control — roughly seven times the result from deciding better, not spending more. A second group is winning differently: small specialists getting found by AI assistants. Early commerce data shows three of every four AI-referred sales involve specialized products outside the biggest hundred categories, because shoppers describe exactly what they need and the assistant matches them to the smaller business that fits — but only when that business's information is clear enough to understand and recommend. The backdrop: 58% of small business employers now use AI agents, and in a survey of nearly four thousand owners, 86% said they need more guidance to get the full benefit. Owners are not short on tools; they are short on someone to set the system up correctly, keep it legal, and measure it honestly.

Your Move

Three steps. First, find your dormant asset: pull the list of past customers and unanswered quotes from the last twelve months — reactivated well, it is usually worth more than your next ad campaign at a fraction of the cost. Second, before any automated outreach touches that list, do the compliance check: written consent on file, calling and texting hours set by state, and every opt-out word honored, including the Pennsylvania rules that start tomorrow. Third, run it as a test — reach part of the list, hold part back, and measure the difference in real revenue; keep what wins and kill what does not. That is exactly what we build at The Agency: not a pile of tools to learn, but a working system — a decision-driven follow-up engine for the customers you have already earned, consent and hours handled correctly in every state you serve, and pages written clearly enough that AI assistants find and recommend you. Start with the free AI audit at theagencyllc.net and see, in plain numbers, where AI can actually move your revenue.

Key Takeaways

  • First published AI decisioning results (October 15, 2026): a reactivation campaign produced 137% more opens, 207% more clicks, and 13.8% more deposits per 1,000 people targeted versus control — the only variable changed was the AI choosing the content.
  • An onboarding journey using decisioning produced 1,200+ incremental first purchases, a 42% higher conversion rate, and converters worth about 3x the average; a churn campaign hit a 10% purchase rate against 1.4% in the control group.
  • Pennsylvania Act 47 takes effect October 18, 2026: 9 AM–7 PM Monday–Saturday, no Sundays or holidays, prior express written consent for AI-voice marketing, and opt-out words beyond STOP — QUIT, END, REVOKE, OPT OUT, CANCEL, UNSUBSCRIBE.
  • Early commerce data: 75% of AI-referred sales involve specialized products outside the top 100 categories — small specialists get found when an AI assistant can clearly understand what they do.

Full transcript

Alex: Welcome to Scale with AI for Saturday, October seventeenth, twenty twenty-six. I am Alex.

Jordan: And I am Jordan. Today we are talking about marketing, but not marketing the way most owners mean it. Not more posts, and not more ads. We are talking about A I that decides, customer by customer, what to say, when to say it, and when to stay quiet. That decision layer is where the measurable money showed up this week.

Alex: This is The Shift. For two years, most businesses have used A I in marketing as a copy machine. It writes a subject line, it drafts a post, it makes a picture. That saves time, and time matters. But this week, the first published customer results came in for a different kind of system, one that makes decisions in real time. Instead of sending the same campaign to everyone, it watches what each person just did, and chooses the next message, the next offer, and the next moment, for that one person.

Jordan: Let me explain how that actually works, because the mechanism is the whole story. A traditional campaign is a schedule. You build it on Monday, it goes out on Thursday, and everyone gets the same thing. A decisioning system is a loop. Someone opens a message, visits a pricing page, or goes quiet for two weeks. Each of those is a signal. The system scores which next step is most likely to move that specific person, sends only that step, and learns from what happens. No two people walk the same path.

Alex: And here is why owners should care. Most of your revenue is not hiding in strangers. It is hiding in people who already know you. Past customers who went quiet. Quotes you sent that never got an answer. That list is an asset sitting in your system, doing nothing. A decision loop is built exactly for that list.

Jordan: This is The Law, and today it has a date on it. Tomorrow, Sunday, October eighteenth, a new Pennsylvania telemarketing law takes effect, and it is one of the toughest in the country. Marketing calls and texts are allowed only between nine in the morning and seven in the evening, Monday through Saturday. Sundays are completely off limits, and so are legal holidays. And a text counts exactly the same as a call now, including a voicemail that lands without the phone ever ringing.

Alex: Two more parts owners miss. First, an automated or A I voice reaching out for marketing needs the person's prior express written consent. Not a handshake, not a past purchase by itself. Written consent you can produce. Second, opting out is broader than the word stop. If a customer replies quit, end, revoke, opt out, cancel, or unsubscribe, that counts, and you must honor it. If your system only listens for stop, it is not ready.

Jordan: The cost of getting this wrong is not abstract. Violations carry penalties per contact, with higher penalties when the person reached is sixty or older, on top of the federal rules that already apply. And the point is bigger than Pennsylvania. States are writing their own rules now, with different hours and different consent language. If your follow-up runs across state lines, your system has to know which rulebook applies before it ever reaches out.

Alex: This is Who Is Doing It, and we start with those decisioning results, because they are unusually well measured. One business ran a reactivation campaign aimed at customers who had gone quiet. Compared with a control group, it produced one hundred thirty-seven percent more opens, two hundred seven percent more clicks, and nearly fourteen percent more purchases for every one thousand people targeted. The only thing that changed was the A I choosing the content. Same audience, same offers, different decisions.

Jordan: A second onboarding journey produced more than twelve hundred extra first-time purchases, and lifted the rate at which newcomers became buyers by forty-two percent. Those who converted also spent about three times the average. And a churn-prevention campaign reached a ten percent purchase rate, against just over one percent in the control group. Roughly seven times the result, from deciding better, not spending more.

Alex: None of those wins came from making more content. They came from matching the right message to the right person at the right moment, and testing honestly against a control group. Your version is simpler. One dormant-customer list, one well-built sequence, one clear measure. Did the people we reactivated buy more than the people we left alone? If you cannot answer that, you are not doing decisioning. You are just sending.

Jordan: There is a second group doing it, small specialists getting found. Early commerce data shows three out of every four sales referred by A I assistants involve specialized products outside the biggest one hundred categories. Shoppers describe exactly what they need, and the assistant matches them to a smaller business that fits. But only if your information is clear enough for an assistant to understand and recommend. If an A I cannot tell what you do, who you serve, and why you are different, you are invisible in that moment.

Alex: And the backdrop makes this urgent. A new October survey finds fifty-eight percent of small business employers now use A I agents, with another fourteen percent planning to. In another survey of nearly four thousand owners, eighty-six percent said they need more guidance to get the full benefit. Owners are not short on tools. They are short on someone who will set the system up correctly, keep it legal, and measure it honestly.

Jordan: This is Your Move, and it has three steps. Step one, find your dormant asset. Pull the list of past customers and unanswered quotes from the last twelve months. That list, reactivated well, is usually worth more than your next ad campaign, and it costs a fraction as much to reach. Step two, before any automated outreach touches that list, do the compliance check. Written consent on file, calling and texting hours set by state, and every opt-out word honored, including the Pennsylvania rules that start tomorrow. Step three, run it as a test. Reach out to part of the list, hold part back, and measure the difference in real revenue. Keep what wins. Kill what does not.

Alex: That is exactly what we build at The Agency. Not a pile of tools you have to learn, but a working system. A decision-driven follow-up engine for the customers you have already earned, consent and hours handled correctly in every state you serve, and pages written clearly enough that A I assistants find you and recommend you when a new customer describes what they need. If you want to know where your business stands, start with the free A I audit on our website. It takes a few minutes, and it shows you, in plain numbers, where A I can actually move your revenue.

Jordan: Your move is simple. Do not buy another tool this weekend. Open your own customer list, count the people who went quiet, and ask what one respectful, well-timed, fully compliant message could be worth. Then build the system that sends it right, every time.

Alex: I am Alex.

Jordan: And I am Jordan. This has been Scale with AI for Saturday, October seventeenth. We will see you tomorrow.

Sources

  1. Optimove — first AI decisioning customer results (GlobeNewswire) ↗
  2. Pennsylvania Act 47 of 2026 — telemarketing and AI outreach ↗
  3. SBE Council — Small Business Check Up, October 2026 ↗
  4. Shopify — AI-referred commerce data ↗

Your next conversation

Get Your Free AI Audit

Start a conversation ↗