Blog
My AI agent secured an $80 Verizon offer
A real story from today, with the real screenshots. One prompt to my AI agent turned a vanished $20 loyalty offer into a confirmed $80 order and a technician visit pulled forward to the next morning.

What happened
My Verizon bill was $94.99 a month for gigabit internet. This morning a $20 a month loyalty offer sat in my account with a Redeem button, then vanished before I could redeem it. No explanation. So I gave my AI agent one prompt: keep pushing until they give me something. It ran the Verizon chat while I watched, and every decision came back to me for a yes or no.
What it closed
The order email confirms it: Fios 1 Gig at $90 minus the $10 autopay and paper free discount, Whole Home WiFi fee waived, estimated $80 a month even, taxes $0. No contract, a new router and extender included, and a five year price guarantee listed under special offers. The order also carries the optional Google AI Pro perk at $10 a month, offset for 6 months, removable anytime. The $10 shipping fee came off when I asked for a technician visit instead, confirmed at $0, and the rep pulled the visit forward from November 10 to the next morning: the ticket email confirms November 5, 10 AM to noon, tech bringing the equipment.
My agent also asked what staying put would cost over the year, then asked Verizon for a one time $60 credit to cover the gap. Verizon did not confirm the credit. It pinned down the alternative in writing instead: a $60 Verizon gift card, digital, issued once the new service activates. That lives in the chat notes, not the order email, so fulfillment after activation is the part to watch. One asterisk stays: the ticket email carries a generic repair charges may apply line against the chat's $0 promise, and the equipment return deadline conflicts across the emails (December 4, 5 and 13).
The day at a glance

The proof

The chat where the $80 offer landed, cropped to the rep exchange so it reads on a phone. Personal details blacked out in the original, messages untouched.

The ticket email confirming the next morning visit.
The numbers


My take
Maybe Verizon had AI on its side. I definitely had AI on mine. I did not need to become a better negotiator. I needed to stop doing all the negotiating myself.
Steal this workflow
Here is my account. I pay [your amount] a month for [service], and my current discounts and terms are: [paste what you know]. Open my provider's support chat and negotiate a better price. Rules: use only my real account facts, never invent competitor prices or offers, and never accept or change anything without my yes. Before any commitment, show me the full offer in one table: monthly total with taxes and fees, equipment costs, contract or price lock terms, what happens to my existing discounts, and any trial that starts charging later.
Before you accept: get the real monthly total with taxes and equipment, ask what they will not give, and keep it all in writing. After you accept: read the confirmation line by line and check the first real bill. One rule above all: your agent negotiates, you decide.
Come say hi
I am putting real cases like this one, with the actual prompts and chats, into my $29 AI cheatsheet, which is almost ready. Join the free Discord to hear first when checkout opens, or DM me on Instagram. More real cases live on my site: jadenhong.io case studies.
Written November 4, 2026, updated the same evening after the order confirmation email landed. Figures come from my own account, the chat and the order email.
Read next
Blog
When Does AI Move First?
The newest AI agents start work on their own and message you first. The real question is when they decide to move: on a schedule, on an event, or on their own timing.
Blog
ChatGPT Is Now a Fifth of Google Search, OpenAI Says
OpenAI says ChatGPT now runs one fifth as many searches as Google. What that number actually counts, why it still is not website traffic, and the three checks a small business should run this week.
Blog
7 AI Skills Worth a Closer Look
Seven AI tools and skills worth reviewing for real work, with practical use cases and honest limits.


