28:49Austin 2026
The Retention Roadshow
Connected AI Marketing and MCP with Rytis Lauris
Rytis Lauris shows how connecting AI to marketing data changes the questions a retention team can answer—from weekly performance reviews to segments and deliverability checks.
Watch it. Put it to work.
ABOUT THIS SESSION
Your data. A more useful AI.
Rytis Lauris shows how connecting AI to marketing data changes the questions a retention team can answer—from weekly performance reviews to segments and deliverability checks.
For retention marketers and marketing leaders exploring practical AI workflows.
- Speaker
- Rytis Lauris
Co-Founder & CEO, Omnisend - Recorded at
- The Retention Roadshow ↗
June 2026
Speaker roles and platform examples reflect the session’s original context. This is an archived conversation.
TAKE IT BACK TO YOUR DESK
Ideas to put to work.
- 01
Start with the scattered-data problem
Identify a recurring decision that currently requires several reports or manual exports.
Read this part · 02:00 ↓ - 02
Connect context before asking for action
Begin with a connected analysis task and define what the assistant is allowed to do.
Read this part · 08:39 ↓ - 03
Turn reporting into a conversation
Use a weekly brief to ask follow-up questions about specific changes, not just summarize totals.
Read this part · 15:37 ↓ - 04
Build segments from the account
Review the actual audience logic and exclusions before turning an AI recommendation into a send.
Read this part · 17:59 ↓ - 05
Keep a human in the review loop
Schedule recurring checks and assign a person to evaluate the recommendations.
Read this part · 25:52 ↓
READ THE SESSION
The edited transcript.
A condensed, edited reading version based on the YouTube captions and session chapters. Repetition and unclear audience audio have been removed; the discussion is paraphrased for clarity, rather than presented as a verbatim transcript. Timestamps refer to the original video.
Original recording on YouTube ↗Start with the scattered-data problem
Marketers often have the information they need, but it lives in separate tools, exports, and reports. Generic AI can explain best practices while still knowing nothing about the actual account. That leaves a person doing the work of gathering context and translating the answer back into the marketing platform.
Rytis frames connected AI as a way to reduce that handoff. A useful answer should respond to the brand’s customers, campaigns, and history, rather than reproduce a strategy that could apply to any store.
Connect context before asking for action
The session introduces MCP as a connection between an AI assistant and tools that expose relevant data and actions. The important distinction is between asking a model what a segment might look like and letting it inspect an account to propose a segment grounded in actual activity.
Start with analysis and read-only access where appropriate. More access creates more possibilities, but it also calls for clear decisions about what the assistant may change. The marketer remains responsible for the question and the action that follows.
Turn reporting into a conversation
A weekly brief can become the beginning of an investigation instead of another dashboard to scan. Ask what changed, where the change came from, and which areas deserve attention. Follow-up questions help move from a number to the campaigns or customer groups behind it.
This works best when the request includes the period, comparison, and business question. The point is not to collect a longer report. It is to shorten the distance between noticing a change and deciding what to investigate next.
Build segments from the account
A generic upsell recommendation might suggest recent purchasers or high spenders. With account context, the assistant can consider actual spend, engagement, purchase timing, and the size of the resulting audience. Rytis contrasts that with manually rebuilding a broadly worded recommendation.
Review the proposed logic before using it. A segment can be technically valid and still be inappropriate for an offer. The connection saves preparation time; the marketer still needs to check exclusions, customer experience, and the purpose of the campaign.
Keep a human in the review loop
Deliverability is often investigated only after emails start landing in spam. The session proposes using connected analysis to make regular checks easier and surface changes earlier. AI can help inspect signals, but a diagnosis or suggested fix still needs review.
Rytis closes with the human tasks that remain: ask useful questions, assess recommendations, prioritize work, and apply judgment. Connecting data makes the assistant more relevant; it does not make every answer correct or every proposed action worth taking.
KEEP THE IDEAS COMING
Stay in the conversation.
3 SESSIONSPLAY IT BACK. PUT IT TO WORK.
28:49
36:10The Retention Roadshow
Retention Redefined 2026
26:17San Diego 2024
Decoding Emotions: How AI is Revolutionizing Advertising Resonance
YOUR NEXT STEP: IN THE ROOM
Bring your questions.
Meet your people.
Join Commerce Roundtable in 2027 for focused learning and conversations with people working through the same challenges.