- Personalized push, 18:300.58
- Win-back email0.41
- 10% off next monthBlocked · margin
- Do nothing0.31
For marketing teams
Agentic lifecycle marketing on autopilot.
Give each customer their own lifecycle agent. It picks the next best action across email, push and in-app, writes it in your brand voice, sends it at the moment they’re most likely to act, and proves the lift against a holdout. No journeys to build.
Your 12-day run ends tonight
One 5-minute session keeps it going, Sam. Your usual route is 2 minutes away.
Sent when Sam usually trainsAgents + messaging + in-app
Lifecycle marketing, without the journey builder.
One agent for every customer that knows their habits and reaches them in their inbox, on their phone and in your app. No segments, sequences or templates to maintain.
Lifecycle agents
An agent for every customer that picks the next best action toward your objective, from activation to win-back, and proves its lift against a holdout.
Explore lifecycle agentsPersonalized messaging
Email and push written for one customer in your brand voice, sent at the moment they’re most likely to act rather than on a batch schedule.
Explore messagingIn-app experiences
Cards, progress and milestones chosen for each customer and shown inside your product, so lifecycle doesn’t stop at the inbox.
See in-app features01 · Your data, activated
The agentic lifecycle marketing platform
Your data goes in, agents decide and write for each customer, guardrails check every send, and it goes out on the right channel. No journeys, segments or templates in between.
02 · You stay in control
You set the objective and the rules. Agents do the work.
No sequences to maintain, no A/B tests to run, no engineering ticket for every campaign. Your team decides what good looks like and what is off limits, and agents iterate on the rest.
Objectives not schedules
Pick the outcome, such as D30 retention or renewals. Agents score every action against it, including doing nothing.
Timed for each user
Every message is sent at the time each user is most likely to take action, based on their time zone and past behavior.
Guardrails you control
Frequency caps, each customer’s own volume level, consent, quiet hours, margin limits and brand rules. Every send is checked against them first.
Always on brand
Upload images tagged by tone and use case, set your voice and style, and list what should never be said. Agents pick what to use.
Give feedback and agents learn
New situations, low confidence and new offers go to your team in Slack as one question. Only the affected sends wait.
Uplift, not vanity metrics
Every decision and outcome is measured against a holdout according to your KPIs. Agents learn from the results and improve over time.
Morning review done. 48,210 customers reviewed, 9,960 sends planned for their own hours, 31,804 left alone.
Ultra goes live today. The launch note is on the offer.
Oh, nice! Thanks Jordan
@Nadia I’d like to start using the new Ultra offer with customers. What exactly is the offer and when does it end, which link should I use, and should everyone see it, or only people likely to find it valuable?
I’ll leave it out of every message until you reply. Everything else goes out as planned.
@Nadia I’d like to start using the new Ultra offer with customers. What exactly is it, which link should I use, and who should see it?
First month of Ultra at the Pro price, until 11 November. Use yourapp.example/ultra?ref=launch. Only free users who keep hitting the plan limits, not everyone.
Thanks, Nadia. Saved to the Ultra offer, and I’ll start showing this to relevant users.
Results
Leading consumer platforms are using Trophy to deliver personalized experiences.
Study streaks and milestones for 350,000 students on a revision app
+9%14-day retention since adding TrophyRevisionDojoRead the story →A weekly writing habit for 1M authors, live in about an hour
+22%14-day retention since adding TrophyCampfireRead the story →Leaderboards and XP for a loyalty platform’s retail merchants
+22%Loyalty points redemption rate vs control groupLealRead the story →
Powering engagement for 1M+ users
03 · Ready for rollout
Live in weeks. Safe by default. Fits your stack.
With you from the first week
- A shared Slack channel with the people who actually build Trophy
- Co-build integrations with data sources, messaging providers and your product
- Ongoing guidance on configuring objectives and guardrails
- Weekly KPI reporting cadence for continuous improvement
Your customers’ data, kept safe
- Consent, preferences and brand rules enforced on every send
- Single sign-on, MFA and role-based access for your team
- SOC 2 and ISO reports, data residency options and a DPA. See our trust center for more.
Fits the stack you already run
- Pre-built integrations for common data sources and messaging providers
- Custom integrations built for in-house or niche systems
- No product changes required to use Trophy
Questions
What marketing teams ask us
Straight answers about agents, results and getting started.
Agents
How is Trophy different from journeys and campaigns?
Journeys send every customer down the same branches of triggers, delays and templates, usually one for onboarding, one for engagement and one for win-back. Trophy replaces them with an agent for each customer. You set the objective, brand kit and guardrails; agents decide what to send, when and on which channel for each person.
How do agents decide what to send, and when?
Every customer gets their own agent. It looks at what they did, scores every action it could take against your objective, including doing nothing, and picks the one most likely to change what they do next. Then it waits for the moment that customer is most likely to act.
Who writes the messages, and will they sound like us?
Agents write each message from your brand kit: images tagged by tone and use case, your voice and style guidelines, and the phrases and topics they should never use. Emails send from your own domain with your branding.
Can agents use our offers and discounts?
Yes, inside the rules you set. Margin limits block discounts that aren’t worth giving, and new offers go to your team in Slack before agents use them. Often the best message has no offer at all.
Will my customers get more messages?
Usually fewer. You cap messages per customer per week and per month, agents never send outside the volume level a customer picks, and they only act when an action beats doing nothing. On most days, most customers get nothing at all.
What happens when an agent isn’t sure?
It asks. New situations, low confidence and high-risk decisions such as a new offer go to your team in Slack as one question. Only the sends it touches are held; everything else goes out on time, and it never asks the same thing twice.
Objectives and results
What objectives can agents work toward?
An overarching objective such as activation, retention or reactivation, with the specific results that matter weighted by what they’re worth to you, so an upgrade counts for more than a completed lesson. Each agent works toward the objective that fits where its customer is now.
How do we know it’s working?
Every decision is measured against a holdout agents never touch. You compare the two over time to see the incremental lift on your KPIs, such as retention and return rate, not just opens and clicks.
Do we still need to A/B test campaigns?
No. An A/B test compares two versions and ships one winner to everyone. Agents are closer to running thousands of tests at once, learning from how each customer responds and picking what works for each person, all measured against one holdout.
Can we try it on a small group first?
Yes. We’ll carve out a small share of your customers as a pilot to measure the lift before you roll out to everyone.
Getting started
What data does Trophy need?
The events you already track, such as sessions, page views and purchases, with each customer’s attributes, subscription state and time zone. Agents learn each customer’s habits from them.
Do we need engineering to get started?
A little, once. Your engineers send the events you already track, and we build the integrations with you. After that your team runs objectives, guardrails and the brand kit itself, with no engineering ticket for every campaign.
Which channels can agents use?
Email, sent from your own domain; push, delivered natively through Firebase Cloud Messaging, APNs or Expo; and in-app experiences inside your product. Agents pick the channel that works best for each customer.
Do we have to rebuild our campaigns?
No. There are no journeys to rebuild. Bring your brand kit, offers and objectives; agents decide who gets what and when, and your team approves new offers and answers their questions.
