Timing Optimization
“For each customer, will they respond to outreach sent in the next 4 hours?”
Book a demo and get a free trial of the full platform: data science agent, fine-tune capabilities, and forward-deployed engineer support.
By submitting, you accept the Terms and Privacy Policy.
Loved by data scientists, ML engineers & CXOs at
A real-world example
For each customer, will they respond to outreach sent in the next 4 hours?
Most marketing platforms send outreach at a fixed time or use basic timezone heuristics. But optimal engagement windows vary by individual — influenced by work schedules, app usage patterns, and recent interactions. Sending at the wrong time means the message is buried by the time the customer checks their inbox.
How KumoRFM solves this
Relational intelligence for optimal actions
Kumo learns each customer's response patterns from the full temporal graph of outreach events, responses, and contextual signals. The binary classifier predicts whether a specific customer will respond if contacted now — enabling real-time send-time optimization that adapts to changing behavior, not static rules.
From data to predictions
See the full pipeline in action
Connect your tables, write a PQL query, and get predictions with built-in explainability — all in minutes, not months.
Your data
The relational tables Kumo learns from
CUSTOMERS
| customer_id | timezone | segment |
|---|---|---|
| C-3001 | US/Eastern | high-value |
| C-3002 | US/Pacific | growth |
| C-3003 | Europe/London | enterprise |
| C-3004 | Asia/Tokyo | new |
| C-3005 | US/Central | high-value |
OUTREACH
| outreach_id | customer_id | channel | send_hour | timestamp |
|---|---|---|---|---|
| OUT-601 | C-3001 | 9 | 2026-02-20 09:00 | |
| OUT-602 | C-3001 | push | 14 | 2026-02-20 14:00 |
| OUT-603 | C-3002 | 11 | 2026-02-20 11:00 | |
| OUT-604 | C-3003 | 8 | 2026-02-21 08:00 | |
| OUT-605 | C-3004 | push | 20 | 2026-02-21 20:00 |
RESPONSES
| response_id | customer_id | outreach_id | action | timestamp |
|---|---|---|---|---|
| RSP-701 | C-3001 | OUT-601 | clicked | 2026-02-20 09:12 |
| RSP-702 | C-3001 | OUT-602 | ignored | — |
| RSP-703 | C-3002 | OUT-603 | opened | 2026-02-20 11:45 |
| RSP-704 | C-3003 | OUT-604 | clicked | 2026-02-21 08:05 |
| RSP-705 | C-3004 | OUT-605 | opened | 2026-02-21 20:30 |
Write your PQL query
Describe what to predict in 2–3 lines — Kumo handles the rest
PREDICT COUNT(RESPONSES.*, 0, 4, hours) > 0 FOR EACH CUSTOMERS.CUSTOMER_ID
Prediction output
Every entity gets a score, updated continuously
| CUSTOMER_ID | TIMESTAMP | TARGET_PRED | True_PROB |
|---|---|---|---|
| C-3001 | 2026-03-12 09:00 | True | 0.89 |
| C-3001 | 2026-03-12 14:00 | False | 0.22 |
| C-3002 | 2026-03-12 11:00 | True | 0.76 |
| C-3003 | 2026-03-12 08:00 | True | 0.93 |
| C-3004 | 2026-03-12 20:00 | True | 0.71 |
| C-3005 | 2026-03-12 07:00 | False | 0.18 |
Understand why
Every prediction includes feature attributions — no black boxes
Customer C-3001 at 09:00 ET
Predicted: True (0.89 probability)
Top contributing features
Responded to 4 of 5 morning emails (RESPONSES)
80% AM response
35% attribution
Average response latency < 15 min at 9 AM (RESPONSES)
12 min avg
27% attribution
Timezone = US/Eastern, weekday (CUSTOMERS)
US/Eastern
22% attribution
No outreach fatigue — last contact 3 days ago (OUTREACH)
3 days gap
16% attribution
Feature attributions are computed automatically for every prediction. No separate tooling required. Learn more about Kumo explainability
PQL Documentation
Learn the Predictive Query Language — SQL-like syntax for defining any prediction task in 2–3 lines.
Python SDK
Integrate Kumo predictions into your pipelines. Train, evaluate, and deploy models programmatically.
Explainability Docs
Understand feature attributions, model evaluation metrics, and how to build trust with stakeholders.
Bottom line: Send each message at the moment the customer is most likely to respond. Lift open rates 20-35% and click-through rates 15-25%, translating to $2-4M in incremental engagement-driven revenue.
Related use cases
Explore more next-best-action use cases
Topics covered
One Platform. One Model. Predict Instantly.
KumoRFM
Relational Foundation Model
Turn structured relational data into predictions in seconds. KumoRFM delivers zero-shot predictions that rival months of traditional data science. No training, feature engineering, or infrastructure required. Just connect your data and start predicting.
For critical use cases, fine-tune KumoRFM on your data using the Kumo platform and Data Science Agent for 30%+ higher accuracy than traditional models.
Book a demo and get a free trial of the full platform: data science agent, fine-tune capabilities, and forward-deployed engineer support.
