← Back to JunoDesk
Propensity Modeling

Target Who's Likely to Convert

Rank prospects and existing customers by likelihood to buy, churn, or respond — so targeting stops being a guess.

Book a Demo → See LTV Modeling →
Value vs. Propensity, Every Customer
TARGET SEGMENT High Low PREDICTED VALUE →
Target Segment Rest of the List
Why It Matters

Spend On Who's Worth It

Every customer gets a score — so budget goes toward who's actually worth reaching, not whoever's easiest to find.

Chase the Real Opportunity

The top decile converts with or without a nudge. The real opportunity sits in the middle — customers who are genuinely on the fence, where outreach actually changes the outcome.

Catch Churn Before It Happens

Risk shows up in the score weeks before the cancellation does — early enough for your client's team to reach out while there's still an account to save.

One Score, Every Channel

The same ranked list feeds email, paid retargeting, and outbound — every channel working off the same priority, not its own idea of who matters.

The Process

How Does Propensity Modeling Work?

You send customer data. We send back a ranked, actionable list.

01 — Score Customers

A probability score, not a guess

We pull transaction history, behavioral signals, and service history — data already sitting in your warehouse, not a new pipeline to build.

The model learns what separates customers who took the action from those who didn't, then scores every single customer 0–1 for it.

Propensity Scores
400K Customers
0–1 Score Range
Customer Predicted Action Score Decile
#48213 Purchase 0.94 10
#19042 Purchase 0.81 9
#55107 Purchase 0.42 4
#33890 Purchase 0.06 1
02 — Segment Into Deciles

Ranked into deciles, not one cutoff

Scores are sorted into ten equal groups — decile 10 is your most likely converters, decile 1 your least.

The top decile alone captures roughly 35% of predicted buyers, but they'd likely convert anyway. Deciles 3–4 hold the real opportunity — customers who are genuinely undecided, where a nudge actually changes the outcome.

Decile Breakdown
Real opportunity Converts anyway D1 D2 D3 D4 D5 D6 D7 D8 D9 D10
35% Top Decile
22% Deciles 3–4
03 — Act By Segment

A different play for every segment

Your highest scorers convert without an incentive — spending on them is wasted budget. Your lowest scorers won't convert regardless of what you offer.

  • Suppress spend on won't-converts and already-converting segments.
  • Concentrate incentives on the genuinely undecided middle.
  • Re-score every cycle as new behavior comes in.

The real ROI sits in the middle — customers who are genuinely undecided, where the right offer actually changes the outcome.

Segment Deciles Score Range Recommended Action
High Propensity 9 – 10 0.75 – 1.00 Already converting — no incentive
Mid Propensity Opportunity 3 – 4 0.30 – 0.55 Targeted offer
Low Propensity 1 – 2 0.00 – 0.15 Suppress spend
Recommended Budget Allocation
High Propensity
5%
Mid Propensity
85%
Low Propensity
10%
Contact

Ready to Go Beyond Reporting?
Let's Talk

Let's talk about your next client, project, or pitch.

Book a Demo →