B2B Enterprise ABM & Demand Generation EnginePlaybook3 min readUpdated September 2026

Finding the Usage Signals That Actually Predict an Upsell

Most expansion outreach happens on a calendar trigger, like ninety days after signup, rather than a behavioral one, which means a rep might pitch an upsell to an account that's barely using the product while a genuinely ready account waits for its scheduled check-in. The accounts actually ready to expand usually show a specific usage pattern well before any calendar-based trigger would have caught it.

Finding that pattern in your own usage data, rather than assuming a generic timeline applies, is what turns cross-sell and upsell outreach from a guess into something closer to a real signal.

How do you find what your past expansions had in common?

Pull your recent expansion deals and check their usage data in the weeks before they expanded: were they hitting a usage limit, using a specific feature heavily, or adding new users at a faster rate than typical. Look for a pattern that shows up across most of these accounts, not just one, since a single account's story can be a coincidence rather than a reliable signal.

A common pattern in seat-based products is a team consistently operating near its seat limit for a sustained stretch before expanding; a common pattern in usage-based products is hitting a specific feature or volume threshold repeatedly rather than just once. What your own pattern looks like depends on your specific product and pricing model.

Include a few accounts that grew but never expanded in this same review, since comparing them against the ones that did expand often reveals which specific usage difference actually separates the two groups, rather than just describing what expanded accounts looked like in isolation.

Should expansion triggers use usage patterns or calendar dates?

Once you've identified a real usage pattern that precedes expansion, build an alert that flags an account when it starts showing that pattern, rather than continuing to rely on a fixed check-in schedule that catches some ready accounts too late and others not at all. This usually means pulling from your product's usage data directly, not just CRM fields that get updated manually.

The alert doesn't need to be complex to start. Even a simple weekly export flagging accounts crossing a specific usage threshold, reviewed manually by an account team, beats a purely calendar-based check-in until you have the volume to justify a more automated system.

Start the alert on a single, well-validated signal rather than trying to combine several usage factors into one composite score right away. A single clear signal is easier to trust, easier to explain to an account team, and easier to check against real outcomes as you refine it.

Match the outreach to the specific signal that triggered it

An account flagged for expansion because it's near a seat limit needs a different conversation than one flagged because it's heavily using a specific advanced feature not included in its current plan. Generic "time to upgrade" outreach that doesn't reference the specific signal feels like a sales push; outreach that references the actual pattern, like noting the team is close to its seat limit, feels like a relevant, timely observation.

Give account teams the specific signal alongside the account name when a trigger fires, not just a flag to reach out, so the actual conversation can reference something concrete and true about that account's usage rather than a generic upgrade pitch.

A short internal note template, filled in automatically with the specific metric and threshold that fired, saves the account team from having to dig through a dashboard themselves before every outreach call.

Watch for a signal that predicts expansion for one segment but not another

A usage pattern that reliably predicts expansion among your larger accounts might not hold for smaller ones, or a pattern common in one industry vertical might not translate to another. Check whether your identified trigger holds consistently across your different customer segments before rolling it out as a single universal signal, since a trigger that's reliable for one segment but noisy for another will generate a lot of low-quality alerts if applied everywhere equally.

It's reasonable to end up with different triggers for different segments rather than forcing one signal to work universally across a customer base that doesn't actually behave uniformly.

Common mistake: triggering outreach on total usage growth alone

Total usage going up looks like a positive signal, but if that growth is coming from a customer simply using the product more within their current plan's limits, without approaching any actual constraint, there's no real expansion trigger yet, and reaching out prematurely with an upgrade pitch can read as pushy rather than helpful.

The signal that actually predicts a willingness to pay more is usage bumping against a real constraint or limitation in the current plan, not just usage trending upward in a general sense. Confirm the account is genuinely near a boundary, not just more active than before, before treating it as expansion-ready.

Build an expansion trigger in this order:

  1. Pull recent expansions and check their usage in the weeks before, such as usage limits, heavy feature use or faster user growth.
  2. Turn the pattern that shows up across most of them into an alert that flags accounts as they begin to show it.
  3. Tailor outreach to the signal that fired, such as a seat limit versus heavy use of a feature outside the current plan.
  4. Test whether the trigger holds across segments, since it may predict expansion for large accounts but not for small ones.
  5. Ignore total usage growth alone, because usage rising within plan limits is not yet a real expansion trigger.
Executive Capability Standard

What Good Looks Like

A working expansion trigger is built from real usage patterns found in past expansions, flags accounts by behavior rather than a fixed calendar date, and gives account teams the specific signal behind each alert rather than a generic upgrade flag.

Building The Capability (5-Stage Skill Ladder)

1. Learn:Pull your last several expansion deals and review their usage data in the weeks beforehand to find any pattern shared across most of them.
2. Do Manually:Build a simple manual weekly check for accounts crossing your identified usage threshold and have an account team member review it before automating anything.
3. Delegate:Assign a RevOps or customer success operations owner to maintain the trigger logic and refine it as you learn which alerts actually convert.
4. Automate:Build automated usage-based alerts directly from your product analytics into your CRM so account teams see a live flag without a manual weekly export.
5. Buy:Bring in a RevOps or data consultant to help define and validate the initial usage-based trigger if your product analytics aren't yet set up to support this kind of pattern analysis.

How to Get Started

Frequently Asked Questions

How many past expansions do we need to look at before trusting a pattern?

Enough that the pattern shows up consistently rather than in just one or two accounts, though the exact number depends on how many expansions you've had. If your expansion volume is still low, a hypothesis based on a handful of cases is a reasonable starting point, but treat it as a hypothesis to test rather than a confirmed rule until you've seen it hold across more accounts.

Should product or sales own building the expansion trigger?

It works best as a joint effort between product or data teams and sales or customer success. Product or data teams usually have the usage data needed to define the trigger, while sales and customer success know which signals feel actionable in a real conversation. Building it in isolation tends to produce a trigger that is technically sound but useless, or intuitive but not backed by data.

What if an account shows the expansion signal but isn't actually ready to buy more?

That's expected some of the time; a usage signal improves your odds of a well-timed, relevant conversation, but it doesn't guarantee a close. Track how often a triggered outreach actually converts so you can refine the signal over time, rather than expecting a perfect hit rate from the first version of any trigger.

About the numbers

This guide doesn't quote a sourced benchmark. Figures in it are estimates or general guidance, so check them against your own numbers.

Related Guides