Quick summary: Every rule firing at the same priority puts renewal risk in the same pile as routine updates. Here are eight adjustments that can change that.
Technology has made it possible to stay up to date on customer inquiries, interactions, requests, and a wide array of other actions. But that capability can become too much of a good thing.
Being overwhelmed by alerts is pretty common among customer success teams. Faced with dozens of notifications every week, your CSMs may begin to dismiss them on reflex.
The danger of alert fatigue.
When CSMs start to ignore alerts because they’ve become noise, they’re likely to overlook the ones that represent real account risk.
For example, say you are running a CS team at a mid-market SaaS company with 600 customers and 6 CSMs who receive about 45 alerts each week. They’ve probably stopped reading most of them. Maybe one rule in particular is resulting in a high number of alerts— “license utilization below 80%.”
While this is hypothetical, the reality is that this is a pervasive problem with high stakes. Bottom line: it’s putting revenue at risk.
The good news is this can be fixed.
Related: What if the way you define customer value is wrong?
Tip 1: Measure the noise before you cut it.
Measurement matters. Quantifying the impact of alerts can help you defend any adjustments. For instance: look at how much volume each rule generates and which CSMs receive it. Then, measure the action rate. How often does each alert drive customer interactions?
Make sure that all alerts can pass the following criteria:
- They are actionable.
- They are owned by a specific person.
- There is meaningful business risk if it’s ignored.
If the answer to any of these is no, the alert shouldn’t be landing in someone’s inbox.
Tip 2: Tier by business impact.
Not all alerts are equally actionable, and the frequency of rules firing isn’t an indication of severity. Considering business impact and tiering alerts can provide the context to reduce CSM stress. For instance, tiers like “act today,” “watch this week,” and “information only” modulate the need for action and minimize angst among your team members.
What’s of most value to you is a measure of the potential impact on renewal and revenue. Those impacts vary.
For instance, a recent NPS of 5 from a previously “green” customer needs immediate action,
Of course, alerts notify you about both risks and opportunities. Opportunity alerts like expansion cues, adoption milestones, and reference-customer-ready accounts are important and should be put through the same test and tiers.
Tip 3: Give every alert an owner and an action.
Sending alerts to your entire team drives inaction because everyone assumes that someone else will own the issue. Typically, no one acts.
Tie every alert to a workflow and to the CSM who owns it. Then tie it to a concrete next step, whether a task, a play, or a predefined response.
Instead of sending pings, send to-dos.
Tip 4: Segment so the signal fits the account.
Numbers are relative: 15% usage for a single-product SMB is a big deal. For a multi-product enterprise, it can be just normal seasonal noise. Setting relevance by account attributes not only addresses fatigue for your CSMs but boosts the value of your service to your customers.
There are a number of relevant attributes you might focus on depending on your accounts that go beyond ARR. For instance: product complexity, strategic weight, current health, and your coverage model. A high-touch strategic account needs granular alerts. A long-tail, self-serve account doesn’t need individual pings, but automated plays that trigger when thresholds are crossed.
Tip 5: Let a health score do the filtering.
Instead of separate alerts on usage, support tickets, NPS, and logins, use a composite score for greater relevance with less fatigue.
This is the change that will have the most impact. Two design choices can kill most false alarms:
Focus on trajectory rather than absolutes. For instance, alert based on the direction and speed a score is moving rather than its current value. A score of 7 may be problematic if it was 9 last month; or it may be fine. A single value can’t tell you which is the case. Looking at velocity can help separate a decline from noise.
Set thresholds by segment, so that expected behavior for a group doesn’t trip the alert wire. Onboarding accounts and long-tail accounts don’t share a definition of normal, and a score built per segment keeps their differences from reading as risk.
Changes in healthy accounts can be surfaced by adding a relationship/sentiment dimension to your health score. This can help you spot accounts where the relationship is quietly degrading.
Tip 6: Consolidate what’s one problem.
If a customer’s onboarding is slipping, you don’t need five separate alerts, but one alert that says, “This account is behind on their journey.” If a key stakeholder is disengaged, you don’t need an alert per user activity check, but one relationship-score change that surfaces the whole picture.
Route using your health score so related alerts collapse into one account-level change. Alert at the account level, not per user. Set alerts to fire the first time a condition is met, so nothing re-fires daily. And once a CSM has dismissed or acted on an alert, it shouldn’t resurface. Let low-priority items roll into a digest.
Tip 7: Give leaders one portfolio view.
Your CSMs aren’t the only ones drowning. Your leadership team can also benefit from a consolidated view of portfolio health.
For instance, use one portfolio view of health-score changes, at-risk segments, and open risk-signal counts instead of scattered inboxes to build a capacity check on who’s overloaded. Pull alert volume by CSM and rule periodically to see who’s getting hammered, and by what. Review these on a set cadence.
Tip 8: Treat alert hygiene as a habit.
The value and relevance of alerts don’t remain static over time. Scheduling quarterly reviews offers an opportunity to retire unnecessary rules, retune thresholds and re-check segments.
