Your internal data tells you what merchants do inside your app: billing, usage and support. Public storefront observations tell you what changes around it: a competing app appears, the store redesigns, the merchant upgrades to Shopify Plus or starts hiring. Retention monitoring works best when you combine the two, review the accounts where both point the same way, and start a useful conversation before the merchant decides.
An external observation is a reason to look, not confirmed churn. This guide shows how to tell the difference and what to do with each.
Understand the merchant mix in your install base
You cannot spot unusual changes without knowing what normal looks like. Segment your base before you watch it.
| Segment | Source | Why it matters |
|---|---|---|
| Plan and revenue to you | Your billing | Prioritizes review effort |
| Tenure | Your billing | New and long-tenured merchants churn for different reasons |
| Usage depth | Your product data | Low usage usually comes before an uninstall |
| Shopify plan, including Plus | Public storefront | Larger merchants have more options and more scrutiny |
| Vertical and country | Public storefront | Some segments fit your product better than others |
| Other apps they run | Public storefront | Complementary apps raise switching cost; competing apps lower it |
Look at retention by segment, not only in total. A healthy overall number can hide a segment that is quietly leaving.
Watch for competing app installations
When a competitor's app appears on a store that also runs yours, possible explanations include:
- The merchant is evaluating a replacement.
- They are using the other app for a different use case or store.
- An agency is testing tools during a redesign.
- The other app came bundled with a suite they adopted for another reason.
All four are worth knowing. Only the first is a retention risk, and you cannot tell which from the storefront alone. That is what the conversation is for. For how competitor detection works and its limits, see tracking competitor installs and uninstalls.
Recognize possible contraction or account changes
Other external changes can come before contraction, downgrade or uninstall, or can signal expansion:
| Observation | Possible meaning | Risk or opportunity |
|---|---|---|
| Competing app detected | Evaluation, parallel use, agency testing | Risk |
| Many apps removed or added at once | Redesign, rebuild, new agency, replatform | Either; your app may be dropped in the rebuild |
| Store password-protected or offline | Pause, closure, rebuild | Risk |
| Upgrade to Shopify Plus | More budget, new requirements, stack review | Either; often expansion |
| Hiring a new ecommerce or functional lead | New decision maker who may revisit tools | Either |
| Advertising drops or stops | Seasonality, budget cuts | Mild risk; context only |
| A negative review of your app | Specific frustration | Risk, and the clearest one |
Distinguish an account signal from confirmed churn
Keep two categories separate in your reporting.
| Type | Examples | Status |
|---|---|---|
| Confirmed churn | app/uninstalled webhook received; subscription cancelled; downgrade processed | Fact, from your systems |
| Account signal | Competing app detected; Plus upgrade; redesign; new hire; low usage | Possibility to review |
Signals should create review tasks, not churn forecasts. A signal that turns out to mean nothing is a normal outcome, and you should record it as one.
Prioritize accounts for review
Not every signal deserves a call. Prioritize using two questions: how much the account is worth to you, and how many independent pieces of evidence point to risk.
| One signal | Two or more independent signals | |
|---|---|---|
| High value | Review this week; light check-in | Review now; account owner reaches out |
| Lower value | Monitor; automated or pooled outreach | Review this week |
Strong combinations usually mix one internal and one external observation, for example falling usage plus a detected competitor, or a support escalation plus a redesign.
Use context to start useful customer conversations
Customers know their storefront is public, but a check-in should be about helping, not about what you observed. Good openings refer to their goals and to what has changed in your product.
Illustrative examples (not real customers):
- After a Plus upgrade: "Congratulations on the move to Plus. A few features become available on Plus that teams like yours often set up in the first month. Want a 20-minute walkthrough?"
- During a likely redesign: "If you're working on a new theme, here's how to keep [your feature] in place during the switch so nothing breaks on launch day."
- When usage has dropped: "We've shipped [feature] since you set up your account. It covers [use case]. Worth a quick look together?"
Avoid opening with "We noticed you installed [competitor]." It puts the customer on the defensive and assumes a motive you cannot know. If they are evaluating alternatives, an open question such as "Is there anything you'd like the app to do that it doesn't today?" usually surfaces it.
For signal definitions and relevance windows, see the buying signals reference.
Combine external observations with internal data
Join both views on a stable store identifier, usually the shop's myshopify domain, and keep it in one table your CS team reviews.
| Column | Source |
|---|---|
| Shop domain, plan, MRR, tenure | Your billing (Shopify app billing or your own billing system) |
| Install and uninstall events | Your app's webhooks |
| Usage trend (last 30 versus prior 30 days) | Your product analytics |
| Open tickets and recent escalations | Your helpdesk |
| Competing apps detected, Plus status, recent stack changes | External storefront observation |
| Reviews of your app | Shopify App Store |
| Review status and outcome | Your CS process |
Internal data shows behavior inside your product. External data shows changes around it. Neither is complete alone.
| Shop | MRR | Usage, 30d | Tickets | Competitor | Plus | Next step |
|---|---|---|---|---|---|---|
| Hearth Home Goods | $299 | Down 41% | 2 open | Competitor A | Plus | Review this week |
| Ridgeway Bikes | $99 | Flat | 0 | None | No | Monitor |
| Saltwater Skincare | $199 | Down 12% | 1 escalated | None | New Plus | Expansion check-in |
Measure retention outcomes
Track the process and the results separately.
| Metric | What it tells you |
|---|---|
| Logo retention by segment | Whether you keep merchants |
| Revenue retention (including downgrades) | Whether you keep revenue; downgrades can hide inside a stable logo count |
| Accounts reviewed per week | Whether the process is running |
| Time from signal to first conversation | Whether you act while the signal is relevant |
| Outcome of reviewed accounts | Retained, expanded, downgraded, churned, or nothing wrong |
| False alarm rate by signal type | Which signals are worth reviewing |
If you can, hold out a small, comparable group of flagged accounts that you do not contact, and compare outcomes after a quarter. Without a comparison, it is hard to know whether outreach changed anything.
Where Alert fits
Alert covers the external half of this view. It does not connect to your Shopify store or your systems, and it does not see your billing, usage or support data.
- Customer Lens shows the merchants where Alert detects your app on the storefront, which competitors run inside that base, how the detected base changes over time (new versus lost), and key accounts that also run a competitor.
- App Moves dates installs and uninstalls of your app and your competitors' apps on each store.
- Plus Upgrades, Job Postings and Ad Radar add account context to the merchants in your base.
- Signal export sends these views to CSV so you can join them with your billing, usage and support data.
Storefront detection covers apps that place code on the storefront. If your app works only in the Shopify admin, Customer Lens cannot detect your own installs from the storefront.
Start a 14-day free trial to see which competitors appear inside your detected install base.
FAQs
Can I see why a merchant uninstalled my app?
Not from Shopify's uninstall event alone. Some teams ask in a short exit survey or follow-up email. External observations, such as a competitor appearing around the same time, can suggest a reason but cannot confirm it.
Is a competitor installed alongside my app always a churn risk?
No. Merchants sometimes run two apps in one category for different use cases, test tools during a redesign, or adopt a suite that overlaps with you. Treat it as a reason to check in.
Who should own install base monitoring?
Usually customer success or account management, with product reviewing the patterns each month. Sales may own expansion signals such as Plus upgrades.
How often should we review at-risk accounts?
Weekly for high-value accounts, monthly for the rest. The point is to reach merchants while a change is recent, so a review that happens once a quarter is usually too late.