Campaign strategies to avoid (and why)
Strategic anti-patterns to steer clear of, even when they seem like the obvious choice.
Table of Contents
These aren't edge cases — they're the most common mistakes in in-app experience programs. If you catch yourself reaching for one of these, this article explains why to steer away from it.
The Appcues App is in beta and rolling out gradually. If you don't see it yet at app.appcues.com, it hasn't been enabled for your account. See Introducing the Appcues App for more.
"Set it and forget it"
What it looks like: Building four walkthrough tours, an onboarding Checklist, and a resource center — and calling it done.
Why to avoid it: A static onboarding stack without experimentation, measurement, or iteration looks finished but isn't working as hard as it could. Every campaign should ship with a metric, an experiment design, and brand-appropriate copy. If you're not planning to revisit the campaign after launch, treat that as a red flag.
The exception: Pure deflection experiences (one specific support question), passive resources (a Launchpad with help articles), and static compliance messages can be shipped and left alone.
Experience-first thinking
What it looks like: "Should we use a Flow or a Checklist?" — before defining what outcome the campaign is trying to drive.
Why to avoid it: When the experience type comes before the problem, the recommendation is a guess. Start with discovery: what behavior are we trying to change, for whom, and what does success look like? The right experience type follows from the answers — not the other way around.
A Flow for every announcement
What it looks like: Every new feature, every update, every announcement gets a multi-step Flow.
Why to avoid it: This is the most common over-build pattern. It trains your team to default to the heaviest experience type and trains your users to dismiss everything. Diversify — if you've been using Flows, try a Pin or an Embed instead. Coordinating different experience types to tell a story is what works; repeating the same experience type creates fatigue.
Over-segmentation
What it looks like: A unique segment for every experience. CSV file uploads treated as live segments. Complex "not a member of segment X" logic stacked across many rules.
Why to avoid it:
- CSV uploads are point-in-time snapshots — they're stale the moment you upload them. Use dynamic behavioral or attribute logic instead.
- Stacking exclusion rules across segments slows qualification and becomes unmaintainable.
- A unique segment per experience adds complexity without adding precision — one-off targeting directly on the experience is often cleaner.
Under-segmentation
What it looks like: Treating all admins as one group. Shipping the same experience to every user regardless of where they are in their journey.
Why to avoid it: Two admins on the same plan can be in completely different lifecycle stages. Reach for event-based segmentation — what users have or haven't done — before accepting a broad attribute-based segment.
Wrong-signal segmentation
What it looks like: Segmenting on who you think should see the experience rather than working backward from the outcome you're trying to drive. Over-promoting one feature instead of mapping the combination of experiences that move the KPI.
Why to avoid it: Persona alone is rarely the right signal. Pair it with behavioral data and work backward from the business outcome, not forward from the audience definition.
Wrong-stage targeting
What it looks like:
- Selling expansion to users who haven't activated
- Pushing adoption-stage habit-building to users who haven't discovered the feature yet
- Notifying end users about admin-only actions
Why to avoid it: Every recommendation must pass a stage-appropriateness test. Identify which lifecycle stage the campaign targets before selecting an experience type. If the stage doesn't match the intent, that's a mismatch worth fixing.
Skipping the recovery path
What it looks like: A one-shot Modal or Flow with no way for users to come back to it after dismissing.
Why to avoid it: Every campaign should include three things:
- An opt-out path
- Event tracking on the dismissed cohort
- A way for users to re-enter voluntarily — typically a Launchpad item or a follow-up email
If the user dismisses an experience, that's data — not the end of the conversation.
Ignoring brand context
What it looks like: Shipping copy that feels bolted onto the product rather than native to it.
Why to avoid it: Customer-facing content needs to feel like an extension of the customer's product. If you haven't shared brand or voice guidelines, sort that out before drafting copy at scale. Generic copy isn't done copy.
Celebrating completion rate as success
What it looks like: Reporting that a Flow has an 80% completion rate and treating it as a win.
Why to avoid it: Completion tells you whether users clicked through, not whether they took the downstream action. A Flow can have high completion and drive almost no conversions. Always pair completion with a Key Action that tracks the actual behavior — and on any A/B test, both variants should share the same Key Action.
Defaulting to Flows plus Checklist for every onboarding
What it looks like: Assuming that every new-user onboarding needs a Flows-plus-Checklist stack as the starting point.
Why to avoid it: Flows plus Checklist is a strong combination for teams that can invest in maintaining it — keeping steps current, updating completion logic, and iterating on the sequence. But it's not always the right fit. Evaluate whether your team has the bandwidth and whether the onboarding journey maps well to a Checklist structure. When it doesn't, consider alternatives — like a Welcome Tour plus Launchpad recall — that deliver the same outcome with less ongoing maintenance.