Loading tutorials…
Loading tutorials…
Mailchimp gives you three ways to slice a single audience: Segments, Tags, and Groups. They overlap. They confuse new users. Picking wrong locks you into years of duplicated work. Here's the decision tree.
Who this is forMailchimp account holders with at least 1,000 contacts who feel their segmentation is messy. If you have 'newsletter signup,' 'VIP,' and 'product launch' as separate audiences instead of one audience with metadata, this tutorial fixes the foundation.
What you'll need
Step 1
Segments = dynamic queries on your audience. Tags = static labels you apply manually or via automation. Groups = subscriber-selected preferences.
SEGMENTS: dynamic. Mailchimp queries your audience based on rules (e.g., "opened any campaign in last 30 days"). Updates daily, automatically. Use for behavioral targeting.
TAGS: static. You (or an automation) apply tags to contacts manually. Stays until you remove. Use for explicit metadata that doesn't change based on behavior ("attended-webinar-2026," "high-value-lead," "support-ticket-opener").
GROUPS: subscriber-selected. Shown to subscribers on signup forms or preference pages. They check which groups they want ("Newsletter," "Product Updates," "Sales Alerts"). Use for content preferences.
Critical distinction: Segments READ data. Tags STORE data. Groups CAPTURE data from subscribers.
Step 2
Tags = facts about a contact that don't change based on email behavior. Lead source, conversion event, deal stage.
GOOD tag examples:
- "source:facebook-ad" (where did they sign up?)
- "lead-magnet:saas-checklist" (which lead magnet?)
- "attended-webinar-2026-q2" (event attendance)
- "purchased-product:hoodie" (specific product purchase)
- "support-tier:premium" (customer status)
BAD tag examples (these should be segments instead):
- "opened-last-30-days" (behavior — use a segment)
- "clicked-pricing-page-last-7-days" (behavior — segment)
- "VIP" (depends on $value behavior — segment)
Rule of thumb: if the data CHANGES based on email behavior, it's a segment. If it's a permanent fact, it's a tag.
Step 3
Groups = preferences subscribers actively choose. Newsletter type, content frequency, product interests.
GOOD group examples:
- "Newsletter type: Weekly digest / Product updates / Sales alerts" (frequency preference)
- "Interests: Men's / Women's / Kids" (apparel preferences)
- "Industry: SaaS / E-commerce / Agency" (B2B segmentation)
BAD group examples:
- "Customer status: VIP / New / Lapsed" (this is segment territory, not group)
- "Source: organic / paid / referral" (use tags, not groups)
Rule of thumb: if subscribers should be able to opt INTO or OUT OF the category themselves, it's a group. If you assign it, it's a tag.
Build groups in: Audience → Settings → Groups → Create Groups. Choose group display: checkboxes (multi-select), radio buttons (single-select), or hidden (you assign).
Step 4
Most lists need ~10-15 segments. RFM (Recency, Frequency, Monetary), engagement-based, and source-based.
Engagement segments (use for re-engagement and deliverability):
- "Engaged 30d" = opened or clicked in last 30 days
- "Engaged 90d" = opened or clicked in last 90 days
- "Inactive 90d" = no open + no click in 90 days
- "Inactive 180d" = no open + no click in 180 days (sunset candidate)
RFM segments (e-com):
- "VIPs" = 4+ orders in last 365 days AND $300+ in last 90 days
- "New customers" = 1 order in last 30 days
- "At-risk" = 1+ orders + no order in last 60 days
- "Lapsed" = 1+ orders + no order in last 180 days
Source segments (use tag-based logic):
- "From paid ads" = Tag contains "source:meta-ad" OR "source:google-ad"
- "From content" = Tag contains "source:blog" OR "source:podcast"
Build each in: Audience → Segments → Create Segment.
Step 5
If you've been using audiences/tags/groups inconsistently, document what you have, then rebuild.
Audit current state: Audience → Settings → Groups (list all groups). Audience → Tags (list all tags). Audience → Segments (list all segments).
Document which "thing" each one actually represents (behavior? preference? fact?).
Build the canonical structure: segments for behavior, tags for facts, groups for preferences.
For each existing tag/group that's actually behavioral data, build the equivalent segment.
Delete or archive the now-redundant tags/groups.
Document the new structure in a shared doc: 'Our Mailchimp taxonomy.' Onboard team members to this doc — taxonomy drift over months will undo all this work.
Step 6
Campaigns target segments or groups. Automations trigger on tags. Use the right tool for each job.
CAMPAIGN TARGETING: in the campaign builder, choose your audience, then add filters/segments. Use the segments you built in Step 4.
For "VIPs only" campaigns: target the "VIPs" segment.
For preference-based campaigns: target the relevant group (e.g., subscribers who selected "Sales alerts" group).
AUTOMATION TRIGGERS: use tags for explicit trigger points.
- "Tag added: attended-webinar" → trigger post-webinar follow-up automation.
- "Tag added: purchased-product:hoodie" → trigger product-care automation.
AUTOMATION FILTERS: use segments + groups inside automation logic.
- Welcome series → filter "Is in segment: Engaged 30d" → only send Email 4 to engaged subscribers.
Step 7
Segment thresholds should match your current scale. VIP at $20K/mo isn't the same as VIP at $200K/mo.
Calendar reminder every 90 days: re-pull segment counts.
VIP segment should be ~5-10% of active subscribers. If it's 30%, the threshold is too loose. If it's 1%, too tight.
Engaged 30d segment should be 40-60% of active subscribers in a healthy account.
Inactive 180d segment should be <30%. If higher, your re-engagement automation isn't working or you have a deliverability issue.
Adjust thresholds and re-document in your taxonomy doc.
Common mistakes
Using multiple audiences when one would suffice
What goes wrong: Contacts duplicated across 2-3 audiences means billing 2-3x. A 10K-contact 'list' across 2 audiences bills as 20K — jumping plan tiers unnecessarily. Compound over years and it's $1-3K wasted subscription.
How to avoid: Consolidate to one audience. Use tags, groups, and segments inside. Multi-audience is for legitimately separate brands only.
Tagging behavioral data instead of segmenting
What goes wrong: Tag 'recent buyer' applied today is still 'recent buyer' in 18 months because tags don't auto-update. Campaigns targeted at this tag include people who haven't bought in a year. Targeting accuracy drops 50%+.
How to avoid: For any behavioral query (opened, clicked, purchased in last X days), use a segment. Segments auto-update; tags don't.
Using groups for things subscribers shouldn't see
What goes wrong: Customer status group ('VIP / New / Lapsed') exposes internal labeling to subscribers on the preferences page. Customers see 'VIP' or 'Lapsed' next to their account and react negatively.
How to avoid: Use tags for internal labels subscribers shouldn't see. Groups should only contain options subscribers WOULD legitimately choose between.
Inconsistent tag naming (no taxonomy)
What goes wrong: Tags like 'vip', 'VIP', 'vip-customer', 'VIP Customer' all exist on the same audience because team members tagged ad-hoc. Segments looking for 'VIP' miss subscribers tagged 'vip'.
How to avoid: Document a tag taxonomy: lowercase, prefix-grouped ("source:facebook-ad," "purchased:hoodie"). Use the same format everywhere. Audit tags monthly for typos.
No engagement-based segmentation for deliverability protection
What goes wrong: Campaigns send to entire audience including 180-day inactives. Disengagement signals tank deliverability. Open rate drops from 28% to 16% across the list over 6 months.
How to avoid: Always exclude "Inactive 180d" segment from regular campaigns. Run them through re-engagement automation instead; sunset non-responders.
Static segments that never get recalibrated
What goes wrong: VIP threshold set at $200 lifetime spend in Year 1 doesn't fit Year 3 when the list has grown 10x. VIP segment is now 40% of list, meaning 'VIP' is meaningless. Targeting becomes noise.
How to avoid: Calendar reminder every 90 days. Re-pull segment distributions. Adjust thresholds to keep VIPs at 5-10% of active list, engaged at 40-60%, inactive at <30%.
Recap
Done — what's next
How to set up a Mailchimp account from scratch (audience, compliance, sending domain)
Read the next tutorial
Hand it off
Segmentation is the foundation every campaign sits on. Get it wrong on Day 1 and you pay for it forever. A specialist who's structured 50+ Mailchimp audiences will design your taxonomy in 2-3 hours and document it for the team. Typical engagement is $300-500 at $14-16/hr.
See specialist rates
Multiple audiences ONLY when you have legitimately separate brands or business lines (different websites, different customer bases, different products). 99% of accounts should use ONE audience with tags/groups/segments inside. Multi-audience doubles billing and creates duplicate work.
Not directly, but the workflow is: build a new segment with the same logic, verify it returns the same contacts (give or take), then deprecate the old tag. For tag → segment migration of behavioral data, the segment will be MORE accurate because it auto-updates.
200-500 active tags is normal for mature accounts. Above 500, audit for redundancy and typos. Above 1,000, you've probably been using tags for things that should be segments — restructure.
No. Segments are included on every Mailchimp plan including Free. They're dynamic queries against your audience, not stored data. Build as many as you need.
Yes. Mailchimp segments support AND/OR logic with up to 5 conditions on Free, more on paid plans. For complex segments (e.g., "VIPs who haven't opened in 30 days"), combine: "Has tag VIP" AND "Has not opened email in last 30 days."
No — segments are evaluated when campaigns send, not in real-time during the day. Mailchimp's infrastructure handles segments on millions of contacts. Don't over-engineer for performance; build the segmentation you actually need.
Mailchimp
Mailchimp's onboarding hides the decisions that matter most — audience structure, single vs double opt-in, and sending-domain authentication. Skip them and you'll be untangling them in 6 months. Here's the setup that doesn't rot.
Mailchimp
Four automations cover 80% of the email revenue most lists ever generate: Welcome, Birthday, Anniversary, and Re-engagement. Each takes 30-60 minutes to build right. Skip them and you're sending the same campaign to everyone like it's 2009.
Mailchimp
Deliverability is the silent killer of email revenue. Open rates dropping from 28% to 16% feels like 'the algorithm changed,' but it's almost always one of three things: authentication, warmup, or list hygiene. Here's the playbook.
Mailchimp
Open rate dropped from 28% to 16% in 30 days. The instinct is 'subject lines.' Usually it's not. Deliverability, list hygiene, and audience drift account for 70% of open-rate drops. Here's how specialists isolate the cause.
Mailchimp
DIY email marketing is a great call — until it isn't. Email should be 20-35% of total business revenue for most online businesses. If yours is at 8-12%, the gap is your list isn't being worked. Here's the honest framework for when to hire.