Almost Everything You Wanted to Know About Email Marketing
Everything You Wanted to Know About Email Marketing Silverpop
Jan 13, 2015
Welcome message from author
This document is posted to help you gain knowledge. Please leave a comment to let me know what you think about it! Share it to your friends and learn new things together.
Transcript
Almost Everything You Wanted to Know About Email Marketing
2
You have email marketing questions. So do we.
We also have some answers. Lots of them.
3
How to use this document:
1. Browse from the start; or2. Go to a specific section of interest.
4
Permission
Permission is the foundation.
You are correct. The U.S. CAN-SPAM Act
does not require permission to email.
See number 1.
Permission IS required by law in the EU, Australia,
New Zealand, Korea, Canada.
The theory that relevance trumps permission is just
that – a theory.
Recipients will punish you hard with SPAM
complaints if you don’t have permission.
Recipients will still punish you with SPAM complaints
if they don’t remember giving you permission.
ISPs tell you to gain permission. They monitor/block by engagement and
complaints.
Single, Double or Confirmed
Opt-in?
Yes. Any approach will do.
Single opt-in = higher completion rate; lower
quality.
The vast majority of companies use the single
opt-in method.
Double Opt-in = An extra step, fewer opt-ins.
Use Double Opt-in when you use aggressive
acquisition methods.
Confirmed Opt-in sends an email notifying the
new subscriber they’ve opted in – and to
unsubscribe if in error.
20
List Rental/Buying
Simple really. Don’t EVER buy email lists.
List rental? Work with a list broker
with a stellar reputation.
Yes, there’s a difference in the meaning of the
words “buy” and “rent.” Learn the difference.
Buy = someone sells you a list that you send.
DON’T DO IT! You are a spammer.
Rent = a third-party sends your email to people that
have given permission. You must then convince them to
opt-in to your program.
Don’t expect huge growth from list rental.
Focus on deeply qualified leads that will drive a
higher percentage of opt-ins.
Map your offering to the original reason for opt-in. Contest entrants probably
don’t buy $200K timeshares.
Never risk your primary sending IP address with
outside names.
30
Welcome Emails
Use a Welcome Email. Better yet, a Welcome
Series.
Better yet – an onboarding program.
No welcome email? Then you are leaving to chance
the new subscriber experience.
Explain your value. Link to preferences. Provide
an incentive.
Welcome emails help set expectations. Can solve
problems.
Ask progressive survey questions during the
Welcome Series to build your database.
Don’t lead with a coupon. Let full-price buyers act
first!
Classify new users into active and inactive, and try incenting inactives with a small discount.
Clearly outline your return, shipping and other key policies.
Ask for the first order in your final email of the
series.
41
Deliverability
Email marketing is NOT Direct Email. ISPs are NOT
the same as the Post Office.
You the marketer are responsible for your email
deliverability – not your email service provider
(ESP).
Your ESP helps you with infrastructure and
strategies – but there is no magic bat phone to the
ISPs.
Permission is the foundation to high engagement, low complaint rates.
Even the best email marketers receive spam complaints. The key is to keep the complaint rate
LOW.
Why do I receive spam or abuse complaints?
Your recipients think you send too many emails; they are irrelevant; or they don’t remember
subscribing.
Most ISPs will block/filter your emails if you receive more than 3 complaints
out of 1,000 they receive from you.
How do we reduce spam complaints then?
Don’t Hide the Unsubscribe Link. Make it
Easy to Unsubscribe.
Use a highly recognizable “from” name.
Use a consistent “from” name.
Don’t use spammy or potentially deceitful
subject lines.
Authenticate your emails. DKIM. SenderID.
Shared/Pooled or dedicated IP addresses?
Companies that send infrequently and low
volumes should probably be on a shared IP.
Like your toes in winter, your sending IP addresses
need to be ‘warmed’.
A tiny percentage of your messages will be
‘blackholed’ by ISPs.
Non-responders will slowly and silently kill
your deliverability stats. Manage them proactively.
Test deliverability statistics of each new
template using a deliverability monitoring
service.
62
Opt-in Forms
Capture the data required to deliver relevant emails.
Move opt-in forms on the bottom of your home
page to the top/above the fold.
Do it now. We’ll wait. Oh, IT, Ecommerce, etc.
doesn’t want to give up that real estate?
Tell the roadblockers this: Companies that move their opt-in forms up typically see opt-in
conversions increase 50% to 500%!
Allow users to opt-in with existing social network
credentials, and capture that rich data!
68
From and Subject Lines
69
Yes, from or sender names matter? A lot. So
choose wisely.
70
Use your most recognized and trusted brand name.
Don’t veer.
71
Yes, use different sender names for different email
streams. But always incorporate your brand
name.
73
Resist the urge to use a person’s name.
74
But I see everybody doing that these days. Is that
your reason for doing it – everybody is?
75
Do you know who John Doe is? I didn’t think so. So why use a person’s
name?
76
Yes, there are exceptions. If the person’s name is
also a recognized brand in itself.
77
Personal names are also okay if you're automating email streams on behalf
of a salesforce.
Subject lines can make a huge difference. But your brand/sender name is still the most important thing.
79
Subject lines drive opens – AND clicks and
conversions.
80
Short. Medium. Long. Any length subject line length
works.
81
Research studies suggest that longer subject lines
outperform short and medium length.
82
But, keep the key information within the first 40-50 characters.
“Free” Use it. It works. Look in your inbox if you
don’t believe us.
84
Test subject lines all the time. It is the easiest
thing to test.
85
‘From’ names are just as easy to test, but DON’T.
86
Pay attention to automated SPAM scoring
tools.
87
Creative
88
Use personality and a real human’s voice WITHIN the
email body.
89
Feature pictures of actual employees within the
emails.
90
Related: Avoid cheesy stock photos of perfect people, shaking hands, lightbulbs, globes, etc.
91
Use “alt” tags to provide context when images
don’t automatically load.
92
Use “bulletproof buttons” – HTML text and tables behind you CTA image.
93
Keep key customer actions in the same
location across templates.
94
Show more than you tell.
95
Designing Emails for Mobile
96
Take a “mobile first” approach. Go ahead,
Google what that means…
97
OK, so design your emails and landing pages to be
friendly for mobile devices. Period.
98
Smartphones are everywhere, and people
have email in their pocket.
99
The finger is the new mouse, just 40 pixels
wider.
100
Don’t make a user side scroll or zoom to read
your content.
101
How Do We Measure Success?
102
Think in terms of “process” or
“operational” and “output” or “success”
metrics.
103
Output/Success: Measures against goals
and objectives, i.e., conversion, revenue,
retention, cost savings.
104
Process/Operational: Measures email tactics ,
i.e., opens, clicks, unsubscribes, spam
complaints.
105
Except perhaps if you are a publisher.
106
Open rates DON’T matter
107
Are you crazy? People have to open their emails to click and take action.
108
You are correct. But just because recipients open an email, doesn’t mean
they’ll take action.
109
Opens and clicks are irrelevant, conversion
matters.
110
Did more people buy or sign up for your event?
111
Focus your email metrics on those that support
achieving key company objectives.
112
Ever met someone that got a raise for increasing
open rates?
113
I didn’t think so. Focus on: conversion; revenue;
engagement; cost savings. These things
matter.
114
Metrics Benchmarks
115
There are no average open rates. Industry
matters.
116
Still want to know? OK. How does 23% sound?
Happy now?
117
Go ahead, tell your boss you are “average.”
We’ll wait.
118
Thought so. How’s that raise looking
again?
119
There’s average, lower quartile and upper
quartile in each industry. Higher is better.
120
Oh yeah, we have a benchmark report. You should grade yourself.
121
The real metric: interaction.
122
Frequency
123
The best frequency is the one that delivers the best combination of revenue,
LCV and acceptable churn.
124
The more behavior-based the content, the less you
need to worry about frequency.
125
Ignore frequency controls for your most behavior-
driven emails.
126
Manage frequency across ALL campaigns from your
brand. See it from the customer’s viewpoint.
127
Test frequency. This is your only answer.
128
Re-test frequency every 90 days to keep your
approach fresh.
129
When to Send
130
The best day to send email? Monday. Tuesday. Wednesday. Thursday.
Friday. Saturday. Sunday.
131
Everybody has their own opinion…
132
Send based on historical recipient open and click
times.
133
Consider multiple time zones within a list.
134
Weekends rock for some brands. Give it a try…
135
Inactives / Reactivation
136
Inactives are a HUGE problem for almost every
company.
137
The average company can expect that between 30%-
50% of their email database is inactive.
138
So how do you define an inactive subscriber?
139
Inactives combine timeframe + email
activity + purchase/offline behavior.
140
For example, no email opens and clicks or
purchases for 6 months.
141
But every company must develop a definition that
makes sense for their business.
142
Determine who your Inactives are. Then target
them differently.
143
ISP algorithms are watching whether your recipients are opening
and clicking.
144
Sending “we want you back” emails to inactives typically only activates up
a few percent.
145
Proactively re-engage or purge inactives while adding new records to
maintain audience.
146
Wanna drive your metrics through the roof? Kill inactives mercilessly.
147
Reactivate via a 2-3 step program with escalating
offers. No action = purged record or, reduced
frequency.
148
Tell the user you're taking them off the list, and give them a way to prevent it.
149
Reactivation programs are not the answer. But
do them anyway.
150
Preference Centers
Implement a world-class preference center.
Seriously. Implement a world-class preference
center.
Ask for what you need to segment, provide relevant
content.
154
Empower new subscribers through choice.
155
Step 1) Capture birth date. Step 2) Send Happy
Birthday emails.Step 3) Print Money.
156
What’s the right number of form fields?
8
157
Just kidding. Use common sense. Test It. Balance data needs with lower form completion rates.
158
Be radical. Consider a database keyed on
something other than email.
159
Use existing social credentials during sign
up.
160
Let users control which communications they
receive and each one’s frequency.
161
Support recipient choice, and know that some will opt-out of everything if
you act shady.
162
Don’t force a full opt-out. Offer subscribers the
option to pause communications?
163
Testing
Test. Test Again.Test everything.Test, test, test.
165
But why? Because testing is the only real answer to all your darn questions.
166
Where should I start?
167
Most email marketers start with subject lines because of the ease of
testing.
168
When testing, use your ultimate goal (revenue
and other conversions) to determine winners.
169
What else can we test: pre-header text, layout, copy, personalization,
offers, buttons, timing – pretty much everything.
170
Only conduct true A/B tests – random splits in
parallel.
171
Test 10% of your list with each variant, then send the winner to the other
80%.
172
Use A/B testing to prove or disprove pre-defined
assumptions.
173
Don’t go crazy testing ‘from’ names. In that
case, consistency wins.
174
Misc./Resources/Final Notes
175
Stop saying (and thinking in terms of) ‘e-blast’.
ESPs are different. Find the one that is right for
YOU.
177
Balance existing need with future plans, and
pick a vendor who works for both.
178
Pay attention to 4-5 KPIs send-over-send, week-
over-week and year-over-year.
179
Spend less time modifying lists in Excel, and more time building rules-driven programs.
There are no short cuts.
Don’t agree with these Slides?
Then Yes, it depends and test it.
Thank you!
On Twitter: @Silverpopwww.slideshare.net/silve
rpopwww.silverpop.com
Related Documents
