Thursday, August 17, 2006

email testing

why don't people test their emails?

i don't just mean testing for accuracy, i mean testing different subject lines. different copy. different creative. even different tone.

the studies are out there - if you test you will improve your results.

yet, still a resistance to test.

we have a feature that allows for ab testing in our software. it is covered in every training session, yet no one uses it.

i think the story is different when the emails are a primary revenue source but for many clients they choose not to test.

they use the copy/creative that has been approved internally and don't find out what their audience prefers.

is it too little time? not enough training or awareness? they just don't care?

i have not got this one nailed down yet but i think it is about what they are measured on.

if your neck is on the line to get the email out on time and with no mistakes then where is your focus going to be.

i think this is an educational effort to the next level of management.

the metric should be about results. it could be registrations or downloads but it needs to be tangible.

i think the the missing piece is a well thought out strategy on email.

email is easy to setup and send so nearly anyone can do it. it can become fragmented within larger organizations.

think about why you are talking to your customers and what you expect them to do.

try testing something in your next campaign. subject line, from name, copy or creative it does not matter - just try testing something.

i think people will be surprised by their results.

the milk house...


i had to share. this is the milk house. my office away from the office.

i write here.

i code here.

this is my thinking place.

Wednesday, August 16, 2006

Email Metrics - Outcomes

I recently commented on the Marketing Sherpa blog about email open rate metrics. I mentioned a feature that we have in our email marketing software called outcomes.

Whereas your open rate is not an accurate number, your click rate should be but to get more analysis from the success of your campaign you want to know more than a link was clicked on.

That is where we use outcomes. As an example, if you were promoting a webcast and directing people to a sign-up page it would be nice to know not just how many people you sent to that page from your email but how many actually registered. Using outcomes you can track when that recipient hits the confirmation page. It is an awesome way to show ROI for your email efforts.

here is the official word from the user manual...

Outcomes

inMailBox allows you to track pre-defined user actions that might be outside of the
focus of your campaign message, anywhere on the target website.

You simply place a snippet of HTML code on the page(s) you wish to monitor and if a user eventually arrives on this page (from your inMailBox campaign), the system will capture the event and add it to the information in Report Manager.

- Once an Outcome is defined, it will be tracked in all subsequent campaigns.
- If creating custom code, make sure you first give the Outcome a unique name.
- Copy and paste the code into the target webpage – just under the body tag.

Monday, August 14, 2006

Email Open Rate Metrics

I am always asked about this and there is always a swirl of confusion.

Email open rates are not an exact measure.

I just read a very good summary of the problem on the Marketing Sherpa blog.

The Annoying Imprecision of Email Open Rate Metrics

There is some good information in the comments too.

One thing i do agree with and tell clients all the time is that your open rate should be looked at over time. You can see trends and pick up problems there.

Sunday, August 13, 2006

RSS vs. Email: Are they really in competition?

i have seen alot about the subjet. many have predicted that RSS will kill email because it is a pull technology and it can get around the deliverability issues. so far email is safe.

will that change?

when you look at RSS, it is about one to many communication - it is syndication. it does that well.

when you look at email, it can be used as a one to many broadcast tool however it is best as a one to one medium. the power of email lies in its ability to form conversations. RSS can't do that.

i think in the new world of marketing - building conversations, one to one relationships, dialog, trust, that is where the consumer is taking us. the broadcast model has it place and i am not ruling it out, but to reach the saavy, the weary online consumer, you have to go one to one.

so, will RSS kill email? i don't think so.

will it carve out its own niche and be another tool, that when used effectively is killer? yes.

Friday, August 11, 2006

how do I get more traffic to my blog?

I must be the most common second question of a new blogger. "How do I get more traffic?"

The first question is how do i setup a blog?

I found an interesting article that gives some suggestions on the site askdavetaylor.com.

Thursday, August 10, 2006

newsletter sign-up form fields

How many fields should i have on my email newsletter sign-up form?
Which fields?

Simple - you need one. Email address.

The more fields you add the higher the barrier for people to sign-up.

Think of the analogy of starting a conversation with a stranger. Do you walk up and ask for all of their personal info? if you did, do you think they would give it to you.

You start a conversation by sharing a piece of information each. "Hi, my name is Chris. What is yours?"

As the conversation progresses you trade info back and forth and build trust.

The same is true for email. It is a conversation and you need to build trust.

If you live up to your end of the deal, you can ask for more info later.

All you need to start the conversation is an email address.

milk house thoughts?

ah yes. the benefit and draw back of starting a new company is the ability to work from home.

it is great because...i work from home and it sucks... because i work from home.

I bought a 70 acre 150 year old dairy farm out in the country and when things got rolling i converted the old milk house to my office.

It is a little, rustic 12x15 foot building about 250 feet away from the house.

i have high speed satellite, a firewall and wireless lan. Everything you need in a milk house.

i have spent a lot of time out here designing software, coding and everything else required to get the company running.

So that is where the name comes from...

image hosting

another client question...

when we send out html emails should we host the images on our webserver or with you (the esp) ?

First, do you need to use images?

Test it out. Try a campaign with text only, formatted text and then with images. See which one has the best results. If you don't need the images to get your message out - then save them.

The answer to the question is, it depends.

Do you control your site? Can you ensure that that the images won't be moved, changed or deleted? Does your site have a fast, highbadwidth connection to the net?

If you answer yes to all of the above, then go ahead. Your place or mine it does not matter.

However, if you are not sure about the above, then host them with your ESP. It is safer.

back to basics

i run a small email service provider (esp) and we offer hosted self serve or full service email marketing to clients.

somedays, you just have to step back and remember that not everyone has spent as much time working with email as you would imagine.

i am dealing with a new client now who has been having problems with their emails not getting delivered.

i started in explaining about SPF and spam filters and advanced whitelist tools - then i discovered they had not even had an opt out link in their emails.

in canada, you legally must have a simple unsubscribe method. they did not. worse, they did not know they needed one.

all that to say two things...

first, start with the basics. i assumed a level of knowledge and sophistication which did not exist.

second, the role of an esp is not only about delivering the technology but being a partner to navigate through the complexities of email marketing.