Spam Scoring
Spam spam spam spam spam. We all get it, we all hate it – so what are we doing potentially adding
to the problem by using email as a marketing tool? Discuss.
But seriously, spam is a big issue for email marketing,
and it’s something that you need to consider a great deal if you want your
message to be delivered to as many people on your list as you can.
A good 3rd party email provider will provide a
built-in spam scoring engine for you to use but if yours doe not then do not
worry as there are several others available for free on the web, or as
downloadable freeware you can install your self, to do the job.
The basic way that spam scoring works is that you “send” a
test email through the system, which in turn “reads” it, and returns to you a
score of how much it “looks” like a spam email. This score acts as an indication of how likely that email then is to get
blocked by a spam filter and, by association, not delivered to your intended
recipient.
The spam scoring software will look at the email in two
distinct ways. Firstly, one a
human-style level, by reading the actual copy of the email and assessing how
likely it is to be spam. Offensive words
(no need for me to explain what these are), words with common spam like content
(such as viagara) or common phrases which sound human but are often used to
“trick” or try to trick a user into opening an email (such as “great to see
you!”) are all likely to grab the attention of the spam software and affect
your overall score. You can then use the
report to look back at the copy you have written and re-write it to achieve a
better ranking, and thereby improve the deliverability of your email. This is probably a reason why, if you happen
to work in the pharmaceutical side of things, email marketing probably isn’t
going to be your best route to the target market.
The second way that spam scoring works is to look at the
actual HTML source code of the email and report back on any possible “tricks”
that spammers employ to get through spam filters. These can include things such as having
nothing but images in an email with no text content so spam filters cannot read
the copy, or using copy of a colour that is the same as the page background –
white on white for example. Obviously,
most of us won’t be using these techniques anyway because it doesn’t make sense
in any way, but it’s still worthwhile from a best practice point of view.
The results of your spam score should be used to rework
the email if required to achieve better deliverability. This can sometimes mean that the copy needs
to be re-written, especially the subject line which is very important, and at
times this can seem like a pain for all involved. Developers may not want to
put the copy back in, account managers may state that the client has already
signed off the copy, and the client may think along the lines of “why didn’t
you get the copy correct in the first place”. But, spam scoring must be done
with the email built in HTML as if you were going to send it out, due to the
two different ways that it is “read” by the tools. Scoring on copy alone would not give a full
reflection of the communication.
If you and your client is serious about reaching as many
people as possible on their list, this is a critical step that should be
followed.
Final tests - Preparing
your test list
Before you send the email for real, and even though the
client and other stake holders will have signed off the email as a HTML
document before by viewing it offline, as a print out, or on a website as a
HTML page, it is essential that they get a “live” version sent to them from
within the bulk email system so that they get the actual email version to
review and sign off as it will be seen by each and every one of the recipients
on the list.
In order to do this you will need to prepare a test, or
seed, list containing a set of data for each stakeholder in the project who
needs to sign off.
The complexity of this list will vary depending on how
much personalization there is in your template. If there is none, then you simply need a list with the email addresses
in it. However, for each personalized
element in the email you will need your list of stakeholders to be completed
accordingly. Otherwise, when the
stakeholders receive the email there will be gaps, omissions and possibly
errors in the email which will, of course, make them not sign it off and throw
it back. It will also make you, NOT the other
departments, look bad. Therefore, you
must ensure that whoever it is providing you with data for the seed list
ensures that all elements are in place.
Again, as I said before, it’s not your responsibility to
ensure that all the right people are on this list, that’s the job of the
account team, but rest assured that because you as a digital campaign project
manager are effectively the last link in the chain, you and your team will look
bad – so you need to do everything you can to avoid this.
When you have your seed list, you can upload the data into
the bulk system, and should be able to flag it as a seed list so that when the
actual email goes out live these people are not sent to again, unless they want
to be.
All that remains then is to send the email to the list and
wait for the comments to come back. If
you’ve covered off all of the points above then at this stage you should simply
be receiving sign-off to say it is good to go. If you’ve done something wrong, i.e. a bug, then this will come back to
haunt you so hopefully you’ve checked everything up to this point.
Often, even though everything has been signed off at this
point, clients or other stakeholders will continue to make copy or design
amends to the document. This is an
annoying yet frequent problem and hard to avoid. In general I would recommend being flexible
to a certain point. If the copy amend is
a small one such as a single word, then it isn’t too difficult to do directly
in the template itself. If the amends
are more involved, or there are quite a few of them, it’s probably best to go
back a few stages to the copy writing phase and take it offline again. Use your own judgement, all clients are
different after all, but be flexible as you see fit.
Hopefully though you will get back a sign-off from all
parties to say you can proceed to the “go live” stage and schedule the email
for the actual send.
Uploading your live
data and scheduling the send
Uploading your live data is likely to be exactly the same
process as uploading your seed list, the only difference being the way you flag
or classify the data within the 3rd party database. Depending on the number of people in your
list it could take some time so be prepared for quite a wait, even once the
data file itself has been uploaded.
When you have uploaded the data you will most likely be
presented with a data report which will list the following items:
· New records
· Amended records
· Non-valid addresses
· Bounces
· Un-subscribed
· Total addresses in batch that will be sent
You should make a note of these numbers and send them to
the account team for them to record. These numbers are very useful and will give you an indication that the
data has been uploaded in the right format and all is as it should be. If, for example, you’ve uploaded a dataset of
100,000 recipients, but the system is reporting that there are only 30,000 in
the send list, something major has gone wrong and you should pause the process
and speak to whoever supplied the data.
It’s perfectly normal to lose a small percentage of
recipients to bounces and un-subscribes. These are simply addresses that the 3rd party system knows,
probably from your previous campaigns, are no longer valid or no longer wish to
receive communications from you. This
will take time to build up, so logically the very first email you send from the
system should have none, but once you have sent a few there will be numbers in
this. Therefore, aside from the very
first send, it is extremely unlikely that you will send to 100% of your
list.
A good 3rd party system will give you the un-subscribes,
bounces and other address which you can no longer send to in an exported file
such as a CSV of XLS. You should take
these files and give them to the account team. They in turn should feed them back to the data provider so that these
names can be removed from the database and not be sent back to you in the
future – it’s just good data practice to keep your lists clean and tidy.
Scheduling the send
Once you’ve uploaded your data, and got full sign off from
all stakeholders – not just the account team but the client as well – you’re
ready to send the email.
You should be able to just hit a button and send the
email, but my advice is to resist this urge and use the scheduled sending
option that most 3rd party system provide, and set the email for
some time in the near future.
Fifteen minutes delay is often a good idea even if you
want it to go “as soon as possible”. The
reason for this is that it still gives you, at the very last hurdle, a little
bit of a buffer zone for protection. You
don’t need to tell other people that this has been done, as far as they are
concerned the email has been sent. What this delay gives you though is a little
bit of space to breathe incase you or any other member of the project team notices,
at the final hour, a problem or error with the email.
Most times this will not be the case, but I have
personally experienced the horror of sending an email to thousands of people
only for an account manager to run up five minutes after the send shouting “you
haven’t sent it yet have you? It’s the wrong list!” Even with all of your sign
off procedure in place mistakes can and do happen – so this last bit of room
that you yourself put into the process gives you that last bit if
professionalism that, should disaster strike, allow you to rectify it at the
eleventh hour.
To wrap up…
And that’s it. If
you follow all of the above, not to the word, but use the various points as a
general guide to best practice in email process, you should cover off most eventualities
that can and do arise. Email sending
isn’t a dark art – it’s pretty straight forward if you look at it in it’s
component parts. But what is almost
frightening about email is the fact that you can, in a very short period of
time, send a message out to thousands of users in which there can be a
potential error, be it technical or communication based. Unlike web pages, you cannot take them back
once you’ve hit send to limit the impact. But, even with this in mind, by
applying careful process and checks to the campaign you can limit the risk
involved at all stages.
One final point – although this process may seem long
winded at time, and you may get hastle off other stakeholders for applying it,
it really is there to help you and them communicate effectively to your target
audience and, at the end of the day, that’ what everyone wants to do – so stick
to your guns.
Howard
–
January 30th, 2007 at 4:14 am
Bmw 1999
Select Year 2007 2006 2005 2004 2003 2002 2001 2000 1999 was a BMW 32