You already know how to send a simple campaign. This course is about what to do when one message sent to everyone at once stops being enough. Quanda offers three advanced campaign types and each of them solves a different problem. Before you start clicking, you need to know which one to choose. Picking the wrong type is the most common reason an advanced campaign ends up not working.

One question that decides the type

Before you create a campaign, answer two things. How many different messages do you want to send, and what determines which recipients get which message.

• One message, everyone the same, one send date. That is the simple campaign from the beginner course, we will not cover it again here.
• Several messages in a row at a regular rhythm, every recipient goes through the whole sequence. That is an email series.
• One message in two variants, and you want to find out which one performs better. That is a campaign with A/B testing.
• Several messages where the next step depends on whether the recipient clicked a specific link. That is a logic campaign.

Email series: a fixed rhythm with rolling entry

A series sends several different emails at a regular interval, for example once a week. Use it to welcome a new customer, to send a series of tips on choosing a product or as a guide to using something they just bought.

Here is the difference to remember right away: recipients enter a series through a tag, not by being selected in a list. Whoever gets the tag starts with the first installment, even if the series has been running for half a year. That makes the series a great fit anywhere new people keep arriving over time.

A/B testing: measurement, not guesswork

A campaign with A/B testing prepares one message in two versions, sends them to small groups of recipients, evaluates the results after a set period and automatically sends the winning version to the rest of the database. With this type you stop guessing whether a shorter or longer subject line works better and start knowing it, based on your own database.

Keep in mind that the minimum number of recipients for a campaign with A/B testing is eleven. On a small database the test makes no sense, the difference between versions would be random.

Logic campaign: branching by behavior

A logic campaign sends follow-up emails based on whether the recipient clicked a tracked link. Whoever clicked gets the message from the "Yes" branch, whoever did not click gets the message from the "No" branch. The structure can have up to three levels.

A typical real-world example: you send an invitation to a workshop and a few days before the event a reminder goes out only to those who have not signed up yet. People who already clicked the button get practical information about the event instead of a reminder.

What to prepare in advance

Advanced campaigns are hard to improvise. Before you create one, have three things ready: a list of all the messages you want to send, a clear rule for who gets which one, and the tags you will use to select recipients. Sketch it out on paper. For a three-level logic campaign that is not excessive caution, it is a necessity.

One last practical note. Quanda lets you have thirty-five active campaigns in the "ready to use" state at the same time, which you can recognize by the checkmark in a green circle. Series and logic campaigns run for a long time and will fill this number faster than you expect. Get into the habit of ending or pausing finished campaigns in the campaign overview under Actions, so the limit does not surprise you at the moment you need to launch something new.

The email series is the workhorse of advanced campaigns. You prepare several messages, set a rhythm and Quanda then sends them to every recipient starting from the first installment, no matter when they entered the series. In this lesson we will walk through the whole series from selecting recipients to launch, and we will look at the feature that lets you chain series together.

Recipients: a tag instead of selecting contacts

You create a series in the top menu under Email campaigns via "Create new campaign", name the campaign and click "Create" in the "Email series" section. You know the "Settings" page from the simple campaign, this is where you choose the language of helper texts and invite colleagues if needed. Once the series has been sent, these settings can no longer be changed.

The "Recipients" page, however, differs fundamentally from the simple campaign. You do not pick individual contacts from a list, you choose one or more tags and click "Save". The recipients of the series become all contacts that have any of the chosen tags. If you chose a brand new tag, you will see a zero, which is fine. The contacts will appear the moment you assign them the tag.

You can assign a tag manually, through an import or through an automation in a form, questionnaire or another campaign. This is exactly where the power of the series lies: you trigger it by a form submission, a specific answer in a questionnaire or a click in a previous campaign. New contacts keep being added even while the series is running, and every newly added contact starts with the first installment.

You always have an overview on the Recipients page. For each installment you see the total number of recipients, the number of recipients for the next mailing and the number of currently blacklisted contacts. Use the "Show list" link to display the waiting contacts.

Editor: up to one hundred fifty installments and a smart shortcut

The email editor of a series contains a list of all installments, and you can create up to one hundred fifty of them. You build the individual emails the same way as in a simple campaign.

A proven way to save yourself work: polish the first installment into its final form including the header, footer and colors, save it as a template, and only then click "Save and add next". Create the following installments from this template and change only the content. You will save hours and all installments will look consistent. Installments can be moved, copied and deleted at any time.

Tags in action: how to connect series

In the "Actions" column you will find the feature that turns a series into a building kit. For each installment you can set a tag to be added to or removed from the contacts that have just passed through it.

• Adding a tag on the last installment enrolls the contact in a follow-up series.
• Adding a tag in the middle starts a parallel series on a different topic.
• Removing a tag takes the contact out of a series that is currently running.

This way you can assemble long-term contact nurturing from several shorter series, which are easier to maintain than one giant series.

Sending and rhythm

On the "Sending" page, first choose the sender. You have two options. Either a specific user with a verified email address. Or the automatic setting, where Quanda changes the sender based on the contact owner. When no contact owner is stored, the company owner steps in, followed by the campaign manager. For smooth delivery you also need a verified sending domain.

Then run the spam test. The result is a number from zero to ten and the closer to zero, the better. The most common offenses are a subject line written in capital letters, unsuitable words in the subject, an email body made up of a single image and weak contrast between text and background.

Send yourself test emails of all installments and read them back to back as one story. Only then switch the series to the "Ready to send" state, choose the start day and time and set the frequency. Monday at 2 p.m. with a frequency of seven days means the next installment goes out every Monday at 2 p.m.

Most email marketing advice ends with the sentence "it depends on your database". A/B testing is the tool that turns this sentence into a concrete number. You prepare two versions of one email, a small portion of recipients receives them, Quanda evaluates the results after a set period and sends the winning version to the rest on its own. In this lesson we will look at what is worth testing and how to set up the test so its result actually means something.

First define the goal of the test

Before you even open the editor, decide what you want to improve. You have two reasonable options and each leads to different changes in version "B".

• Open rate. You leave the email content unchanged and change only the subject line and the preheader. That way you test the single thing that decides whether the email gets opened.
• Conversion, meaning clicks on links. You keep the subject and change the number, color or text of buttons and links.

The fundamental rule is: do not change too many things at once. If in version "B" you rewrite the subject, swap the image and recolor the button, you will learn that it won, but not why. One change per test is boring, but it is the only kind you can learn from.

Creating the campaign and recipients

You create the campaign via "Create new campaign", name it and click "Create" in the "A / B testing" section. The "Settings" page works the same as elsewhere, you choose the language of helper texts, the campaign manager and optionally invite other users.

You select recipients the classic way through tags, combinations of tags and optionally the detailed filter, just like in a simple campaign. The minimum number of recipients for a campaign with A/B testing is eleven. That is a technical threshold, not a recommendation. With a few dozen contacts the difference between versions will tell you very little, because one extra open moves the percentages by whole points.

Two versions in the editor

First create the version "A" email exactly as you would send it normally. Then click "Save and create version B". Quanda copies the version and you make one planned change in it according to your chosen goal. Review both versions in the preview and continue to the "Sending" page.

Test settings and evaluation

On the sending page you choose the sender, the reply-to address and run the spam test, all of which you already know. What is new is the "A / B testing settings" section.

With the slider you set how large a group of recipients takes part in the test and how large a portion of the database waits for the winning variant. Then you choose the winning criterion, for example a higher percentage of opened emails, and the time period over which the results are evaluated. In practice, one day, meaning twenty-four hours, proves to be the most reliable period. A shorter interval only shows you the fastest readers, which tends to be misleading.

Do not forget the test emails of both versions. After checking them, switch the campaign to the "Ready to send" state and send it immediately or schedule it. By scheduling you set not only the time of the first emails but, depending on the test duration, also the moment the winning version goes out.

Here is what happens after sending: based on the parameters you set, Quanda picks group "A" and group "B", sends them the respective versions and after the set time evaluates the results. It then automatically sends the winning version to all remaining recipients.

You cannot learn from a single test

An A/B test pays off as a habit, not as a one-off event. Test the subject line length in one campaign, addressing recipients by first name in the next, the color of the main button in another. After half a year of such tests you have your own list of what works on your database, and that is more valuable than any general recommendation.

Record the results outside Quanda, for example in a shared spreadsheet. The test setup, the hypothesis, the sample size and the outcome. A year from now you will not remember any of those tests, and you will not have to ask the same question twice.

The logic campaign is the most complex type you can build in Quanda, and at the same time the only one that can send different emails based on how the recipient behaved. The deciding signal is a click on a tracked link. Whoever clicked continues down the "Yes" branch, whoever did not click gets the email from the "No" branch. In this lesson we will go through the whole build, including the two places where mistakes happen most often.

What a logic campaign is good for

The most typical use is an invitation to an event with a fixed date. You send an invitation with a sign-up button and a few days before the event a reminder goes out, but only to those who did not click the button. Those who signed up get practical information about the event instead of the reminder, such as the program or directions.

You can apply the same principle anywhere a click signals interest. Whoever clicked on a specific product gets its details. Whoever clicked nothing gets a different offer or a reminder. Unlike a series, which runs at a fixed rhythm the same way for everyone, here the recipients' paths split.

Recipients and emails

You select recipients the same way as in a simple campaign. Filter contacts by tags, or with the detailed filter, and add them to the campaign by selecting them. You know the blue, green and red numbers at the top of the page, they show selected, ready-to-send and blacklisted contacts. Unsubscribed contacts are not shown here at all.

In the email editor you build the individual messages as usual, but with one key extra step. For the link the campaign should branch on, check the option "Track click for logic campaign" in the link creation section. This tells Quanda which click to watch.

Watch out for the first common mistake. If you do not mark any link in the email, all recipients automatically fall into the "No" branch, no matter how much they click. So double-check the tracked links once your emails are finished. In the last email, which no further email follows, you do not need to mark any links.

You create the next email with the "Save and add next" button. Within one logic campaign you can connect several emails in up to three levels.

Building the tree

On the logic settings page you see all created emails as tiles and below them the tree structure. The email created first sits in the first position in the structure, but even that one can be replaced by another. You drag the tiles with the mouse onto the marked spots in the tree.

The structure always branches into "Yes" and "No". If you leave a slot empty, simply no email is sent there. That is a legitimate scenario, for example when you only want to reach those who did not click.

For each level you set, in the "Send next email after" section, the number of days after which the clicks are evaluated and the follow-up emails are sent.

Time counts from sending, not from the click

Here is the second common mistake, and it is worth remembering. In a logic campaign, time is not counted from the moment of the click but from the date and time the previous email was sent.

An example. You set seven days and the first email goes out on April 1 at 10 a.m. Quanda starts sending the follow-up emails on April 8 at 10 a.m., based on who clicked and who did not. Whether someone clicked right away on the afternoon of April 1 and someone else only on the evening of April 7 makes no difference, both get the email from the "Yes" branch at the same moment. Likewise, the email from the "No" branch goes out at that same moment, if you set one up.

This is what distinguishes a logic campaign from an automation in a simple campaign, where time runs from the event for each individual recipient. When you need to react to a click within minutes, reach for an automation. When you are building a scenario with fixed dates, the logic campaign is the right choice.

Checking and launch

Before launching, send yourself test emails of all the messages, including those from the deeper levels, and click through the links. In test emails the unsubscribe function is intentionally disabled and personalization is displayed based on the first recipient on the Recipients page.

Then switch the campaign to the "Ready to send" state and send it immediately or schedule it. You then get statistics for each email in the campaign separately.

Advanced campaigns have more complex statistics than a simple campaign, because each type measures something slightly different. A series tracks individual installments, A/B testing compares versions and a logic campaign evaluates each email in the tree separately. In this final lesson we will look at where the numbers hide and how to turn them into a concrete improvement in your next campaign.

Terms that apply everywhere

The basic overview looks the same in all campaign types, so you only need to learn the terms once.

• Sent, delivered and undelivered count unique recipients. The list of undelivered emails, including the reason, is at the bottom of the statistics page.
• Total opens counts every single open. If one recipient opens the email ten times, it shows ten. Unique opens counts people, so the same case counts as one.
• Total clicks and unique clicks work the same way for links.
• Marked as SPAM and unsubscribed are two numbers to watch even when they are small. A rising trend is a warning that you are sending too often or to people who are not interested in the topic.

Always look at the unique values when comparing campaigns with each other. The totals can be inflated by a few very active readers.

Series: find the installment where people drop off

Series statistics are evaluated for each installment separately and they are cumulative. Different groups of recipients move through the series over time, so one installment gets sent many times and the overview shows the sum across all of its sendings.

In practice this means you cannot compare installments by absolute numbers, because later installments have fewer sendings behind them. Compare percentages. When the open rate holds steady for the first two installments and drops by half at the third, that is exactly where your problem is. The usual culprit is a weak subject line or an installment that fails to deliver on the promise of the previous message.

A/B testing: four views of one campaign

A campaign with A/B testing has four sets of statistics. A separate one for version "A", a separate one for version "B", a comparison of both versions and overall results for all sent emails including the send-out of the winning variant.

The comparison is the most useful. Right at the top you see which version won, what its subject line was and what time interval you set for determining the winner. It is followed by the main statistics of both emails in a single chart, a percentage comparison and the development over time for each version separately. At the very bottom you will find a comparison of clicks on individual unique links, which is the best place to look when you were testing buttons.

Be strict when reading the result. A difference of a few tenths of a percent on a small test group is not a victory, it is chance. Only treat a difference as a finding once it repeats.

Logic campaign: read the branch ratio

In a logic campaign you display the statistics by clicking a specific email in the tree structure. Each email has its own overview with the same terms as everywhere else.

But the number you built the campaign for is not shown directly in the overview. You calculate it from the recipient counts of the follow-up emails. How many people continued down the "Yes" branch and how many down the "No" branch is effectively the level of interest in your offer. Track that ratio across campaigns. If it keeps improving over time, you are doing something right.

An evaluation rhythm and cleanup

The statistics of every email can be exported to an xlsx file, with the individual overviews split into sheets. Take advantage of that and keep your own spreadsheet of results. Decisions are not made by a single number but by how it develops over time.

Establish a simple rhythm. After every campaign, go through the list of undelivered addresses, fix typos and delete permanently blacklisted contacts. Once a quarter, review the running series and logic campaigns to check they still match what you offer.

While you are at it, end or pause the campaigns that have served their purpose. Otherwise the limit of thirty-five active campaigns in the "ready to use" state will stop you at exactly the moment you are in a hurry.